Transmission Systems Design Handbook for Wireless Networks
For a listing of recent titles in the Artech House Mobile Communications Series, turn to the back of this book.
Transmission Systems Design Handbook for Wireless Networks Harvey Lehpamer
Artech House Boston London www.artechhouse.com
Library of Congress Cataloging-in-Publication Data Lehpamer, Hrvoj (Harvey). Transmission systems design handbook for wireless networks/Hrvoj Lehpamer. p. cm.(Artech House mobile communications series) Includes bibliographical references and index. ISBN 1-58053-243-8 (alk. paper) 1. Wireless communication systems. 2. Cellular telephone systems. 3. RadioTransmitters and transmission. I. Title. II. Series. TK5103.2.L45 2002 621.3456dc21 2001056655
British Library Cataloguing in Publication Data Lehpamer, Hrvoj (Harvey). Transmission systems design handbook for wireless networks. (Artech House mobile communications series) 1. Wireless communications systemsDesign I. Title 621.38456 ISBN 1-58053-243-8 Cover design by Igor Valdman
© 2002 ARTECH HOUSE, INC. 685 Canton Street Norwood, MA 02062
All rights reserved. Printed and bound in the United States of America. No part of this book may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording, or by any information storage and retrieval system, without permission in writing from the publisher. All terms mentioned in this book that are known to be trademarks or service marks have been appropriately capitalized. Artech House cannot attest to the accuracy of this information. Use of a term in this book should not be regarded as affecting the validity of any trademark or service mark. International Standard Book Number: 1-58053-243-8 Library of Congress Catalog Card Number: 2001056655 10 9 8 7 6 5 4 3 2 1
Dedicated to the two most important women in my life, my mother and María.
Contents Acknowledgments
xix
1
Introduction
1
2
Basics of Wireless Networks
5
2.1
Historical Background
5
2.2 2.2.1 2.2.2 2.2.3 2.2.4 2.2.5 2.2.6 2.2.7
Existing Wireless Technologies Mobile Networks and the Wireless Local Loop Analog Cellular Systems FDMA TDMA Personal Digital Cellular GSM and General Packet Radio Services CDMA
6 6 7 9 9 10 10 13
2.3 2.3.1 2.3.2
Evolution of Wireless Technology 1G and 2G Wireless Networks 3G Wireless Networks
15 15 15
vii
viii
Transmission Systems Design Handbook for Wireless Networks
2.3.3 2.3.4 2.3.5
CDMA2000 CDMA2000 1xEV-DO Future Directions
19 22 24
2.4 2.4.1 2.4.2
Satellite Networks Fixed Satellite Service Mobile Satellite Systems
26 26 27
2.5 2.5.1 2.5.2
Fixed Microwave Systems Microwave Point-to-Point Systems Microwave Point-to-Multipoint Systems References
28 28 31 44
3
Transmission-Network Principles
47
3.1 3.1.1 3.1.2 3.1.3 3.1.4
Wireline Side of Wireless Networks PSTN Interconnect and Telephony Overview Traffic Engineering SS7 and AIN Telecommunications Act of 1996
47 47 51 54 58
3.2 3.2.1 3.2.2 3.2.3 3.2.4 3.2.5 3.2.6
Digital Transmission Technology About the Transmission Transmission MediaPhysical Layer Transmission (Backhaul) in Wireless Networks DSX-1 Digital Interfaces (North America) North American Digital Hierarchy CEPT Digital Hierarchy
59 59 59 61 62 63 65
3.3
Plesiochronous Versus Synchronous Digital Hierarchy 66
3.4 3.4.1
Multiplexing and Inverse Multiplexing Statistical Multiplexing
70 72
3.4.2 3.4.3
3/1 Multiplexing and Subrate Multiplexing Inverse Multiplexing
73 74
3.5
ATM
79
3.5.1
ATM Basics
79
Contents
ix
3.5.2 3.5.3 3.5.4 3.5.5
Use of ATM Adaptation Layer Inverse-Multiplexing for ATM Protocol QoS in ATM Networks Definition of Availability in ATM
83 86 87 89
3.6 3.6.1 3.6.2 3.6.3 3.6.4
Voice over IP H.323 Network Building Blocks Latency and Jitter Issues Multiprotocol Label Switching IP-Based Wireless Networks
91 91 95 97 99
3.7 3.7.1 3.7.2 3.7.3 3.7.4 3.7.5 3.7.6 3.7.7 3.7.8 3.7.9 3.7.10 3.7.11 3.7.12 3.7.13 3.7.14 3.7.15 3.7.16
Complete T1 Tutorial Signals in a T1 Network Pulse Transmission BERs Overall System Length Single-Cable and Dual-Cable Operation T1 Repeatered Lines Order Wire Lightning Simplex Power Design Data Error Rates Voltage and Temperature Factors T1 Engineering, Installation, and Documentation Troubleshooting and Problem Classification Switch Options, Line Codes, and Framing Fractional T1 T1 (J1) in Japan
102 102 105 105 106 107 107 110 110 112 114 115 116 116 118 118 118
3.8 3.8.1 3.8.2 3.8.3 3.8.4 3.8.5 3.8.6
Complete E1 Tutorial Introduction to E1 Networks Customer Premises Equipment Signal Characteristics Transmission Facilities Pulse Density HDB3
119 119 119 120 120 121 121
x
Transmission Systems Design Handbook for Wireless Networks
3.8.7 3.8.8 3.8.9 3.8.10
E1 Framing Synchronization E1 Framing Formats Spare Bits Global Framing Formats References
122 122 124 127 127
4
Wireless-Network Architecture
129
4.1
2G Wireless-Network Architecture
129
4.2 4.2.1 4.2.2 4.2.3 4.2.4 4.2.5 4.2.6 4.2.7
3G Wireless-Network Architecture Directions in 3G Developments Horizontally Layered Network Architecture 3G Core Network Universal Mobile Telephone System 3G in GSM Networks 3G CDMA Network 3G Traffic Classes
131 131 136 140 141 141 142 145
4.3 4.3.1
3G Transmission Networks Replacing TDM with ATM in Transmission Networks Importance of AAL2 QoS Concept ATM Physical Layer Traffic Modeling and Simulation Tools 2G and 3G Coexistence Transmission-Network Architecture References
148 148 154 158 161 165 168 170 179
5
Theory and Principles of Fiber-Optic Transmission
181
5.1
Basics of Fiber-Optic Transmission
181
5.2 5.2.1 5.2.2
Design Principles Bandwidth and Attenuation Optical Power Budgets and Distance Calculations
183 183 185
5.3
Synchronous Digital Hierarchy
189
4.3.2 4.3.3 4.3.4 4.3.5 4.3.6 4.3.7
Contents
xi
5.3.1 5.3.2 5.3.3 5.3.4 5.3.5
Basics of Synchronous Systems Benefits of SONET SONET Architecture SONET Availability Requirements SDH
189 191 191 195 195
5.4 5.4.1 5.4.2 5.4.3 5.4.4 5.4.5
DWDM DWDM Overview DWDM Capacity (Bandwidth) Requirements Network Growth and Flexibility of DWDM Optical Layers Protection in DWDM Networks
196 196 197 198 200 200
5.5
Optical Switching References
201 202
6
Microwave Point-to-Point System Design
203
6.1
Basic Microwave Transmission Theory
203
6.2 6.2.1 6.2.2 6.2.3 6.2.4 6.2.5 6.2.6
Theoretical Aspects of Microwave Link Design Microwave-Radio-Path Calculation Overview Design Fade Margins Diversity Improvement North American and ITU Objectives Reliability and Availability Overview Effects of Rain on Microwave Propagation
206 206 207 210 212 215 217
6.3 6.3.1 6.3.2 6.3.3 6.3.4 6.3.5
Practical Aspects of Microwave-Link Design Design Overview Protected and Nonprotected Microwave Systems Microwave Repeaters Microwave Path Calculations Microwave Interference Analysis and Frequency Coordination Microwave System Design Guidelines Microwave Lookup Table
219 219 221 222 224
6.3.6 6.3.7
225 226 229
xii
Transmission Systems Design Handbook for Wireless Networks
6.4
Spread-Spectrum Microwave Systems
230
6.5
Microwave Compatibility and Safety
236
6.6 6.6.1 6.6.2 6.6.3 6.6.4 6.6.5
Coordinate Systems, Datums, and GPS About Datums Geometric Earth Models Reference Ellipsoids and Coordinate Systems GPS Useful Facts to Remember
240 240 241 242 244 247
6.7 6.7.1 6.7.2
Managing the MW Radio Network Introduction Managing a Microwave Network with SNMP References
247 247 249 254
7
Transmission-Network Planning and Design
257
7.1
Overview
257
7.2 7.2.1 7.2.2 7.2.3 7.2.4 7.2.5 7.2.6 7.2.7 7.2.8 7.2.9 7.2.10 7.2.11 7.2.12 7.2.13
General Wireless-Network Planning and Design Principles Identifying the Opportunity and Strategic Planning Customer Requirements Analysis Interconnection Spectrum Auctions Clearing Spectrum and Microwave Relocation RF Design Transmission Media and Topology Planning Mobile Positioning Toll QoS Network Performance Sales and Marketing Regulatory Issues Life Cycle of Wireless Networks
259 259 262 265 266 267 268 270 271 272 273 276 278 281
7.3
Transmission System Design
283
Contents 7.3.1
xiii
7.3.2 7.3.3 7.3.4 7.3.5
Transmission System Design Process and Requirements SDH and PDH Transmission Systems SDH Transmission-Network Protection Ring Protection in the Wireless Network Description of TND Deliverables
283 286 286 288 289
7.4 7.4.1 7.4.2 7.4.3 7.4.4 7.4.5 7.4.6 7.4.7
Leased Lines in Wireless Networks Leased Access and Core Transmission Networks Dedicated Leased Service xDSL Switched Leased Service Higher-Speed Switched and Nonswitched Services Leased Lines Network Build Out Owned Versus Leased Transmission Networks
292 293 294 296 297 299 299 301
7.5 7.5.1 7.5.2 7.5.3
302 302 304
7.5.4 7.5.5 7.5.6 7.5.7 7.5.8
SynchronizationStratum, BITS, and GPS Introduction and Historical Overview Strata General Timing Planning Rules in Transmission Networks Interoffice Distribution Intraoffice Distribution SONET Network Timing Synchronization: Issues in PCS Networks Cell-Site Timing in Wireless Networks
308 310 312 313 314 316
7.6 7.6.1 7.6.2 7.6.3 7.6.4
Transmission-Network Optimization Daisy Chaining and Traffic Grooming Voice Compression Signal Propagation Delay Example of Optimized Network Design
317 317 318 321 323
7.7 7.7.1 7.7.2
Transmission Network: Design Examples Small PDH Microwave Transmission Network Complex Transmission Network
324 324 325
xiv
Transmission Systems Design Handbook for Wireless Networks
7.8
Overview of RNC Dimensioning in the 3G Wireless Network Traffic Classes in the UTRAN Network Description of RNC Interfaces User Traffic Modeling Traffic Calculation Guidelines Example of Traffic Calculations
327 328 329 332 334 342
7.9 7.9.1 7.9.2 7.9.3
Alternative Solutions in Transmission Networks Considering Dark Fiber and Dark Copper Partnership with Utilities Optical Laser Communications References
344 344 344 348 350
8
Transmission Equipment
351
8.1 8.1.1 8.1.2 8.1.3 8.1.4 8.1.5 8.1.6 8.1.7
Digital Microwave Radio PDH and SDH Microwave Radios Standard Microwave Radio Configuration Split Microwave Radio Configuration Microwave Antennas Transmission Lines Environmental and Quality Issues Standards and Recommendations
351 351 352 353 358 361 362 363
8.2 8.2.1 8.2.2
Fiber-Optic Equipment SONET and SDH OPGW
363 363 368
8.3 8.3.1 8.3.2 8.3.3
Wireline Equipment Digital-Access Cross Connects CSU/DSU DSL and ADSL
370 370 375 376
8.3.4
Echo Cancellers
379
8.4 8.4.1
Cabling NEC Cable Categorization
384 384
7.8.1 7.8.2 7.8.3 7.8.4 7.8.5
Contents
xv
8.4.2 8.4.3 8.4.4
Digital Cross Connects DS1 Signal Termination Leased Lines and the Network Interface Unit
386 388 392
8.5 8.5.1 8.5.2 8.5.3 8.5.4
Grounding Earth Grounding Basics Ground System Design Fundamentals Types of Grounding Grounding for Wireless Cell Sites
392 392 395 396 398
8.6 8.6.1 8.6.2 8.6.3 8.6.4
Power and Battery Backup ac Power dc Power Batteries Solar Energy
399 399 400 401 403
8.7
GPS Antennas
405
8.8 8.8.1 8.8.2 8.8.3 8.8.4 8.8.5 8.8.6
Quality and Reliability Issues Quality Assurance First Article Inspection Factory Acceptance Testing Equipment Reliability Environmental Specifications Network Equipment Building Standard References
405 405 406 407 407 410 411 412
9
Transmission-Network Deployment
413
9.1 9.1.1 9.1.2 9.1.3 9.1.4 9.1.5
Equipment and Services Ordering Process Planning and Design RFQs, RFIs, and RFPs Negotiating the Statement of Work Negotiating with Telecommunications Providers Negotiating with Equipment and Services Suppliers
413 413 414 421 422 425
9.2
Regulatory Issues
425
xvi
Transmission Systems Design Handbook for Wireless Networks
9.3 9.3.1 9.3.2 9.3.3 9.3.4 9.3.5
Services Engineering Services Project Management Outsourcing Services Due Diligence Network Maintenance
430 430 430 432 435 436
9.4 9.4.1 9.4.2 9.4.3 9.4.4 9.4.5 9.4.6 9.4.7
Project Management in Wireless Networks Definitions Project-Management Organizational Issues Project Stages Leased-Lines Tracking Process Change Orders Postinstallation and Optimization Activities Project-Management Tools
436 436 438 440 441 443 443 445
9.5 9.5.1 9.5.2
446 446
9.5.3 9.5.4 9.5.5
Selection of Key Sites Site Acquisition, Zoning Issues, and Colocation Switch and NOC Site Selection and Building Requirements Cell-Site Selection Microwave Repeater Site Selection Colocated Systems and RF Cell-Site Compliance
449 451 452 453
9.6 9.6.1 9.6.2 9.6.3 9.6.4 9.6.5
Microwave Deployment Microwave System Scope of Work Microwave Site Surveys Microwave Path Survey Housing the Infrastructure Microwave Antenna Mounting Structures
454 454 456 461 464 466
9.7 9.7.1 9.7.2 9.7.3 9.7.4 9.7.5
Measurement of Radio-Frequency Fields Health and Safety Issues Measurements and Sources of Emission Near-Field and Far-Field Regions Power Levels and Power Density Radiation Patterns and Polarization
485 485 486 488 489 490
Contents
xvii
9.7.6 9.7.7 9.7.8 9.7.9
Sources of Electromagnetic Field Induced and Contact Currents Instrumentation Measurement Procedure for Microwave Installations
491 492 493 494
9.8 9.8.1 9.8.2 9.8.3 9.8.4 9.8.5 9.8.6 9.8.7 9.8.8
Fiber-Optic Cables and Their Installation Cabling Design Considerations Fiber Protection Fiber-Optic Cable Types Fiber Count Fiber Splicing Connectors Handling Fiber-Optic Cables Fiber-Optic Cable Installation Procedures
495 495 497 498 501 503 504 506 507
9.9 9.9.1 9.9.2 9.9.3 9.9.4
Operations and Maintenance Growth of Multiservice Networks Network Management System Geographic Partitioning Location-Finding Techniques References
509 509 511 516 516 520
10
Transmission-Network Testing and Commissioning 521
10.1
Definitions
521
10.2 10.2.1 10.2.2 10.2.3 10.2.4
BERT T1 Impairments T1/E1 Testing Out-of-Service Testing In-Service Monitoring of Live Data
522 522 523 524 526
10.3 10.3.1 10.3.2 10.3.3 10.3.4
Transmission-Network Testing Procedure Testing Leased Facilities Testing Microwave Systems DS1 and DS3 Performance Objectives DS1 Test Procedure and ATP Form
526 526 527 528 529
xviii
Transmission Systems Design Handbook for Wireless Networks
10.4 10.4.1 10.4.2 10.5 10.5.1 10.5.2 10.5.3
Fiber-Optic Cable Testing 530 OTDR Test Procedure for Single-Mode Fiber-Optic Cables 531 Characterizing and Testing Fibers for DWDM Applications 532 Packet-Network Testing Testing the Voice Traffic ATM Network Testing Quality Tests in VoIP Networks References
537 537 538 543 547
Appendix: Units Conversion
549
A.1
About Units of Measurement
549
A.2
The International System of Units
550
A.3
Common Units
552
Glossary
553
About the Author
587
Index
589
Acknowledgments I would like to thank my colleague Alfiya Vali-Durrett, a great engineer, for patiently reading and contributing to a number of the chapters in this book. Many thanks also to Dave Olson for saving the electronic version of the manuscript from disappearing into the twilight zone of my crashing computers hard drive. In addition, I would like to express my appreciation to everyone with whom I was unable to spend time during the past 3 years because I was working on this book. Thank you all.
xix
1 Introduction Public telephone operators and new independent wireless operators throughout the world are deploying wireless access in an effort to drastically reduce delivery costs in the most expensive part of the networkthe local loop. Available radio technology enables both existing and new entrants to access subscribers in a rapid manner and deliver their basic telephony products and broadband-enhanced services. While each operator has his or her own series of business, regulatory, and technical drivers, the following questions are common: • What services should be delivered to the various customer segments? • Which technology strategy should be adopted: TDMA, GSM,
W-CDMA, or CDMA 2000? • How can the network be implemented and service provided in a fast-track manner? • Which programs, processes, and suppliers must be established to realize the business plan? • How can the new wireless provider compete against other wireline and wireless operators, and what are the key factors for success? Issues, such as technology choices, security aspects, network quality, marketing, and customer service, are essential to long-term success, affecting subscriber installation equipment, program management, opportunities in 1
2
Transmission Systems Design Handbook for Wireless Networks
emerging nations, license obligations, interconnect strategy, billing systems, and revenue collection. Prior to the license application, it is imperative to formulate the strategic elements of the business plan. The strategy must be developed and agreed upon using a top-down approach, with realistic time scales and budgets, and identified risks. The vital elements of the corporate strategy include personnel, technology, network topology, procurement strategy, regulatory strategy, and project management. Typically, a business plan is based on marketing data and, once the license is granted, implementation. Key issues include geographic rollout, communications links, site requirements, service offerings, and launch strategy. Working closely with the marketing plan, the overall coverage plan and capacity requirements should be established. In line with the chosen technology, the engineering details and specifications need to be generated, agreed upon, and approved. Physical elements, such as structures, antennas, and cabinets, should be professionally specified to ensure expert purchasing. Design activities must be addressed in link planning, transmission networks, switch capacity, and numbering plans. Site surveys have to be performed from both microwave-link and RF-coverage aspects, to ensure exact design. Site acquisition is usually the gating factor in most wireless networks build-out timelines. Once these sites are acquired and planning permission is obtained, the development process (build-out) will accelerate. Some of the key issues here are building services, type of site, zoning issues, access, power, site management, and landlord liaison. Equipment procurement is dependent on the performance of outside agencies (i.e., equipment and systems suppliers). Consequently, there is obvious risk, and a number of key issues will need to be addressed, including competitive tendering, contract management, software development, and change management. Strategic alliances with equipment suppliers and service providers must also be considered. Constructed sites are usually tagged as ready for communications. There is intense activity surrounding installation of all the necessary equipment following the logistical challenge of materials management. At this time, given the pace of the project, it is typical to find numerous sites being developed simultaneously. Some of the key issues are equipment deliveries, rigorous testing procedures, commissioning, optimization, patenting, and integration. Transmission, or transport, design and its integration with the existing networks are important parts of the telecommunications network, as they allow for the transporting of the traffic between various elements of the
Introduction
3
network and provide access to the customer, while enabling connectivity between various networks. The transmission media has been traditionally categorized into three main typescable (copper or fiber-optic), radio (microwave point-to-point, in this case), and satellite. Although the technologies for the backbone network are already well developed, there remain costs and difficulties involved in setting up the local access; the technologies for overcoming these difficulties are continuing to evolve. The question for wireless operators is whether to build and own the transmission network or to lease facilities (lines) from the existing carriers or operators; the answer will be different in different situations. The introduction of third-generation (3G) wireless networks with increased capacity requirements and packet data architecture will also have a great impact on future transmission-network design and deployment. The issues involved in the deployment of various transmission technologies and their impact on the overall wireless-network topology will be presented in this book. Overall strategy and approach to transmission-network planning, design, and deployment will be discussed, but the details of the fiber-optic and microwave designs are outside the scope of this book. Transmissionnetwork design in wireless networks is almost technology independent, although there are differences between the TDMA and CDMA counterparts. This book covers a number of wireless technologies for the purpose of introducing readers to the overall aspects of wireless-network design; it focuses then on the transmission aspects. Many of the recommendations, tables, and descriptions in this book are the result of the authors experience of over 20 years in the engineering field and cannot be traced to any particular standard, book, or article. Regardless of the technology, the transmission network, including its physical layer and issues related to it, is an important part of every wirelessnetwork build-out. In this book, the importance of transmission engineering within wireless networks is given the attention it deserves.
2 Basics of Wireless Networks 2.1 Historical Background The first commercially available radio and telephone system, known as the Improved Mobile Telephone Service (IMTS), was put into service in 1946. This system was quite unsophisticated, as there were no solid-state electronics available at that time. In order to use IMTS, a tall transmitter tower was erected near the center of a metropolitan area. Several assigned channels were then transmitted and received from the antenna atop this tower and any vehicle within range could attempt to seize one of those channels and complete a call. Given these constrictions, the number of channels available did not come close to satisfying the demand. The solution to this problem was cellular radio. Metropolitan areas were divided into cells of no more than a few miles in diameter, each cell operating on a set of frequencies (sending and receiving) that differed from the frequencies of the adjacent cells. Because the power of the transmitter in a particular cell was kept at a level just high enough to serve that cell, these same sets of frequencies could be used at several places within the metropolitan area. Two important characteristics of cellular systems contribute to their usefulness. The first is their controlled handoff. As subscribers are driving out of one cell and into another, their mobile (subscriber) units, in conjunction with sophisticated electronic equipment at the cell sites [also known as base stations(BS)] and the telephone switching offices [also known as mobile switching centers, (MSCs)], are transferred from one frequency set to 5
6
Transmission Systems Design Handbook for Wireless Networks
another with no audible pause. Second, systems were designed to locate particular subscribers by paging them in each of the cells. After locating the vehicle in which the paged subscriber was riding, the equipment assigned sets of frequencies to it and conversation could begin. The initial transmission technology used between the vehicle and the cell site was analog in nature and known as Advanced Mobile Phone System (AMPS); the analog scheme used was called frequency-division multiple access (FDMA). Digital transmission was later developed, resulting in timedivision multiple access (TDMA). In Europe, the selected scheme was an adaptation of the TDMA used in the United States, called Groupe Speciale Mobile (GSM), now known as Global System for Mobile Communications. A few years later, a third group of companies (led by Qualcomm, Inc.) developed spread-spectrum technology called code-division multiple access (CDMA). Thus, there are at least four schemes that may be used for communications between a vehicle and the cell site. Communications between the cell site and the switch utilize transmission media such as microwave, copper pairs, and fiber optics. The continuing growth of cellular communications led government and industry in the United States to search for additional ways to satisfy the obvious need not only for ordinary telephone service but also for special services and features, smaller telephones, and cellular phone use. This search led to the development of the personal communications service (PCS) industry. Additional frequency bands were allocated for their use, and rather than assign them to the first comers or by way of a lottery, the Federal Communications Commission (FCC) [1] auctioned them off through a sophisticated bidding contest that brought the U.S. Treasury billions of dollars. Unfortunately, after acquiring the spectrum, many of the new potential operators went bankrupt and never actually had a chance to use it. In Canada, the decision was based on a beauty contest, a comparison of business cases, technology, and future plans, rather than the bidding process.
2.2 Existing Wireless Technologies 2.2.1
Mobile Networks and the Wireless Local Loop
Mobile communication is defined as a system in which the originator of the message, its recipient, or both, is in motion. User movements in mobile networks affect the communications system in many waysfrom the channel behavior in the physical layer to the system performance in the higher layers [2].
Basics of Wireless Networks
7
Wireless networks are not necessarily mobile; they can also be fixed in nature. Sometimes called radio in the loop (RLL) or fixed radio access (FRA), wireless local loop (WLL) is a system that connects subscribers to the Public Switched Telephone Network (PSTN) using radio signals as a substitute for copper for all or part of the connection between the subscriber and the switch. This connection includes cordless access systems, proprietary fixed radio access, and fixed cellular systems. In WLL application, the subscriber units are stationary without roaming capabilities of cellular (800 MHz) or PCS (1,900 MHz) services. Fixed local loop connections utilizing wireless multiple access networks are rapidly emerging as a viable alternative to cable-based networks for providing voice and data connection to subscribers in countries both with and without mature and established telecommunications networks. Much of the growth will occur in emerging economies where half the worlds population lacks plain old telephone services (POTS). Developing nations look to WLL technology as an efficient way to deploy POTS for millions of subscriberswithout the expense of burying tons of copper wire. In the more developed economies, WLL can help unlock competition in the local loop, enabling new operators to bypass existing wireline networks to deliver POTS and data access. The penetration of WLL into developed countries is assumed to be much lower than in emerging countries. The telephone companies are able to keep up with the demand for new POTS lines, thereby removing unserved demand as an issue. The requirement for WLL in developed countries will come from companies that want to bypass the established local phone companies or customers who want the additional services that WLL can provide. There is a significant difference in network design for the mobile wireless networks, fixed wireless networks, or a combination of the two, as equipment, quality requirements, and usage of the network resources can vary greatly. 2.2.2
Analog Cellular Systems
2.2.2.1 Principle of the Frequency Reuse
The first-generation cellular systems in operation were analog FM radio systems that allocated a single carrier for each cell. Each carrier was frequency modulated by the caller and the carriers were typically spaced at 25-KHz intervals. The allocated bandwidth was relatively narrow and only few channels (typically 12) were available [3]. Central to the cellular concept is the concept of frequency reuse. Although there are hundreds of channels available, if each frequency were
8
Transmission Systems Design Handbook for Wireless Networks
assigned to only one cell, the total system capacity would be equal to the total number of channels. Adjusting for the Erlang blocking probability results in only a few thousand subscribers per system. By reusing channels in multiple cells, the system can grow without geographical limits. Typical cellular reuse (pre-CDMA, that is) is easily rationalized by considering an idealized system. Assuming that propagation is uniform and that cell boundaries are at the equisignal points, then a planar service area is optimally covered by the classical hexagonal array of cells. Using seven sets of channels, one set in each cell, this unit is then replicated over the service area. While real systems do not ever look like these idealized hexagonal tilings of a plane, the seven-way reuse is typical of that achieved in practice. The capacity of a K-way reuse pattern is simply the total number of available channels divided by K. With K = 7 and 416 channels, there are approximately 57 channels available per cell. At a typical offered load of 0.05 Erlangs (50 mE) per subscriber, each site supports about 1,140 subscribers. Assuming that the cells are using omnidirectional antennas, it might be expected that system capacity is increased by antenna sectorization. Sites are, in fact, sectorized by the operators, usually three ways, with each site equipped with three sets of directional antennas, and their azimuths separated by 120°. Unfortunately, the sectorization does not in practice lead to an increase in capacity. The reason is that the sector-to-sector isolation, often no more than a few decibels, is insufficient to guarantee acceptably low interference. This is due partly to the poor front-to-back ratio of the antennas. The properties of electromagnetic propagation in the real world also conspire to mix signals between sectors. The practical result of sectorization is simply an increase in coverage because of the increased forward gain of the directional antenna. Nothing is gained in reuse. The same seven-way cell reuse pattern of omnidirectional cells applies in sectored cells. Viewed from the standpoint of sectors, in this case the reuse is K = 7 × 3 = 21, not 7. 2.2.2.2 Cellular Digital Packet Data
Cellular digital packet data (CDPD) is designed to provide packet data service on an existing cellular telephone network. First, the basic goal of the CDPD system is to provide data services on a noninterfering basis using existing cellular telephone service with 30-kHz channel spacing. This can be achieved by devoting some of the existing channels to CDPD service. Second, CDPD is designed to make use of cellular channels (those temporarily not being used for voice traffic). Basically, the system is used along with AMPS system, and a possible application is for digital AMPS (D-AMPS). The 30-kHz channels with CDPD support bit rates of up to 19.2 Kbps.
Basics of Wireless Networks
9
The degraded radio channel condition, however, limits actual information payload throughput rates to lower levels such as 5 to 10 Kbps. This will introduce additional time delay due to error detection and transmission protocol in the CDPD radio link physical layer using GMSK modulation at standard cellular frequencies at both forward and reverse links. 2.2.3
FDMA
FDMA is based on frequency division multiplexing (FDM). The digital FDMA systems resemble analog FM, but the carrier is modulated by a digitally encoded speech signal. FDMA allocates a single channel to one user at a time. If the transmission path deteriorates, the controller switches the system to another channel. Although technically simple to implement, FDMA is wasteful of bandwidththe channel is assigned to a single conversation whether or not somebody is speaking. Moreover, it cannot handle alternate forms of data, only voice transmissions. The first generation of analog wireless networks, or cellular networks, were based on this principle. 2.2.4
TDMA
During the late 1980s, the wireless industry began to explore converting the existing analog network to digital as a means of improving capacity. The Cellular Telecommunications Industry Association (CTIA) chose TDMA over Motorolas FDMA (today known as NAMPS) narrowband standard as the technology of choice for existing 800-MHz cellular markets and for emerging 1.9-GHz markets. The TDMA-136 specification, which was defined in the United States in 1988 by the Telecommunications Industry Association (TIA), was developed with the aim of digitizing the analog AMPS. To maintain compatibility with AMPS, the TDMA specification stipulates 30-kHz carrier spacing in a three-slot TDMA solution. TDMA is digital transmission technology that allows a number of users to access a single radio frequency channel without interference by allocating unique timeslots to each user within each channel. The TDMA digital transmission scheme multiplexes three signals over a single channel. The current TDMA standard for cellular divides a single channel into six timeslots, with each signal using two slots, providing a 3-to-1 gain in capacity over AMPS. Each caller is assigned a specific timeslot for transmission. Because of its adoption by the European standard GSM Communications, the Japanese Digital Cellular (JDC), and North American Digital Cellular (NADC), TDMA and its variants are currently the technology of choice
10
Transmission Systems Design Handbook for Wireless Networks
throughout the world. Over the last few years, however, a debate has convulsed the wireless community over the respective merits of TDMA and CDMA. The TDMA system is designed for use in a range of environments and situations, from hand portable use in a downtown office to a mobile user traveling at high speed on the freeway. The system also supports a variety of services for the end user, such as voice, data, fax, short message services, and broadcast messages. TDMA offers a flexible air interface, providing high performance with respect to capacity, coverage, and support of mobility and capability to handle different types of user needs. 2.2.5
Personal Digital Cellular
The development of the Personal Digital Cellular (PDC) specification was drafted by the RCR (1990), which later became the Association of Radio Industries and Broadcasting (ARIB). To ensure compatibility with the Japanese analog systems, a carrier spacing of 25 kHz was maintained in a threeslot TDMA solution. There is very little development going on in this technology today, and it is expected that it will be eventually completely phased out. 2.2.6
GSM and General Packet Radio Services
GSM is a set of ETSI standards specifying the infrastructure for a digital wireless service. The new Pan-European digital cellular standard was developed in 1985. The GSM system, which became operational in 1991, has since evolved into the leading global second-generation (2G) standard, in terms of number of subscribers and area of coverage. To ensure interoperability between countries, these standards address much of the network wireless infrastructure, including the radio interface (900 MHz), switching, and intelligent network (IN). An 1,800-MHz version, DCS 1800, has been defined to facilitate implementation in some countries, particularly the United Kingdom. GSM was designed to provide good speech quality with low-cost service and low terminal cost and at the same time provides a range of new services, including international roaming. GSM is an eight-slot TDMA system with 200-kHz carrier spacing. In terms of service, GSM is a mobile integrated services digital network (ISDN), with support for a wide variety of services. IN support in the mobile environment has also been defined for GSM, for example, the virtual home environment, as well as many advanced data services. GSM subscriber data is
Basics of Wireless Networks
11
carried on a subscriber identity module (SIM) card, which is inserted in the mobile phone to make it work. So, the subscriber potentially has an option of either SIM card mobility or terminal mobility across multiple networks. Today, using General Packet Radio Services (GPRS), packet access can also be integrated into GSM. GPRS is a new nonvoice value-added service that allows information to be sent and received across a mobile telephone network. It supplements todays circuit-switched data and short-message service (SMS) and has connectivity to X.25 and Internet Protocol (IP) networks. GPRS involves overlaying a packet-based air interface on the existing circuit-switched GSM network. This gives the user an option to use a packet-based data service. To supplement a circuit-switched network architecture with packet switching is quite a major upgrade; However, the GPRS standard is delivered in a very elegant mannerwith network operators needing only to add a couple of new infrastructure nodes and make a software upgrade to some existing network elements. Services running on mobile IP packet networks can be divided into horizontal and vertical services, where horizontal refers to e-mail, file transfer, information searching, multimedia, and so on, and vertical refers to applications targeting government, police, transport, medical, business, and the like. Theoretical maximum speeds of up to 171.2 Kbps are achievable with GPRS using all eight timeslots at the same time. This is about three times as fast as the data transmission speeds possible over todays fixed telecommunications networks and ten times as fast as current circuit-switched data services on GSM networks. By allowing information to be transmitted more quickly and efficiently across the mobile network, GPRS may well be a relatively less costly mobile data service compared with SMS and circuitswitched data. GPRS facilitates instant connections whereby information can be sent or received immediately as the need arises, subject to radio coverage and when no dial-up modem connection is necessary. This is why GPRS users are sometimes referred to as being always connected. Immediacy (immediate connection) is one of the advantages of GPRS (and SMS) when compared with circuit-switched data. High immediacy is a very important feature for time-critical applications, such as remote credit card authorization, where it would be unacceptable to keep the customer waiting for even 30 extra seconds. GPRS also facilitates several new applications that have not previously been available over GSM networks due to the speed limitations of circuitswitched data (9.6 Kbps) and message length of the SMS of only 160 characters. GPRS will fully enable the Internet applications you are used to on your
12
Transmission Systems Design Handbook for Wireless Networks
desktop, from Web browsing to chat, over the mobile network. Other new applications for GPRS include file transfer and home automationthe ability to access and control in-house appliances and machines remotely. Packet switching means that GPRS radio resources are used only when users are actually sending or receiving data. Rather than dedicating a radio channel to a mobile data user for a fixed period of time, the available radio resource can be concurrently shared between several users. This efficient use of scarce radio resources means that large numbers of GPRS users can potentially share the same bandwidth and be served from a single cell. The actual number of users supported depends on the application being used and how much data is being transferred. Because of the spectrum efficiency of GPRS, there is less need to build in idle capacity that is only used in peak hours. GPRS therefore enables network operators to maximize the use of their network resources in a dynamic and flexible way, along with allowing user access to resources and revenues. It should be noted here that the GPRS is not only a service designed to be deployed on mobile networks that are based on the GSM digital mobile phone standard. The IS-136-TDMA standard, popular in North and South America, will also support GPRS. This follows an agreement to follow the same evolution path toward 3G mobile phone networks concluded in early 1999 by the industry associations that support these two network types. It should already be clear that GPRS is an important new enabling mobile data service that offers a major improvement in spectrum efficiency, capability, and functionality compared with todays nonvoice mobile services. It is important to note, however, that there are some limitations with GPRS, which are described here. GPRS does impact a networks existing cell capacity. There are only limited radio resources that can be deployed for different usesuse for one purpose precludes simultaneous use for another. For example, voice and GPRS calls both use the same network resources. The extent of the impact depends upon the number of timeslots, if any, that are reserved for exclusive use of GPRS. However, GPRS does dynamically manage channel allocation and allows a reduction in peak-time-signaling channel loading by sending short messages over GPRS channels instead. Achieving the theoretical maximum GPRS data transmission speed of 172.2 Kbps would require a single user taking over all eight timeslots without any error protection. Clearly, it is unlikely that a network operator will allow all timeslots to be used by a single GPRS user. Additionally, the initial GPRS terminals are expected to be severely limited, supporting only one, two, or three timeslots. The bandwidth available to a GPRS user will therefore be severely limited. As such, the
Basics of Wireless Networks
13
theoretical maximum GPRS speeds should be checked against the reality of constraints in the networks and terminals. The reality is that mobile networks are always likely to have lower data-transmission speeds than fixed networks. GPRS is based on a modulation technique known as Gaussian minimum-shift keying (GMSK). GPRS packets are sent in all different directions to reach the same destination. This opens up the potential for one or some of those packets to be lost or corrupted during the data transmission over the radio link. The GPRS standards recognize this inherent feature of wireless packet technologies and incorporate data integrity and retransmission strategies. However, the result is that potential transit delays can occur. Because of this, applications requiring broadcast-quality video may well be implemented using high-speed circuit-switched data (HSCSD). HSCSD is simply a circuit-switched data call in which a single user can take over up to four separate channels at the same time. Because of its characteristic of end-to-end connection between sender and recipient, transmission delays are less likely. Whereas the store-and-forward engine in the SMS is the heart of the SMS center and a key feature of the SMS service, there is no storage mechanism incorporated into the GPRS standard, apart from the incorporation of interconnection links between SMS and GPRS. When a new service is introduced, there are a number of stages before it becomes established. GPRS service developments will include standardization, infrastructure development, network trials, contracts placed, network rollout, availability of terminals, application development, and so on. 2.2.7
CDMA
CDMA is a spread-spectrum technology that allows multiple frequencies to be used simultaneously. CDMA codes every digital packet it sends with a unique key, and the CDMA receiver responds only to that key (Walsh codes) and can pick out and demodulate the associated signal. One of the characteristics of CDMA is that it uses multiple levels of diverse receptionfrequency, spatial, time, and path diversity. When a signal is received, it is received from a number of paths, some directly and others reflected from buildings, mountains, or other obstacles. The technology combines all these signals to create a clear and reliable signal. In March 1992 the TIA established the TR-45.5 subcommittee with the charter of developing a spread-spectrum digital cellular standard. In July 1993 the TIA gave its approval of the CDMA IS-95 standard. IS-95 (also
14
Transmission Systems Design Handbook for Wireless Networks
called cdmaOne) systems divide the radio spectrum into carriers that are 1,250 kHz (1.25 MHz) wide. Unlike GSM and other TDMA-based wireless technologies, CDMA frequency reuse theoretical value is 1. Practical value is close to 1.6 due to interference from users in other cells [4]. CDMA is a spread-spectrum technology, which means that it spreads the information contained in a particular signal of interest over a much greater bandwidth than the original signal [5]. When implemented in a cellular telephone system, CDMA technology offers numerous benefits to the cellular operators and their subscribers: • Capacity increases of 8 to 10 times that of an AMPS analog system
and 4 to 5 times that of a GSM system;
• Improved call quality, with better and more consistent sound as
compared with AMPS systems;
• Simplified system planning through the use of the same frequency in
every sector of every cell;
• Enhanced privacy, meaning low detectability of the transmitted sig-
nal by an intended receiver;
• Improved coverage characteristics allowing for the possibility of
fewer cell sites;
• High tolerance to intentional interference (jamming) or uninten-
tional interference;
• Increased talk time for portables; • Bandwidth on demand.
The narrowband CDMA IS-95 specification stipulates 1.25-MHz carrier spacing for telephony services. Each of the 2G standards essentially defines a mobile telephony system; that is, a system that provides mobile end users with circuit-switched telephony services. Aside from voice services, these systems support supplementary services and some low-bit-rate data services. A number of new flavors of CDMA have been proposed for the next generation of wireless systems (3G); common to all of them is the fact that they are wideband, provide seamless interfrequency handoff, pilot-aided coherent reverse link, and fast closed-loop power control in the forward link [6].
Basics of Wireless Networks
15
2.3 Evolution of Wireless Technology 2.3.1
1G and 2G Wireless Networks
First-generation (1G) wireless cellular systems, analog in nature, introduced the convenience of limited mobility to the consumer. The most widely deployed 1G (analog) mobile phone systems include AMPS, Nordic Mobile Telephone (NMT), and Total Access Communications Systems (TACS). An AMP has major network deployments in North America, the Asia-Pacific region, and Central and Latin America. TACS and NMT first deployments were primarily in Europe (NMT in Scandinavia and TACS in the United Kingdom) and then the Asia-Pacific region. Second-generation systems, digital in nature, consist of GSM, IS-136 or D-AMPS, IS-95 or cdmaOne, and PDC. GSM, D-AMPS, and PDC are TDMA-based systems, while IS-95 is a CDMA system. Although digital in nature, 2G systems were focused on voice and had very little data-transfer capability. The Wireless Application Protocol (WAP) is a hot topic that has been widely hyped in the mobile industry and outside of it. It is also sometimes, like GPRS, referred to as 2.5G. WAP is simply a protocola standardized way that a mobile phone talks to a server installed in the mobile phone network. Mobile information services, a key application for WAP, have not been as successful as many network operators expected, and WAP is seen as a way to rectify this situation. WAP services are expected to be expensive to use because the tendency is to be on-line for a long circuit-switched data (CSD) call. WAP takes a client-server approach. It incorporates a relatively simple microbrowser into the mobile phone, requiring only limited resources on the mobile phone while the intelligence is in the WAP gateways. The WAP is envisaged as a comprehensive and scaleable protocol designed for use with any mobile phone. Examples are those with a one-line display to a smart phone, any existing or planned wireless service such as the SMS, CSD, unstructured supplementary services data (USSD), and GPRS, and any mobile network standard, such as CDMA, GSM, or the Universal Mobile Telephone System (UMTS). 2.3.2
3G Wireless Networks
2.3.2.1 Definition
It is important to clarify what is meant by 3G. The radio component of advanced technologies varies in terms of at least three characteristics:
16
Transmission Systems Design Handbook for Wireless Networks
1. The radio frequency (RF) channel width. This can range from 200 kHz for GSM-GPRS to 5 MHz for UMTS. 2. The RF spectrum allocation. This can vary from deployment on currently assigned spectrum at 800 and 1900 MHz for CDMA2000 1x to deployment on newly allocated spectrum at 1900 and 2100 MHz for UMTS. While seldom discussed, this latter spectrum is also suitable for CDMA2000 1x as well. 3. The data rate. Depending on technology, this may range from theoretical rates of 115 Kbps to beyond 2 Mbps. Some advanced technologies are called 2.5G, others are called 3G, and some are, or have been, called both. The International Telecommunication Union (ITU) serves as the arbiter of 3G standards. It does not define 3G in terms of channel width or spectrum allocation, but rather of data rates. By ITU definition, the 3G RF interface can deliver data rates of 144 Kbps or greater. The ITU recognizes Wideband CDMA (UMTS) and CDMA2000 1x as meeting this criterion. Higher data rates will enable end users to experience richer content than is now available and, in conjunction with packet architecture, to gain instant and low-cost access to the Internet. Higher data rates, and especially instant and low-cost access to the Internet, will expand future network traffic. As network traffic expands, operator revenues will increase. However, independent of the capability of technologies, operators must recognize the economic-commercial trade-off of network costs versus data rates. The higher the data rates, the greater the network costs. Eventually, every operator must optimize the data rate it offers to end users in terms of the cost to provide it versus the revenues it generates. While the 2G radio access brought mobile telephony capabilities to the mass market, the 3G radio access is expected to introduce value that extends beyond basic telephony. The widespread growth of the Internet has created a mass-market base for multimedia and information services. The challenge is to merge mobile telephony coverage and the associated user base with the Internet and other multimedia applications. To meet this challenge successfully , 3G radio access must provide the following: • Flexible multimedia management; • Internet access; • Flexible bearer services;
Basics of Wireless Networks
17
• Cost-effective packet access for best-effort services.
Most new multimedia services will be offered via the Internet. Therefore, a characteristic feature of 3G radio access is that it provides mobile Internet. Multimedia requires considerable flexibility; that is, the costeffective ability to support different bearer services with very different requirements, such as different bit rates (constant or variable), real-time or best-effort service, and packet- or circuit-switched service. In addition, 3G radio access must provide full-area coverage (same as 2G voice service), high-peak bit-rate services (384-Kbps full-area coverage, 2-Mbps local coverage) and any kind of service mix. Finally, 3G radio access must use the radio spectrum and network resources in a cost-effective fashion. To succeed, the 3G standards must facilitate efficient migration from 2G radio access. The introduction of multimedia into mobile communication will proceed gradually over time. Thus, a step-by-step migration plan must be developed that begins with the state of present-day 2G systems. Given that there are four separate existing 2G standards (GSM, TDMA, PDC, and CDMA/IS-95), different migration paths must be offered. Throughout the past decade, the ITU Radio Communications Sector (ITU-R) has elaborated a framework for global 3G standards. At the same time, since the early 1990s, the industry has been actively researching 3G radio access. 2.3.2.2 3G Standardization Process
It was expected that the 3G mobile phone system would be available commercially in 2001 or 2002, but it appears likely that it will be delayed for a year or even longer. The idea behind 3G is to unify the disparate standards that todays 2G wireless networks use. Instead of different network types being adopted in the Americas, Europe, and Japan, the plan is for a single network standard to be agreed upon and implemented. In 1998 the ITU called for radio transmission technology (RTT) proposals for IMT-2000, originally called Future Public Land Mobile Telecommunications Systems (FPLMTS), the formal name for the 3G standard. Many proposals were submittedthe DECT and TDMA/Universal Wireless Communications organizations submitted plans for the RTT to be TDMA-based, while all other proposals for non-satellite-based solutions were based on wideband CDMA. The main submissions were called wideband CDMA (WCDMA) and CDMA2000. The ETSI/GSM players including infrastructure vendors such as Nokia and Ericsson backed WCDMA. The North American CDMA
18
Transmission Systems Design Handbook for Wireless Networks
community, led by the CDMA Development Group (CDG), including such infrastructure vendors as Qualcomm and Lucent Technologies, backed CDMA2000. After acquiring Qualcomms CDMA infrastructure division in 1999, Ericsson is now supporting CDMA2000 as well. 2.3.2.3 Third Generation Partnership Project
In December 1998 the Third Generation Partnership Project (3GPP) was created following an agreement between six worldwide standards-setting bodies including ETSI, ARIB and TIC of Japan, ANSI of the United States, and the TTA of Korea. This cooperation into standards setting made 3GPP responsible for preparing, approving, and maintaining the technical specifications and reports for a 3G mobile system based on evolved GSM core networks and the frequency division duplex (FDD) and time division duplex (TDD) radio access technology. For example, ETSI SMG2 activities on UMTS have been fully transferred to 3GPP. Interestingly enough, China, an important market for the wireless industry, and the CDG were not original members of the 3GPP. In the first half of 1999, much progress was made in agreeing on a global IMT-2000 standard that met the political and commercial requirements of the various technology protagonists, such as GSM, CDMA, and TDMA. In late March 1999 Ericsson purchased Qualcomms CDMA infrastructure division and Ericsson and Qualcomm licensed each others key intellectual property rights and agreed to the ITUs family-ofnetworks compromise to the various standards proposals. 2.3.2.4 3G Data Rates
The ITU has laid down some indicative minimum requirements for the data speeds that the IMT-2000 standards must support. These requirements are defined according to the degree of mobility involved when the 3G call is being made. As such, the data rate that will be available over 3G will depend upon the environment in which the call is being made. • High Mobility. 144 Kbps for rural outdoor mobile use. This data
rate is available for environments in which the 3G user is traveling more than 120 km/h in outdoor environments. Let us hope that the 3G user is in a train and not driving along and trying to use a 3G terminal at such speeds.
• Full Mobility. 384 Kbps for pedestrian users and those traveling less
than 120 km/h in urban outdoor environments.
Basics of Wireless Networks
19
• Limited Mobility. At least 2 Mbps with low mobility (less than 10
km/h) in stationary indoor and short-range outdoor environments. These kinds of maximum data rates are often talked about when illustrating the potential for 3G technology to only be available in stationary indoor environments.
2.3.3
CDMA2000
CDMA2000 is a wireless standard that supports 3G services as defined by the ITUs IMT-2000 vision. CDMA2000 networks are backward compatible to cdmaOne (IS-95 CDMA), providing simple and cost-effective migration paths to next-generation wireless services and will offer voice quality and voice capacity improvements while delivering high speed and multimedia data services. The CDMA2000 standard was originally divided into two phases, commonly known as 1x and 3x. To realize the IMT-2000 vision in only 1.25 MHz of spectrum, an evolutionary standard under development, known as 1XEV, will enhance the capabilities CDMA2000 can deliver beyond 1x. CDMA operators around the world have recently defined the requirements for this standard through the CDG. CDMA2000 1x is implemented in existing spectrum allocations and delivers approximately twice the voice capacity of cdmaOne and data rates up to 144 Kbps. CDMA2000 1x also offers backward compatibility with cdmaOne networks along with other performance improvements. The TIA has published the CDMA2000 1x standard (IS-2000). The name 1x is derived from the technical term 1xRTT, which refers to CDMA2000 implementation within 1.25 MHz of existing spectrum. The designation 1x means one times 1.25 MHz, and RTT stands for radio transmission technology; 1x can be implemented in existing spectrum or in new spectrum allocations. Building on the 1x standard and CDMA2000s legacy of investment protection and spectral efficiency, 1xEV will improve data throughput, achieving peak rates of 2 to 5 Mbps, without requiring more than 1.25 MHz of bandwidth. Newly defined operator requirements for 1xEV specify two phases. In the first phase, the requirements request data throughput of up to 2.4 Mbps for an efficient, best-effort approach to data delivery. Phase 2 focuses on real-time voice and data capabilities and performance increases for both voice and data efficiency. CDMA2000 3x, part of the original CDMA2000 standard, provides for capacity increases over 1x and data rates up to 2 Mbps using a multicarrier approach. The name 3x is derived from the technical term 3xRTT,
20
Transmission Systems Design Handbook for Wireless Networks
which refers to N = 3 (i.e., use of three 1.25-MHz carriers). Thus, the multicarrier approach utilizes three 1.25-MHz carriers to deliver wideband 3G services. The idea of CDMA2000 3x (3xRTT) was recently abandoned in favor of the new and different approach. The first phase of CDMA2000 next-generation technology, called 1xRTT, has a data-only option, called 1xRTT Evolution-Data Only, for example. Called 1xEV-DO for short, this is the recently standardized CDMA technology based on Qualcomms high data rate (HDR) and adopted by top-tier equipment manufacturers at the end of 2000. According to data from the CDG, 1xEV-DO is expected to provide peak data rates of 1.25 Mbps for downloading data and 300 Kbps for uploads when users are mobile and moving at vehicular speeds. Average rates of 600 Kbps for downloads and 144 Kbps for uploads are achieved in a fully loaded system. The 1xEV-DO technology offers most of what 3xRTT does but accomplishes it within the existing 1.25-MHz CDMA carrier. For operators, that is a big step toward cost-effective high-speed data because it avoids the hardware and software upgrade changes required with the 5-MHz carrier of CDMA2000 3x. These speeds are further increased with a second evolutionary phase of 1xEV, called 1xEV-DV, which also supports voice. It is expected that the 1xEV-DV standard will be ready sometime in 2002. WCDMA and CDMA2000 networks are capable of streaming audio and video to large numbers of users, while GSMs 2.5G solution (GPRS) is not and must focus on text, graphics, animation, and electronic music. Figure 2.1 illustrates the next generation of CDMA2000 network architecture. It is obvious that the network is much more complex than the previous generations of wireless networks as a result of the 3G services offered to the customer. The BSC will separate packet data, which will be handled by the PDSN, from voice and circuit-switched data, which will be handled by the BSC or MSC. The MSC, visting location register (VLR), and home location register (HLR) provide the CS-CU functionality. The CDMA2000 packet core network (PCN) consists of three elements: the packet data serving node (PDSN), the home agent (HA) and the authentication, authorization, and accounting (AAA) server. The PDSN has switching and routing funtionality and functions as a connection point for the radio and IP/ATM networks. Point-to-Point Protocol (PPP) links are established, maintained, and terminated here. In addition, the PDSN delivers foreign agent (FA) functionality to register and facilitate services for network visitors (roamers). In conjunction with the PDSN, the HA authenticates Mobile IP registrations from the mobile client and maintains current location information. The HA also performs packet tunneling, a
Internet
Other ATM/IP networks
Supporting AAAF servers AAAH Router
VLR
Router
ATM
Media gateway/PDSN
AC
MSC server
Operators ATM/IP backbone
PSTN
HA
VLR HLR
Media gateway/PDSN
MSC PSTN
AC BSC
BSC
MSC PSTN
Access network
Access network
cdma2000
cdma2000
Basics of Wireless Networks
HA
HLR
HLR
WAP/HDML
Access transmission network Core transmission network
AAA Authentication , authorization, and accounting BSC Base station controller PCN Packet core network Internet protocol IP AC Authentication center IWF Interworking function HA Home agent PDSN Packet data service node ATM Asynchronous transfer mode WGW Media gateway HLR Home location register WAP Wireless application protocol
21
Figure 2.1 3G high-level CDMA2000-based wireless network topology.
22
Transmission Systems Design Handbook for Wireless Networks
function that receives packets destined for a mobiles permanent address and routes them to the mobiles new temporary address. The AAA server authenticates and authorizes the mobile client, provides user profile and quality of service (QoS) information to the PDSN and stores accounting data. The AAA server provides true policy management, profile definition, and the ability to offer a range of differentiated services. From premium services like 24-hour, high-bandwidth Internet access to more standard services, such as simple e-mail, the AAA server enables custom service packages with unique characteristics that address target market segments. The AAA servers function in a PCN network is similar to HLRs funtion in a voice network. CDMA2000 networks will support two IP addressing options for packet data: • Simple IP: used when no mobilityis required; • Mobile IP: used when mobility beyond a packet zone is required
based on Moblie IP standards from IETF.
To offer 3G services, operators need to invest in the access and core networks of their systems and will be looking for solutions that are both easy to adopt and provide a wide range of services. Initially, emphasis will be on the introduction of high-speed mobile data services, multimedia services, and services that require a guaranteed QoS. End users will expect to have access to services anywhere and at any time; moreover, they will expect reliable, secure connections during transmissions every time. Third generation is based on a different technology platform (i.e., CDMA, as opposed to the TDMA technology widely used in the 2G world, including GSM). 2.3.4
CDMA2000 1xEV-DO
The TIA recently adopted a specification based on HDR, with the designation TIA/EIA/IS-856, also known as 1xEV. The 1xEV specification was developed by 3GPP2, a partnership consisting of five telecommunications standards bodies: CWTS in China, ARIB and TTC in Japan, TTA in Korea, and TIA in North America. CDMA2000 1xEV-DO supports packetswitched voice and packet-switched high-speed data on separate RF channels. The voice channel facilitates the low latency necessary for transmitting two-way conversation, while the data channel enables the flexible routing and low-cost transmission advantages of a packet network. CDMA2000
Basics of Wireless Networks
23
1xEV-DO provides theoretical data speeds of up to 2.4 Mbps. In theory, using separate channels for voice and data requires more bandwidth than using a combined channel. In practice, the spectrum disadvantage diminishes as data traffic increases. This will be especially true for operators with larger spectrum assignments and large data throughput. Of particular value, in some cases not fully recognized, the migration of cdmaOne to CDMA2000 1x and beyond provides a more flexible use of spectrum compared with the migrations from GSM to UMTS and TDMA/IS-136 through GSM to UMTS. Under present concepts, GSM will not be available for the 1,900 and 2,100 MHz frequencies allocated to UMTS. UMTS will not be available for the 800-, 900-, 1,800-, and 1,900-MHz fre- quencies allocated to GSM. However, operators can deploy CDMA2000 1xEV-DO (and eventually EV-DV) either on newly available 1,900- and 2,100-MHz spectrum or on currently assigned 800- or 1,900- MHz spectrum. Some operators have deployed CDMA2000 1x on current spectrum. The Japanese operator KDDI intends to deploy CDMA2000 1x on newly available spectrum while most operators will deploy CDMA2000 1x on current spectrum. This flexible use of spectrum is an advantage of CDMA2000 1x. By enabling operators to use their current spectrum, this technique can save them the overt costs of bidding for new 3G spectrum or, in the case of beauty contests, the covert costs of petitioning for it. The latter can be considerable, especially when they include onerous conditions for network construction. Sweden, for example, did not charge for 3G licenses; however, it did require each license recipient to spend what would have been $3 billion or more for constructing full nationwide networks within two years of the license award. The regulator has since eased this burden by allowing the license recipients to share up to 70% of the 3G infrastructure. Operators who deploy CDMA2000 1x on currently assigned spectrum do not gain the added capacity that new spectrum provides. However, this disadvantage is to some extent overcome by the more efficient coding algorithm that CDMA2000 1x deploys. This algorithm doubles the theoretical capacity of cdmaOne, although in practice the capacity gain without voice degradation will be closer to 50%. UMTS will also deploy a more efficient coding algorithm and realize the associated capacity gains. On the other hand, because GPRS is network architecture, not an RF interface, it cannot provide capacity gains. For more information on 1xEV-DO see [7]. The TIA/EIA/IS-856 standard was prepared by Technical Specification Group C of the 3GPP2. This standard is evolved from and is a companion to the CDMA2000
24
Transmission Systems Design Handbook for Wireless Networks
standards. This air-interface standard provides high-rate packet data services, and 10 different operating bands have been specified. 2.3.5
Future Directions
The central challenge for the industry lies in creating a stable next-generation network platform that can support operators, service providers, and users efficiently and profitably for the foreseeable future. A significant portion of this challenge involves integrating the existing public-switched network with packet-based technologies. Until recently, the business of telecommunications operators centered around providing basic voice services (i.e., dial tone) to a mass market, and voice and data communications connectivity and services to large and medium businesses. Changes in lifestyles, business practices, technologies, and service possibilities mean that a once simple market is increasingly becoming fragmented. It is now possible to consider the concept of the personalized communications environmenta mix of mobile and fixed voice, data, and media services bundled, tariffed, and marketed on a tailored basis to individual users, with the additional freedom that these users can control and change these services at will. Today, less than 1% of traffic in mobile wireless networks is data. The enormous growth, experienced by todays mobile operators and service providers, has been achieved predominantly by voice services. With the arrival of 3G, initially in the form of packet data capabilities for existing networks, and subsequently by full 3G capabilities over new broadband wireless systems, this is set to change. Looking at the mass market in this way, it is possible to visualize discrete market sectors evolving out of a once undifferentiated whole in the following ways: • Basic voice services for the majority of the population will continue
to be the basic communications requirement, and voice is continuing to grow at a rate of approximately 5% a year. Use of advanced services, such as Caller ID, conferencing, call forwarding, and personal numbering will increase, while penetration of mobile voice and communications into the mass consumer market should continue.
• Internet technologies will have an impact on both the cost and
the functionality of voice services. The ability to both provide highquality voice communications over a packet network using IP and deliver advanced voice services will dramatically lower the cost
Basics of Wireless Networks
25
of voice calls, as well as lead to an increase in interactions between the world of the Web and the public voice network. Users will be able to control their services and access billing information through Web-based interfaces or smart terminals, and call-center services will evolve to mix the two environments for e-commerce applications. • Integrated voice and data services will be building on wideband
and broadband copper and wireless technologies (xDSL, HFC, FWA, CDMA2000, WCDMA, GPRS, and UMTS) and on the shift toward IP and packet technologies that can carry voice, data, and multimedia traffic. Simultaneously, users will be able to switch between various communications applications such as Web browsing, e-commerce, entertainment, e-mail, and voice using similar interfaces on different devices for their control.
The role of wireless networks and telecommunications networks in general in facilitating interoperability and ease of use will become increasingly important to consumers. Content and its presentation will become increasingly important in this scenario, as will QoS, reliability, and security. Most mass-market users expect their applications and services to be extremely dependable and intuitive. In theory, software-based functionality can be placed in end user terminals to enable interoperability, to resolve incompatibilities that would be perceived by customers as application failures, and to make complexity transparent to end users. In reality, this is achieved today by forcing end users to be systems administrators of their complex terminal software, or to engage others to administer their systems for them. Traditionally, the telephone networks have hidden complexity from end users and have resolved incompatibilities among end-user terminals by employing middleware in the networks, and all this should be resolved within the 3G wireless networks. In the planning process, these complexities translate into a fairly standard series of problem areas that along with more speculative considerations must be investigated and resolved as efficiently as possible. The first one is the persistent problem of how to extract more value from existing assets like network infrastructure, base station sites, the terms and conditions of the operating license, or limitations of available spectrum. The second dilemma is strategic in its nature and associated with migrating from 2G to 3G technologieswhat kind of capital investment is required, who are the best suppliers to satisfy technical and economical needs, and so on. Third, there is the
26
Transmission Systems Design Handbook for Wireless Networks
question of deciding on the optimum time to deploy a new system, service, or technology, in what market, and in which order.
2.4 Satellite Networks 2.4.1
Fixed Satellite Service
The space age opened many new opportunities for radio communications between widely separated locations. Instead of high-frequency (HF) terrestrial systems with limited bandwidth or a large number of short-range microwave relays, satellites can link distant locations from a point high above the Earth. By the mid-1960s, launch vehicles were delivering communications satellites to locations in the geostationary satellite orbit, about 35,800 km (22,000 mi) above the equator. In this orbit, the satellites circle the Earth at the same rate as the Earth rotates, making them appear nearly stationary from the Earths surface. Communications between two places on Earth can take place by using these satellites; one frequency band is used for the uplink and another for the downlink. Such satellite systems are excellent for the transmission of data, but they leave something to be desired for voice communications. This is a result of the huge distance and the time (delay) it takes for an electrical signal to make an Earth-satellite-Earth round trip, which can amount to more than one-quarter of a second. A reply from the called subscriber takes another quarter of a second or more, and the result of a half a second or more is definitely noticeable. Consequently, voice communications are seldom carried via geosynchronous satellites. Today, geostationary communications satellites continue to play a major role in telecommunications. From the geostationary orbit, satellite antennas can illuminate a small area (using spot beams), a country, or a larger region encompassing many countries. Thus, satellites can theoretically compete with point-to-point microwave and nonradio media (e.g., optical fiber) in providing communications between fixed points. The fixed satellite service primarily involves communications between fixed Earth stations via satellite (i.e., uplinks and downlinks), although the service can also include certain intersatellite links and feeder links. The fixed satellite service can include communications to multiple, specified fixed locations, but does not include broadcasting functions. The fixed satellite service basically involves four frequency bands: 46 GHz, 78 GHz (for military systems), 1114 GHz, and 2030 GHz. Although there are numerous bands above 30 GHz allocated to the fixed satellite service, only one is presently being used. Microwave frequencies and
Basics of Wireless Networks
27
stationary satellites allow the use of high-gain, directional antennas, much like the fixed service. This reduces the power requirements for the satellite transmitters. The fixed satellite service includes international, domestic, and military systems, and although they often carry the same type of traffic, each group has its own set of users. International and domestic systems operate at 46 GHz and 1114 GHz, while military systems use 78 GHz and frequencies near 20 GHz and 45 GHz.
2.4.2
Mobile Satellite Systems
Another of the wireless telecommunications technologies is the low-Earthorbit (LEO) satellite system, made up of satellites that communicate directly with handheld telephones on Earth. Because these satellites orbit at a relatively low altitude (less than 900 miles), they move across the sky quite rapidly [8]. In a LEO system, the communications equipment on a satellite acts in much the same way as a cell site of a cellular systemit catches the call transmitted from Earth and usually passes it to an Earth-based switching system. Because of the speed of the satellite, it is frequently necessary to hand off a particular call to a second satellite just rising over the horizon. This is similar to a cellular system, except that in this case it is the cell site that is moving rather than the subscriber. Due to the high cost of deploying a satellite mobile network, cost of service is very high and today in use only in very special cases and in the remote and unpopulated areas of the world. This is also the main reason for the failure of the Iridium projecta very small number of subscribers. Despite huge debt and failure to meet subscriber acquisition targets, the Globalstar network is unlikely to be abandoned. In the worst case, Globalstar will follow the precedent set by Iridium and Orbcomm (i.e., bankruptcy protection followed by an asset auction). This will allow the highest bidder to obtain a multi-billion-dollar satellite network at a fraction of its original cost. Globalstar offers significant advantages over other satellite networks for both mobile and fixed applications. Due to its high capacity and low latency, Globalstar is ideal for a wide range of voice, Internet access, e-mail, and monitoring and control applications. A broadband mobile satellite network (i.e., a network that employs both satellites on nongeostationary orbits and high-speed switching and routing techniques) could be considered the new generation of satellite networks [9, 10]. Requirements for lower propagation delays and propagation loss, together with the coverage of high latitudinal regions for PCS, have led
28
Transmission Systems Design Handbook for Wireless Networks
to the initiation of extensive research in this area and also in the area of Internet applications over the satellite.
2.5 Fixed Microwave Systems Fixed microwave (MW) systems typically use microwave frequencies above 1 GHz. Point-to-point microwave systems (often simply called microwave systems) provide wideband communications over the line-of-sight (LOS) paths. Tropospheric scatter systems provide point-to-point service over paths up to 200 km, using highly directional antennas and high-power transmitters. Point-to-multipoint microwave systems are used over LOS paths, often with an omnidirectional master antenna and directional node antennas. The fixed services are grouped within seven categories: HF services, VHF/UHF services, and five categories of microwave services. The microwave categories include common carrier, private, auxiliary broadcasting, cable relay, and federal government. These five microwave groups represent the licensing categories that are established and defined by the FCC or NTIA. 2.5.1
Microwave Point-to-Point Systems
The main purpose of microwave radio link is to transport data and voice traffic from one place to another. The radio link uses the air as the transport medium to send encoded electromagnetic waves. A typical link consists of two radios and two antennas separated by a distance from a couple of hundred meters up to tens of kilometers. The data and voice traffic are fed into the radio using either an electrical or optical line. In the radio the digital signals are coded into analog signals and converted to microwaves with a typical length of a few centimeters. Microwaves are used because they are able to propagate high bit rates safely through the air. The microwaves are sent using a highly directive parabolic-shaped antenna. At the other end the signals are received and restored to the digital format. This works both ways, simultaneously, of course. Microwaves, which are only centimeters (or inches) in length, are small relative to the surroundings, and hence, do not have this bending property. In order to establish a radio link, it is important to have LOS between the two radio position sites. Figure 2.2 shows basically two types of microwave systems (i.e., for low- to medium-capacity access and high-capacity backbone transmission systems). Terrestrial MW point-to-point systems use frequencies from approximately 1 to 60 GHz with maximum hop lengths of around 200 km (125
Basics of Wireless Networks
Split-configuration MW radio
15, 18, 23 GHz
Short-distance low- and mediumcapacity systems Access radio
Split-configuration MW radio
29
6, 7, 8 GHz
Long-distance Backbone medium- and systems high-capacity systems
Central office switch
Figure 2.2 Microwave transmission networks.
miles). Long hops over the water or flat surfaces (deserts, wheat fields, planes) are usually more difficult and more expensive to build. A typical microwave hop consists of parabolic dish antenna, waveguide or coaxial cable, and terminal (radio) equipment on both ends of the hop. LOS is the main prerequisite (although not the only one) for achieving satisfactory communication performance over the microwave point-to-point link. The greatest growth area for the use of digital microwave radio is currently associated with the emergence of new competitive wireless operating companies as part of a liberalized telecommunications environment. It is also becoming increasingly common for newly licensed wireless operators to be granted the rights to self-provide the transmission infrastructure. It is also fairly standard that the terms of such competitive licenses commit the operators to challenging operational obligations; that is, to provide service throughout a certain percentage of the country within an ambitious time frame. Furthermore, the terms require wireless operators to provide service at the earliest opportunity to realize revenues in line with their business plans. The speed of installation of the microwave radios and flexibility to upgrade in line with network requirements mean that almost all mobile operators that are independent from the Post, Telephone, and Telegraph Company (PTT)
30
Transmission Systems Design Handbook for Wireless Networks
organizations have the right to self-provide chosen digital microwave radio as the interconnect solution for base stations. In certain parts of the world, utility and government organizations have long had discretionary rights to build their own networks, and have historically been users of microwave radio. With growing liberalization, many other private users are also recognizing the benefits of digital microwave radio. As the above network applications indicate, there are many reasons that a wireless-network operator, given the right to self-provide (and own) transmission infrastructure, should choose microwave radio as opposed to utilizing leased lines or implementing their own cable-based systems. In summary, the advantages of MW radio systems are as follows: • MW systems are economical compared with fiber or leased lines •
•
• •
• •
over a period of at least 2 to 3 years. An owned transmission network remains under the control and ownership of the end user, which removes sometimes sensitive dependency upon the incumbent telephone company (often a competitor) and provides operational benefits. Modern MW radio architecture has been designed to provide a high degree of flexibility in terms of distance and traffic capacity, enabling links to be designed to fit operator requirements and local conditions precisely. Link capacities can also be field upgraded to cater to a networks growing traffic requirements as subscriber numbers increase. Owned MW networks can be planned to provide a higher QoS than often guaranteed by the leased T1/E1 lines. A MW link can, in the majority of circumstances, be installed and commissioned in a much shorter period of time than cable-based alternatives, because a microwave link does not require the same degree of civil works associated with laying cables. MW radio links can be removed and redeployed to another geographical area, without leaving valuable assets in the ground. MW radio is commercially available, and it can usually be supplied with short lead times.
In todays wireless networks, compact base stations require less time and money when it comes to selecting, acquiring, and preparing a site. The savings can be applied to areas that will generate revenue, such as coverage,
Basics of Wireless Networks
31
capacity, and applications, adding value for both operators and end users. Because of the reduced space requirements, the operator has more choices for site selection. This empowers operators to optimize network design and take control of costs early on by negotiating lower leases. This, of course, means that all the other equipment, including power and transmission equipment, should be as compact and upgradable as possible. One of the solutions is to use split-configuration microwave radio, which has an outdoor RF unit mounted very close (or attached to) the parabolic dish antenna. A small indoor unit (digital circuitry including multiplexer) is placed into the shelter of a small cabinet. Outdoor and indoor units are connected with the regular coaxial cable. Microwave radio is the preferred solution for radio base station (RBS) connectivity in most parts of the world, except in North America, where leased T1 lines are quite common and widely available. Most new wireless service providers (WSPs) in North America, in order to achieve short timeto-market, start with leased facilities (lines) and after a year or so, start redesigning the network using microwave radio. 2.5.2
Microwave Point-to-Multipoint Systems
2.5.2.1 Broadband Wireless Access
Currently, multiple cable TV channels, together with ancillary telephony and low data-rate services, are transmitted to homes and businesses via coaxial cable. However, the capability for interactive services and high-speed data transfer over this medium is limited, and cable TV provision necessitates the installation of underground cables right up to the home or business premises. An alternative technology, local multipoint distribution service (LMDS), which overcomes the inadequacies of cable TV and wireline integrated voice and data communications, is now available. LMDS is a fixed, wireless, point-to-multipoint (PMP) telecommunications platform that facilitates the two-way transmission of voice, data, and video. The system operates in the 2431-GHz bands, which, coupled with a high bandwidth capability in excess of 1 GHz, permits the communication of multimedia services with interactive facilities within a 5-km radius of a central data hub. For PMP architecture, the operator installs base stations around the market area very similar to traditional cellular systems. Those base stations have antennas that transmit and receive on multiple sectors and typical configuration is four sectors using 90º antennas. The subscribers (much like WLL users) use much narrower antennas that are typically installed on their rooftops pointing in the direction of the maximum signal strength from the base station. Similar
32
Transmission Systems Design Handbook for Wireless Networks
to cellular systems, frequencies can be reused in neighbor base stations or sectors, as long as the reuse distances are defined in such a way to avoid interference. These systems are used for the bandwidth on demand. Since the bandwidth is shared with other users, bandwidth available per subscriber is reduced with every new subscriber using the system. For subscribers who demand fixed bandwidth availability, point-to-point systems can be offered that will not share bandwidth with other subscribers. Bandwidth on demand is achieved by use of ATM as a transport mechanism. In Canada, the 27.3528.35-GHz (27-GHz) band is designated for LMCS. In the United States, the 27.3527.5-GHz portion of the band is designated for federal government fixed and mobile systems and intersatellite service, and the 27.528.35-GHz portion is designated for LMDS. The 29.129.25-GHz (29-GHz) band is designated for LMDS in the United States and is allocated to fixed and mobile services in Canada. The 31.031.3-GHz (31-GHz) band is designated for LMDS and fixed pointto-point microwave systems in the United States and is allocated to fixed and mobile services in Canada. LMCS is licensed by LMCS service areas and LMDS is licensed by basic trading areas (BTAs). BTAs are defined in [11], which identifies 487 BTAs based on the 50 United States. Further information on U.S. service areas and licensees is available at [1]. Some of the characteristics of PMP systems include the following: • They follow a cellular deployment structure, with multiple cells cov-
ering certain geographic areas.
• Each cell contains a hub with multiple radio nodes equipped with sec-
tor antennas for PMP and directional antennas for point-to-point connections.
• Statistical multiplexing is used. • They can use FDMA, TDMA, spread spectrum (direct sequence
[DS] or frequency hopping [FH]) over-the-air interface.
• They can use FDD or TDD. • LOS between the hub site and all the customer sites is required,
similar to classic point-to-point microwave systems, and therefore standard microwave propagation and prediction methods are used for the system design.
• Frequency coordination with other spectrum owners (and PMP
service providers) can be difficult and must be carefully executed.
Basics of Wireless Networks
33
The services to be provided, aside from regular cable TV, are telephony, video, videoconferencing, video programming (video on demand), data, full-duplex data communications, fast Internet access, and so on. Some other bands are also used for point-to-point systems, like MMDS 2.1 to 2.69 GHz and other unlicensed and newly allocated bands. Air interface standards for new broadband services in the microwave and millimeter-wave frequency range are being developed by the IEEE Working Group 802.16 (Broadband Wireless Access [BWA]). This group, working with the National Institute of Standards (NIST) and National Wireless Electronic Systems Testbed (N-WEST) is the primary focus of industry efforts to develop transmission standards that will support the goal of broadband wireless systems developers and their hardware providers. Originally established to address services in the 1066-GHz range, the 802.16 group has recently added bands below 10 GHz to its portfolio of standards projects. The most economical and practical infrastructure for providing wireless broadband channels to a high concentration of users will probably include a large number of small cells operating at the very high frequencies. It is becoming increasingly apparent that in order to provide wireless communications with bit rates in hundreds of megabits, a large amount of bandwidth in the millimeter-wave (over 30 GHz) must be utilized. A wide, continuous frequency bandwidth is available only in these higher-frequency bands. Many frequency spectrum regulatory agencies around the world, including the FCC, have allocated several large sections of spectrum in the millimeter-wave region [12]. Millimeter-wave characteristics dictate short-range LOS propagation (rain attenuation is predominated at frquencies about 10 GHz) with minimal refraction and reduced interference while providing a bandwidth capacity approaching coaxial or fiber-optic systems. These millimeter-wave characteristics require a cellular network topology to be based on a large number of small cells which facilitate frequency reuse resulting in a large number of traffic channels per service area and, thus, high network traffic capacity. Under FCC Part 15, the 5964-GHz range is available for general use by unlicensed devices based on severe propagation losses that provide protection against interference. In Europe, the 6263-GHz and 6566-GHz bands are allocated for licensed operation and specifically for mobile broadband systems (MBS). In Japan, the 5964-GHz band is regulated for use by MBS as well. The full-scale deployment of fixed wireless technology in the United States was at almost a standstill from 2000 to 2001, because the FCC began studying possible reallocation of the MMDS and Instructional Television
34
Transmission Systems Design Handbook for Wireless Networks
Fixed Service (ITFS) spectrum in the 25.026.9-GHz band. They are among several frequencies identified by the World Radiocommunications Conference for possible 3G use. In the United States the use of millimeter-wave broadband systems is already established, while in Europe, some countries (notably Germany) have already licensed the 26-GHz band. In the United Kingdom, the government awarded licenses for networks operating at 28 GHz in the summer of 2000, with 40-GHz licenses on offer later. Figure 2.3 shows a typical PMP system architecture based on the dynamic bandwidth allocation and ATM. During 2000, so-called mesh networks were proposed, wherein the problems of LOS to every subscriber would be avoided. Some of the subscriber sites within the range of the cell site will not have LOS with the cell site. If the subscriber station can act as a repeater and bounce signals to neighboring subscriber stations, the likelihood of having a LOS path to desired points is significantly improved. In addition to improving LOS coverage, mesh networks can improve service range for those locations that
PSTN 2426 GHz or 28 GHz
RBS 1
AT CE
AT
RBS 2
Hub site 1
90 deg
AT
C-QPSK modulation Point-to-point
RN
SDH/SONET
RS
RN RBS 3
MSC
E3/DS3
ATM MUX or ATM switch
AT
ATMbased, wireless broadband access system: F-DCAFast dynamic capacity allocation Frequency: 2431 GHz Radio Capacity: 37.5 Mbps (symmetrical) ATAccess termination RSRadio shelf CECircuit-emulation shelf RNRadio node NMSNetwork management system
Figure 2.3 MW PMP system architecture.
BSC
NMS
RS Hub site 2
Basics of Wireless Networks
35
originally could not be reached because of the distance and RF path losses. Of course, at one point the delay may become an issue. Industry is still debating the merits of TDD versus FDD in PMP networks. While TDD requires a single channel for full duplex communications, FDD systems require a paired channel for communication, one for the downlink (hub to remote) and one for the uplink (remote to hub). In TDD, transmit-receive separation occurs in the time domain, as opposed to FDD, where it occurs in frequency domain [13]. While FDD can handle traffic that has relatively constant bandwidth requirements in both directions, TDD better handles varying uplink-downlink traffic asymmetry by allocating time spent on up- and downlinks. It is obvious from this discussion that bursty traffic (data, Internet) favors TDD. TDD requires a guard time equal to the round-trip propagation delay between hub and remote units and increases with link distance. In FDD, sufficient isolation in frequency between the uplink and downlink is required. In other words, FDD is a simpler, but less efficient, solution for broadband access. 2.5.2.2 Wireless Local Access in U-NII Band
During the past 15 years, high-frequency digital microwave radio has proven to be a cost-effective and quality solution for short-haul local access applications. The high growth of cellular, PHS, PCN, and PCS networks has predominantly fueled the widespread use of wireless solutions from 10 GHz to 38 GHz for last mile solutions. Radio frequency allocation below 10 GHz continues to be used for wideband backbone traffic. The emphasis and requirement for microwave radio equipment operating below 10 GHz have been on new modulation techniques designed to squeeze more data into less bandwidth. Along with fiber networks, these spectral efficient backbone radios feed both wired and wireless local access applications. Although technological advances continue to provide lower-cost highfrequency products and more efficient backbone solutions, these bands are often congested and still too expensive for future wireless applications. As personal communications products and services evolve, faster and less expensive solutions are in high demand. One of the important developments in the telecommunications industry has been the FCC ruling 15.407 in favor of the U-NII frequency spectrum. On January 1997 the FCC allocated 300 MHz of spectrum for U-NII in the 5-GHz band to be used with U-NII products (see Table 2.1). The FCC believes that the creation of the U-NII band will stimulate the development of new unlicensed digital products that will provide efficient and less expensive solutions for local access applications.
36
Transmission Systems Design Handbook for Wireless Networks Table 2.1 FCC U-NII Bands Band 1
Band 2
Band 3
Frequency (GHz) 5.155.25
5.255.35
5.7255.825
Power (W)
0.2 (EIRP)
1.0 (EIRP)
4.0 (EIRP)
Application
Indoors only Campus
Wireless access
The U-NII band is divided into three subbands at 5.15 to 5.25, 5.25 to 5.35, and 5.725 to 5.825 GHz. The first band is strictly allocated for indoor use and is consistent with the European high-performance local area network (HIPERLAN). The second and third bands are intended for high-speed digital local access products for campus and short-haul microwave applications. The FCC rules for products operating in bands 2 and 3 of the U-NII band are best suited for digital microwave applications over distances in excess of 10 miles. FCC spectral efficiency and maximum power requirements for these bands facilitate the deployment of highly reliable microwave links for both data and telephony transmission. Figure 2.4 depicts the relationship between the maximum equivalent isotropic radiated power (EIRP) and occupied bandwidth of the transmitted signal in accordance with the regulations. 4500 4000
EIRP (mW)
3500 3000
5.7
2000
Hz
25 G
5.8 25
2500 1500 1000
5.255.35 GHz
500 0
5
10
Bandwidth
Figure 2.4 Maximum EIRP in the FCC U-NII bands 2 and 3.
15
20
Basics of Wireless Networks
37
The most effective use of the band is by means of robust modulation schemes capable of carrying high-speed Ethernet or multiple T1/E1 digital circuits. Modulation techniques, such as binary phase-shift keying (BPSK), frequency shift keying (FSK), and quadrature phase-shift keying (QPSK) are best suited to provide the most cost-effective and reliable interconnection solution. The U-NII frequency band is an ideal solution for short-haul applications. Unlike high-frequency microwave links above 10 GHz, the U-NII band is not affected by outages due to rain attenuation. Above 10 GHz, rain is a considerable factor in determining the maximum distance of a properly engineered microwave path. High-power amplifiers and larger antennas help; however, these solutions are expensive and often not applicable. Microwave transmission is also less affected by free space loss at 5.25 to 5.825 GHz than high-frequency microwave. Even with FCC limitations on power output and antenna gain in the U-NII bands (5.3 and 5.7 GHz), microwave paths can operate full duplex, using both bands 2 and 3, over 10 miles with 99.995% reliability. The microwave system performance using both bands 5.3 GHz and 5.7 GHz is limited by the FCC transmitter and antenna rules for band 2. The use of dual-band operation, however, does have the benefit of separating the system transmitters and receivers by approximately 480 MHz. This significantly simplifies the equipment transmitter and receiver design, resulting in a lower-cost product. Dual-band operation also promotes frequency reuse, allowing the use of 200 MHz of bandwidth as opposed to 100 MHz in single-band operation. The use of 2- or 4-ft, highly directional parabolic antennas (with gains of approximately 27 and 33 dBi, respectively) in the U-NII band improves the overall performance of the system. As shown in Figure 2.5, a high level of availability is typical for paths in the 8.5-mile range while meeting FCC rules for maximum EIRP and 99.995% of the time availability requirement for RBS-BSC connection (see later chapters on MW design for more details). Antenna gain can exceed 6 dBi as long as the peak power spectral density is reduced proportionately. Parabolic antennas also offer additional isolation from colocated or adjacent microwave signals. One of the concerns with operation in a license-exempt or ISM band is that of interference from other unlicensed band users. With proper use of the spread-spectrum technology and network planning, interference can be almost eliminated. When deploying networks that are PMP in nature in the ISM bands, there is essentially a choice of two spread-spectrum technologies to be considered: DS and FH. It is generally acknowledged that the DS systems can support higher bit rates than FH systems; however, this comes at a
38
Transmission Systems Design Handbook for Wireless Networks 60
2 FT2 FT antenna
Outage (min/yr)
50
2 FT4 FT antenna
40 30
99.995%
20 10 0
99.990% 3
3.5
4
4.5
5
5.5
6 6.5 7 Distance (mi)
7.5
8
8.5
9
9.5 10
Figure 2.5 Typical point-to-point microwave path performance in the U-NII band.
cost of reduced immunity to interference. For resisting narrowband transient or even nontransient interference in PMP networks, FH systems will have an advantage over DS networks. If the FH system experiences a bad hop, or possibly even more, it will merely send the data out on the next clear hop and no data is lost. With a DS system, if the interference is high enough, the link will fail and data transmission will be interrupted. In the U.S. market, wireless access is most successful where the existing infrastructure is weakest. There is little need in downtown New York or Toronto, for instance, for wireless fractional T1 connectivity because the fiber or copper is already installed and easily obtained. Where wireless networks can make the largest contribution and reach the customers are in the tier-two and tier-three markets. These markets are typically less built out and as a result have far fewer ambient 2.4-GHz radio signals floating around. In these environments, coverage of up to 20 miles is easily achievable. There are many system tradeoffs in selecting an optimum backhaul solution for the hub site. For small towns (of less than 100,000 population) with less data traffic, 28 T1 backhaul capacity may be enough. For larger towns with large numbers of subscribers with high data-rate needs, OC3 and higher capacity may be needed. For lower backhaul capacity needs, multiple T1 connections or point-to-point radio link connections may be sufficient. Point-to-point microwave links are appropriate for backhaul when the required capacity is less than a few T1s to DS3 and the cost of fiber near the
Basics of Wireless Networks
39
base station is too high. This is generally the case in the rural and less populated areas. Also, many times the base station hub is located on a ridge that may be inaccessible by fiber. There are unlicensed band (industrial, scientific, medical [ISM] and U-NII bands) products available for backhaul applications from a number of suppliers. The capacity of these products is generally low (16 Mbps) for multichannel multipoint distribution service (MMDS) cell backhaul applications. This type of low capacity is only useful in cells in a rural area with less population. A typical MMDS cell in a larger city needs OC-3 or higher backhaul capacity. In a combined MMDSunlicensed band system, an MMDS PMP system can satisfy the backhaul needs of the unlicensed band cells. In this scenario the upstream link will be the limiting factor in terms of capacity because the MMDS system modulation rates in the upstream are typically QPSK and 16-QAM (quadrature amplitude modulation). Downstream capacity is generally higher due to availability of 64-QAM modulation. Many applications in a building require portable broadband connections to a notebook or a laptop PC. In-building wireless LANs (WLANs) at 2.4, 5.8, and 24 GHz are now available at multimegabit data rates. Integration of MMDS network with the in-building WLAN will be required for many of these applications. Also, as Bluetooth and HomeRF technology is developed, a low-cost in-building network will provide an easier way for mobile computing and communications devices to communicate with one another and connect to the Internet at high speeds without the need for wires or cables. The broadband MMDS network of the future needs to be able to interface seamlessly in these advanced in-building wireless networks. Various architectures are proposed to use license-exempt bands to supplement and enhance the broadband MMDS networks. Specifically, unlicensed band networks can be used in conjunction with MMDS networks to fill holes in coverage due to hills or buildings or extend the coverage of the MMDS network. MMDS and unlicensed bands can also be used for low- to medium-capacity backhaul requirements. As more devices become available that comply with Bluetooth and HomeRF and other in-building networks, it makes sense to integrate these networks seamlessly with broadband MMDS networks. 2.5.2.3 Bluetooth
Bluetooth is a de facto open standard for short-range digital radio. It is designed to operate in the unlicensed ISM applications band, which is generally available in most parts of the world (Table 2.2). The specification includes air-interface protocols to allow several Bluetooth applications to intercommunicate simultaneously, and to overcome external sources of interference, such
40
Transmission Systems Design Handbook for Wireless Networks Table 2.2 Bluetooth Frequency Bands
Area
Frequency Band (GHz)
Number of Channels
United States, 2.40002.4835 Europe, and most other countries
79
Spain
2.44502.475
23
France
2.44652.4835
23
as microwave ovens. The short range referred to above is defined as up to 10m (30 ft) in normal operation, although greater range and penetration can be achieved through higher output powers under some circumstances. Bluetooth is considered to be a PMP system, although it can also be a point-to-point system depending on the application. Bluetooth technology allows for the replacement of the many proprietary cables that connect one device to another with one universal short-range radio link. For instance, Bluetooth radio technology built into both the cellular telephone and the laptop would replace the cumbersome cable used today to connect a laptop to a cellular telephone. Printers, PDAs, desktops, fax machines, keyboards, joysticks, and virtually any other digital device can be part of the Bluetooth system. But beyond replacing the cables, Bluetooth radio technology provides a universal bridge to existing data networks, a peripheral interface, and a mechanism to form small private ad hoc groupings of connected devices away from fixed network infrastructures. Designed to operate in a noisy radio frequency environment, the Bluetooth radio uses a fast acknowledgment and frequency-hopping scheme to make the link robust. Bluetooth radio modules avoid interference from other signals by hopping to a new frequency after transmitting or receiving a packet. Compared with other systems operating in the same frequency band, its radio typically hops faster and uses shorter packets. This makes the Bluetooth radio more robust than other systems. Short packages and fast hopping also limit the impact (interference) of domestic and professional microwave ovens. Use of forward error correction (FEC) limits the impact of random noise on long-distance links. The encoding is optimized for an uncoordinated environment, since Bluetooth radios operate in the unlicensed ISM band at 2.4 GHz. A frequency hop transceiver is applied to combat interference and fading. A shaped, binary FM modulation is applied to minimize transceiver complexity and a time-
Basics of Wireless Networks
41
division duplex scheme is used for full-duplex transmission with the gross data rate of about 1 Mbps. The Bluetooth baseband protocol is a combination of circuit and packet switching. Slots can be reserved for synchronous packets. Each packet is transmitted in a different hop frequency. A packet nominally covers a single slot, but can be extended to cover up to five slots. Bluetooth can support an asynchronous data channel, up to three simultaneous synchronous voice channels, or a channel that simultaneously supports asynchronous data and synchronous voice. Each voice channel supports a 64-Kbps synchronous (voice) link. The asynchronous channel can support an asymmetric link of maximally 721 Kbps in either direction while permitting 57.6 Kbps in the return direction, or a 432.6-Kbps symmetric link. The different functions in the Bluetooth system include the following: • Radio unit; • Link control unit; • Link management; • Software functions.
The Bluetooth system supports both point-to-point and PMP connections. Several piconets can be established and linked together ad hoc, where each piconet is identified by a different FH sequence and all users participating on the same piconet are synchronized to this hopping sequence. The topology can best be described as a multiple piconet structure. Voice channels use the continuous variable-slope delta (CVSD) modulation voice-coding scheme, and never retransmit voice packets. The CVSD method was chosen for its robustness in handling dropped and damaged voice samples. Rising interference levels are experienced as increased background noise: even at bit error rates up to 4%, the CVSD coded voice is quite audible. The Bluetooth air interface is based on a nominal antenna power of 0 dBm. The air interface complies with the FCC rules for the ISM band at power levels up to 0 dBm. Spectrum spreading has been added to facilitate optional operation at power levels up to 100 mW worldwide. Spectrum spreading is accomplished by frequency hopping in 79 hops displaced by 1 MHz, between 2.402 GHz and 2.480 GHz. Due to local regulations, the bandwidth is reduced in Japan, France, and Spain. This is handled by an internal software switch. The maximum FH rate is 1,600 hops per second. The nominal link range is 10 cm to 10m (4 in to 30 ft), but can be extended to more than 100m (300 ft)by increasing the transmit power. The link type
42
Transmission Systems Design Handbook for Wireless Networks
defines what type of packets can be used on a particular link. The Bluetooth baseband technology supports two link types: • Synchronous connection oriented (SCO) type (used primarily for
voice); • Asynchronous connectionless (ACL) type (used primarily for packet data). Different master-slave pairs of the same piconet can use different link types, and the link type may change arbitrarily during a session. Each link type supports up to 16 different packet types. Four of these are control packets and are common for both SCO and ACL links. Both link types use a TDD scheme for full-duplex transmissions. There are three error-correction schemes defined for Bluetooth baseband controllers: • 1/3-rate FEC code; • 2/3-rate FEC code; • Automatic repeat request (ARQ) scheme for data.
The purpose of the FEC scheme on the data payload is to reduce the number of retransmissions. However, in a reasonably error-free environment, FEC creates unnecessary overhead that reduces the throughput. Therefore, the packet definitions have been kept flexible as to whether or not to use FEC in the payload. The packet header is always protected by a onethird rate FEC; it contains valuable link information and should survive bit errors. An unnumbered ARQ scheme is applied in which data transmitted in one slot is directly acknowledged by the recipient in the next slot. For a data transmission to be acknowledged, both the header error check and the cyclic redundancy check must be okay; otherwise a negative acknowledge is returned. The Bluetooth baseband provides user protection and information privacy mechanisms at the physical layer. Authentication and encryption are implemented in the same way in each Bluetooth device, appropriate for the ad hoc nature of the network. Connections may require a one-way, two-way, or no authentication. Authentication is based on a challenge-response algorithm and represents a key component of any Bluetooth system. It allows the user to develop a domain of trust between a personal Bluetooth device, such as allowing only the owners notebook computer to communicate through
Basics of Wireless Networks
43
the owners cellular telephone. Encryption is used to protect the privacy of the connection. Bluetooth uses a stream cipher well suited for a silicon implementation with secret key lengths of 0, 40, or 64 bits. Key management is left to higher-layer software. The goal of Bluetooths security mechanisms is to provide an appropriate level of protection for Bluetooths short-range nature and enable use in a global environment. Users requiring a higher level of protection are encouraged to use stronger security mechanisms available in network transport protocols and application programs. Public WLANs could be the superior solution for providing nextgeneration wireless services to indoor and campus hot spots. Public WLANs can handle large volumes of data at significantly lower costs, offer a migration path to speeds of 100 Mbps and higher, and deliver additional capacity with pinpoint accuracy compared with leading 3G technologies. That is why 3G wireless-network operators need public wireless LANs to serve the most demanding users in the most demanding locations (hot spots). It also explains why funding, coverage, and roaming are huge challenges confronting independent public WLAN operators (including IEEE 802.11b, Bluetooth, IEEE 802.11a, and HiperLAN/2). Bluetooth can provide access in secondary locations, integrated with pay phones, point-of-sale terminals, and ATMs. IEEE 802.11a offers a migration path to speeds of 54 Mbps and higher. 3G wireless networks, WLANS, and Bluetooth are expected to complement each other and create a total wireless access solution. The following examples demonstrate how 3G wireless systems and Bluetooth could work together, providing local intercommunication as well as wide-area connectivity in a wide range of applications [14]. These are not definitive and by no means exhaustive, but aim to show how complementary standards can work together to provide a greater level of service than either could achieve separately. Vending Machines in Shopping Malls
All the automatic vending machines within a confined area can, through a Bluetooth access system, be connected to a central vending machine administration unit that in turn uses a 3G access system to call for maintenance or supplies. Minor problems can be relayed to the mall technician directly through a Bluetooth communicator. Pricing changes can be sent from central administration and locally broadcast to all Bluetooth vending machines. E-Mail Delivery to the PC
Third-generation terminals will be able to handle several channels simultaneously (e.g., voice, fax, and data each requiring different channel characteristics
44
Transmission Systems Design Handbook for Wireless Networks
and speeds). With predictions of terminal penetration being very high (every member of the population above the age of 12 in a few years), the PC itself does not have to be a 3G terminal in order to receive e-mails on the move. A Bluetooth-3G terminal can receive e-mail as a data transmission and forward it, via Bluetooth, to the PC (assuming it is within close proximity). When the reception is complete, the PC can notify the user via Bluetooth and a short message to his mobile terminal that he has e-mail, and if an item is urgent, this fact can also be forwarded. This concept allows the 3G terminal to be the local head-end for a variety of applications that are locally interconnected via Bluetooth. The Underground Train
Underground facilities suffer from poor coverage on cellular systems. Many underground rail operators are overcoming this by installing systems designed to provide driver and station staff with a reliable communication network. Systems such as Terrestrial Trunked Radio (TETRA) in Europe provide sufficient spare capacity to carry some passenger traffic too. Carriages equipped with Bluetooth transceivers would provide a gateway between the train TETRA system and the users 3G Bluetooth terminal, and the TETRA system would provide the gateway to the surface public networks. The Bluetooth Headset
The 3G Bluetooth terminal mentioned in the above example does not in fact need to be in the users hand or pocket during most of the noted transactions. The user will have a Bluetooth headset, allowing him to leave the terminal in his briefcase too. This may provide voice control and recognition functionality, removing most of the need for a keyboard or display on the 3G terminal. These suggestions may raise the question as to where the 3G terminal in fact should reside. Much of the functionality delivered by 3G systems will be directed toward a data terminal device such as a PC or palm-top computer, and it may be logical to build the 3G terminal into it.
References [1]
Federal Commuications Commission, http://www.fcc.gov/wtb/uls.
[2]
Linnartz, J. P., Narrowband Land-Mobile Radio Networks, Norwood, MA: Artech House, 1993.
Basics of Wireless Networks
45
[3] Winch, R. G., Telecommunication Transmission Systems, New York: McGraw-Hill, 1993, pp. 6276. [4] Wheatly, C., Trading Coverage for Capacity in Cellular Systems: A System Perspective, Microwave Journal, July 1995. [5] Viterbi, A. J., CDMA: Principles of Spread Spectrum Communication, Boston, MA: Addison Wesley, 1995. [6] Zeng, M., et al., Recent Advances in Cellular Wireless Communications, IEEE Communications Magazine, September 1999. [7] TIA/EIA Interim Standard, CDMA2000 High Rate Packet Data Air Interface Specification, TIA/EIA/IS-856. [8] Gardiner, J., and B. West, Personal Communication Systems and Technologies, Norwood, MA: Artech House, 1995. [9] Chotikapong, Y., et al., Evaluation of TCP and Internet Traffic via Low Earth Orbit Satellites, IEEE Personal Communications, June 2001, pp. 2834. [10] Jamalipour, A., The Role of Satellites in Global IT: Trends and Implications, IEEE Personal Communications, June 2001, pp. 511. [11] Commercial Atlas and Marketing Guide 1992, 123rd Edition, Skokie, IL: Rand McNally, 1992, pp. 3839. [12] Gavrilovich, C. D., Broadband Communication on the Highways of Tomorrow, IEEE Communications Magazine, April 2001, pp. 146154. [13] Bolcskei, H., et al., Fixed Broadband Wireless Access: State of the Art, Challenges, and Future Directions, IEEE Communications Magazine, January 2001, pp. 100107. [14] Intercai Mondiale, Ltd., Bluetooth as a 3G Enabler, White paper, 2000.
3 Transmission-Network Principles 3.1 Wireline Side of Wireless Networks 3.1.1
PSTN Interconnect and Telephony Overview
Based on the statistics, most of the voice calls from wireless phones are still directed to regular wireline telephones (POTS), and therefore every wireless network at some point will have to interconnect with the wireline telephone system. Although not very complicated from the technical point of view, interconnect is full of complex regulatory and legal and political issues. This means that every wireless operator should have a small team of people dedicated to and prepared to deal with regulatory and interconnect issues. Interconnect (MSC-PSTN connection) issues are completely separate from and far more complex than the leased-lines issues dealing with the backhaul (radio-base stationbase-station controller [RBS-BSC] connection). In a wireline network, if each locale were limited to three or four telephones, it would make sense to connect each phone to all other phones and find a simple method of selecting the one desired. However, with as many as three or four thousand phones in a locale, such a method is not feasible. It is more appropriate in this instance to connect each phone to some centrally located office and perform switching there. This switching can be a simple manual operation using plugs and sockets or it can be done with electromechanical devices or with electronics. In any case, this central office solution is the one that has been chosen by the telecommunications industry. 47
48
Transmission Systems Design Handbook for Wireless Networks
As these thousands of telephones are connected to the central office (CO), we have what is called a star configuration; all lines are particular to one and only one station, and all terminate at the CO. These connections are called the local exchange plant, and the telephone company handling this function is called the local exchange carrier (LEC). The connections themselves are often called the local loop or otherwise referred to as the last mile. In more technical terms, the section closest to the customers premises is called the distribution plant and that section closest to the CO, the feeder plant. But when a particular telephone call is not originated and terminated within the particular COs geographic coverage, the question is how do we get to another city, or another state, or even another country? The answer, of course, is to connect these COs to a higher-level CO. We clarify the distinction by applying numbers to these levels of offices: The local office (or end office) is called a Class 5 office. The office to which it connects is called the Class 4 office, and so on, with the top level, the Class 1 office, appearing in only a few places in the country. It is worth noting that the only office that has people as its subscribers is the Class 5 office, while other offices in this hierarchy have lower-level COs as their subscribers. Lines connecting switching offices to switching offices, rather than to subscribers, are called trunks. This section of the telephone infrastructure (the section leading upward from the Class 5 offices) is handled not by the LECs but by the interexchange carriers (IXCs), the long-distance carriers. The entire structure is called the hierarchy of switching systems. The whole network is usually called the PSTN. When AT&T was the only long-distance carrier, any time a telephone number was dialed using an area code, the LEC knew that it must be handed off to AT&T. Today there are a number of long-distance carriers and it is no longer obvious what an LEC is supposed to do with a particular longdistance call. To whom should it be handed off? In political terms this is called equal access, which means that a requesting long-distance carrier could require that the LEC examine the number and hand off the call to the proper long-distance carrier. This handoff was from the CO of the LEC to the point of presence (PoP) of the IXC. This PoP could be in a building adjacent to the telephone companys CO, or it could be in some convenient site in the suburbs where it could serve several of the telephone companys COs. Clearly, as time went on the very pure hierarchy of switching systems was becoming somewhat corrupted; new hierarchies in the long-distance part of the network were being applied on top of the old one. Although it is not pertinent to the topology of this network, it should be recognized that the interconnections between these various COs can
Transmission-Network Principles
49
be twisted copper-pair carrier systems utilizing copper pairs (e.g., T1/E1), microwave, satellites, and certainly fiber optics. However, this hierarchical network is not the only network in todays telephone system. Other types of networks include the following: • Local-area networks (LANs), are limited-distance networks connect-
ing a defined set of terminals. They could connect workstations in an office, offices in a building, or buildings on a campus.
• Wide-area networks (WANs), link metropolitan or local networks
usually over common carrier facilities.
• The IN is a concept that centralizes a significant amount of intelli-
gence rather than installing this intelligence in individual COs.
• The Synchronous Optical Network (SONET in North America,
synchronous digital hierarchy [SDH] in the rest of the world) is a particular set of standards that allows the interworking of products from different vendors. It usually embodies a fiber-optic ring that will permit transmission in both directions. The Internet is a packet network (rather than a circuit-switched network), but it is an overlay network in its nature.
• The common-channel-signaling (out-of-band signaling) network is
especially important; it works closely with the PSTN. In the original PSTN, signaling (e.g., call setup) and voice utilized the same common trunk from the originating switching system to the terminating switching system. This process seized the trunks in all of the switching systems involved. Hence, if the terminating end was busy, all of the trunks were set up unnecessarily. In the mid-1970s the common-channel-signaling network, separate from the voice network, was established; it utilizes the protocol called Signaling System 7 (SS7). With this system a talking path was not assigned until all signaling had been satisfactorily completed.
The PSTN described so far has a star configuration with local loops (usually one loop per subscriber) terminating in a CO. This CO completes connections from one local loop to another local loop, or from one local loop to a trunk that terminates on some other CO. This CO has gone through a number of fundamental technological changes. The step-by-step system, which is still in operation in many countries, utilized what is known as the Strowger switch [1]. The intelligence in the
50
Transmission Systems Design Handbook for Wireless Networks
system was located in relays mounted on each switch. In the crossbar system, still electromechanical in nature, the intelligence was separated from the actual switch allowing this common control to be used repeatedly to set up and tear down calls and never sit idle. At the onset of electronics, the electromechanical control of the common control system was replaced with electronics, and the network, or matrix, was usually replaced with tiny glass-encapsulated reed switches. Hence, only a part of the switch was electronic. In the next generation, the stored program operation of a digital computer was applied to the switch, although the network remained a complex of reed switches. In the final generation, called a digital switch, the talking path was no longer an electrically continuous circuit. The speech being carried was digitized. It is important to note that this final generation depicted a significant change from the previous generations in that there was no longer an electrical talking path through the switch. It was, in fact, operating in a digital (rather than analog) domain. However, whether the system was analog or digital, one thing must be recognized: there was an actual talking path, a circuit, from the calling party to the called party. This talking path was established at the beginning of a call and held for the duration of a callit is called circuit switching. This system is not very efficient, because a voice source alternates between talk spurts (active) and silent periods. Silent periods constitute over 50% of the transmission time of voice calls in each direction. There is, however, a different kind of connection, and we see it today in a number of applications called packet switching (PS). In a packet-switching system, the information being transmitted (whether it is data or digitized voice) is not sent in real time over a dedicated circuit; rather, it is stored in a nearby computer until a sufficiently sized packet is on hand. Then a computer seizes a channel heading in the general direction of the destination, and that packet of data is transmitted at very high speeds and after that the channel is released; so, except for some necessary supervisory information (destination, error checking codes, etc.) the channel is 100% efficient. When the distant station gets that message, no more than a few milliseconds later, it responds with the necessary handshaking informationagain, by accumulating a packet of data, seizing a channel, and bursting the information out over that channel. It is again 100% efficient. As mentioned earlier, the packet networks in the world (actually overlay networks to the PSTN) are used extensively for data; only recently are we seeing them used for voice and penetrating wireless networks of the next generation (3G).
Transmission-Network Principles 3.1.2
51
Traffic Engineering
Traffic engineering, as it applies to traditional voice networks, is determining the number of trunks necessary to carry a required number of voice calls during a period of time. For designers of a voice over X (VoX) network, the goal is to properly size the number of trunks and provision the appropriate amount of bandwidth necessary to carry the amount of trunks determined. The performance level, often called the grade of service (GoS), is a function of load and the capacity provided. Performance declines at an exponential rate as the load increases on a given amount of capacity [2]. There are two different types of connections to be aware of: lines and trunks. Lines allow telephone sets to be connected to telephone switches, such as private branch exchanges (PBXs) and CO switches. Trunks connect switches together. An example of a trunk is a tie line interconnecting PBXs (ignore the use of line in the tie-line statement; its actually a trunk). Companies use switches to act as concentrators because the number of telephone sets required are usually greater than the number of simultaneous calls that need to be made. For example, a company may have 600 telephone sets connected to a PBX, but may only have 15 trunks connecting the PBX to the CO switch. In telecommunications, an Erlang is a nondimensional unit with a value between 0 and 1 which indicates how busy a telephone facility is over a period of time (usually one hour). Agner Krarup Erlang (18781929) was a Danish mathematician who invented the formula commonly used to forecast telecommunications traffic. Number 1 applied to a particular telephone circuit would indicate busy 100% of the time. Erlang B is a calculation for any one of these three factors, if you know or can predict the other two: • Busy-hour traffic (BHT), or the number of hours of call traffic dur-
ing the busiest hour of operation; • Blocking, or the percentage of calls that are blocked because not enough lines are available; • Lines, or the number of lines in a trunk group. By monitoring daily and weekly variations in traffic intensity, the BHT may be determined. This is a continuous one-hour period during which traffic in a part of the network is at its most intensive. This may occur at different times of the day, depending on which category of subscribers is dominant. The BHT peak is a mean value, measured over several days. The
52
Transmission Systems Design Handbook for Wireless Networks
aim is to calculate for minimal congestion, while obtaining full utilization of the network. In order to provide a decent level of service, it is necessary to base traffic engineering on a GoS during the peak or busy hour. GoS is a unit of measurement of the chance that a call will be blocked. For example, a GoS of P(0.01) means that one call will be blocked in 100 call attempts, and a GoS of P(0.001) results in one blocked call per 1,000 attempts. It is important to look at call attempts during the days busiest hour. The most accurate method of finding the busiest hour is to take the busiest days in a year, sum the traffic on an hourly basis, find the busiest hour, then derive the average amount of time. In North America it is common to use the 10 busiest days of the year to find the busiest hour. Regardless of which method is used, the intent is to use a number that is sufficiently large in order to provide a GoS for busy conditions and not the average hour traffic. An exchange capacity is not just expressed as the number of possible simultaneous calls. Variations in the holding time of a call (call setup time + conversation time + disconnection time) impact the dimensioning of the exchange, as it may be that the busy hour for switching the calls is not the same as the busy hour for the control equipment. An extended version of Erlang B allows you to determine the number of people who, when blocked, retry their calls immediately. Most of the common models of capacity assume that calls are either served immediately or are blocked and overflow. In some applications, blocked calls can be delayed and served later, which leads us to Erlang C distribution, where capacity estimation is a function of delay criteria instead of blocking. Although the Erlang C traffic model was initially developed for telecommunications applications, it is applicable to any queuing system that meets the following criteria: • Call arrivals are random (Poisson arrivals). • Calls that do not find an idle server enter the queue. • Queue discipline is first come, first served (the queue discipline does
not affect the average call delay, but only the waiting time distribution, meaning the probability of a callers not being served immediately and having to wait).
• The queue length is infinite. • Service channels form a full-availability group with exponentially
distributed service times.
Transmission-Network Principles
53
Other performance parameters developed in conjunction with the Erlang C formula include the probability of a call having to wait longer than a specified period, the average delay on all calls and on delayed calls, and queue size measurements. Erlang C is used for special applications but rarely is used in either wireless or wireline telecommunications network design. The Erlang B and C traffic models make certain assumptions about the nature of the call arrivals. One of them is the assumption that call arrivals are random (Poisson arrivals), and although this is quite reasonable in most applications, it can cause inaccurate results when there is a sudden peak of calls. In many ways, packet-switching theory is comparable to circuitswitching theory, in that the principles of blocking, delay, and alternate routing apply as well as the performance estimation using the same mathematical principles. The major and significant technical difference between packet and circuit switching is the process of assigning the bandwidth. In circuit switching, bandwidth is preassigned per user for the duration of the call, while packet switching allocates total bandwidth per user on a dynamic basis (dynamic bandwidth allocation). The tremendous increase in traffic capacity per circuit through use of packet switching caused a reanalysis of the voice network concepts in combined voice and data networks. New models and simulation methods are currently being developed and tested in order to be able to design and dimension networks capable of carrying voice and data traffic at the same time. For example, in a small wireless network, a simplified calculation for the PSTN traffic (number of required T1 lines between base-station controller [BSC] to PSTN) would look like the following: Assumptions • Small town with a population of 100,000; • 50 milliErlangs (mE) per subscriber; • 5% market penetration first year; • 0.1% GoS; • Private self-contained network80% of traffic is PSTN (no IMTs,
handoffs).
Calculations
First, we determine the number of potential subscribers: • 100,000 [population] × 0.05 [market penetration] = 5,000 subscribers;
54
Transmission Systems Design Handbook for Wireless Networks
Then we determine the total traffic expressed in Erlangs: • 5,000 × 0.05 [Erlangs per sub] × 0.8 [PSTN traffic] = 200 [Erlangs
traffic];
From Erlang B tables, we then determine the number of T1/E1 circuits: • From Erlang B table with 0.1% GoS = 238 voice channels (DS0s); • 238 voice channels/24 DS0s per T1 = 10 T1s (238/30 = 8 E1s).
This is a coarse estimate of the number of T1s or E1s the wireless operator will have to order from the local telephone company (PSTN) to connect to the BSC. We assume here that 80% of wireless subscribers will call outside of the network and their calls will be routed to the PSTN. Typically only a very small percentage of calls are mobile-to-mobile calls within the same network. Most practical voice modeling at the private or public network level for voice assumes a nonstationary Poisson process dependent on time of day and day of week. As mentioned earlier, Erlang B is a very popular formula for trunk sizing. For certain sizing in COs, Erlang-Engset is used, since the number of subscribers is limited. 3.1.3
SS7 and AIN
In traditional telephony, the basic call control procedure is divided into three phases: call setup, the data-conversation phase, and call teardown. Messages on the signaling link are used to establish and terminate the different phases of a call. Standard in-band supervisory tones and recorded announcements are returned to the caller on appropriate connection types to provide information on call progress. Until 1976, the signaling in Bell Systems was almost entirely on a per-line or per-trunk basis and the signaling information for a particular channel was carried on the same channel as the voice or other message information. This type of signaling had many disadvantages, such as long setup times, inefficient trunk usage, restricted signaling periods, and security problems. Common Channel Interoffice Signaling (CCIS) was introduced by AT&T in 1976 to reroute signaling information over a separate path from the voice path. CCIS was an analog system that provided capabilities to support automated calling card and sophisticated routing services, such as 800 service free to the caller.
Transmission-Network Principles
55
The signaling protocol in all modern circuit-switched networks is SS7, which is an internationally standardized, general-purpose CCS system. This protocol enables the exchange of messages through the network for routing, resource reservation, call admission, address translation, call establishment, call management, and billing. Signaling has a special meaning in the telecommunications world: It refers to the information needed to set up, route, monitor, and terminate a call across either a physical or virtual circuit. SS7 network call handling is normally described as being separate from the actual voice connection. This is only partially accurate. SS7 is a separate network, but the call handling is still related to and tied to a specific call reference in the voice path [3]. SS7 is the global standard for telecommunications but it has the Comite Consulatif International de Telegraphique et Telephonique (CCITT) No. 7 and American National Standards Institute (ANSI) No. 7 version. This standard covers procedures and protocols used by network elements in the PSTN to exchange data packets over a digital network. CCITT began specifying the 2G common-channel signaling system in the mid-1970s. More recently, additions have been issued as individual ITU recommendations. The purpose of exchanging data is to enable wireless and wireline call setup, routing, and control, as well as voice communications. The ITU definition is an umbrella standard that allows for variants such as the ANSI and Telcordia (formerly Bellcore) standards used in North America, and the European Telecommunications Standard Institute (ETSI) standard used in Europe. SS7 performs a number of functions: • The fundamental operations of an ordinary phone call: call setup,
management, and call teardown;
• Toll and toll-free wireline services; • Advanced services such as call forwarding, caller ID, and multiparty
calls;
• Wireless services including wireless roaming; • Subscriber authentication; • Local number portability (LNP).
The introduction of stored program-controlled (computerized) switching systems starting in the 1960s made it possible to go beyond direct dialing
56
Transmission Systems Design Handbook for Wireless Networks
of calls to offer customized telecommunications services to meet users needs. The earliest customized services to be offered included customercontrolled call forwarding, three-way calling, and network-based speed dialing. Recently these have been supplemented by services that depend upon the calling partys number, such as caller identification, automatic recalling of the last call received, and call blocking. These services depend upon the combination of low-cost memory (storage of information) and processing power that is enabled by state-of-the-art electronic technologies. However, these types of services have traditionally been implemented by making changes and additions to the large software programs that run the switching systems. Making changes to these large mainframe-like systems is very costly and time consuming. Furthermore, since switches are purchased from different suppliers and come in a multiplicity of types, implementing new services has traditionally required the development and deployment of new generic switching software by multiple suppliers for multiple switch types. This costly and time-consuming process is not consistent with the rapid deployment of a wide range of new telecommunications services that are customized to meet users needs. Thus, the LECs have implemented the advanced intelligent network (AIN) as a client-server approach to creating new services. In this approach, the switches act as clients that interface with software-based functionality in server nodes called service control points (SCPs), service nodes, and intelligent peripherals. The switches, service nodes, and intelligent peripherals implement building-block capabilities that can be mixed and matched by the SCPs to create new services for network users. Because all switches implement comparable building-block capabilities, new services can be created and deployed quickly by implementing new functionality in the server nodes. SS7 separates voice channels from signaling. SS7 data packets go over 56Kbps (mostly in the United States) or 64-Kbps (in most other countries) bidirectional channels called signaling links. Signaling occurs on dedicated channels that are different from the voice/data channels. This provides faster call-setup times and more efficient use of voice circuits. Out-of-band signaling is also needed to support IN services. SS7 has three kinds of switches or signaling points: 1. Service switching point (SSP); 2. SCP; 3. Signal transfer point (STP).
Transmission-Network Principles
57
SSPs originate, terminate, or tandem calls and send signaling messages to other SSPs to set up, manage, and release voice circuits. They may also send a query message to an SCP, which is a centralized database, to figure out how to route a call such as a toll-free call. Network traffic between signaling points may be routed via a packet switch called STP. The hardware and software functions of the SS7 protocol stack are divided into levels connected loosely to the Open Systems Interconnect (OSI) seven-layer model defined by the International Standards Organization (ISO). The message transfer part (MTP) of the SS7 protocol stack is divided into three levels. The lowest level, MTP Level 1, is equivalent to the OSI physical layer. MTP Level 2 ensures accurate point-to-point transmission of a message across a signaling link and is equivalent to the OSI data link layer. MTP Level 3 provides message routing between signaling points in the SS7 network and is equivalent to the OSI network layer. The ISDN user part (ISUP) defines the protocol used to set up, manage, and release trunk circuits that carry voice and data between calling parties. ISUP is used for both ISDN and non-ISDN calls. In some parts of the world such as China and Brazil, the telephone user part (TUP) affects basic call setup and teardown. TUP handles analog circuits. The transaction capabilities applications part (TCAP) supports information exchange between signaling points using the signaling connection control part (SCCP) connectionless service for IN. An SSP uses TCAP to query an SCP to determine the routing number(s) associated with a dialed toll-free number. The SCP uses TCAP to return a response containing the routing number(s) back to the SSP. Calling-card calls are also validated using TCAP. When a mobile subscriber roams into an MSC area, the integrated visitor location register requests service profile information from the subscribers home location register (HLR) using mobile application part (MAP) information carried within TCAP messages. SCCP provides connectionless and connection-oriented network services above MTP Level 3. SCCP allows messages to be addressed to specific applications (called subsystems). SCCP is used as the transmission layer for TCAP-based services, such as toll-free calls, calling card, local number portability, wireless roaming, and PCS. While wireless provides the physical access mechanism for the telephone and other appliances, the AIN provides the software-based functionality to people on the move. Home- and visitor-location registers (AIN service control points) keep track of where nomadic users are and provide the information required to direct incoming calls to those users. AIN can screen or block
58
Transmission Systems Design Handbook for Wireless Networks
incoming calls according to the calling number, time of day, day of week, or other parameters specified by the called party. AIN functionality allows multitier telephones to access cordless base stations, high-power vehicular cellular base stations, low-power pedestrian cellular base stations, and LEO satellite systems, depending on which is most economical and available at any given time. As wireless telephony transitions toward nomadic multimedia computing and communications, the advanced intelligent network will provide access control (security-related) mechanisms, interworking functionality, screening, customized routing, media conversion, and other middleware functionality to support people on the move. 3.1.4
Telecommunications Act of 1996
Prior to the 1996 Telecommunications Act, the service offerings of the Bell operating companies (BOCs) were governed by a modified final judgment (MFJ), or consent decree. The MFJ prohibited the BOCs from entering certain lines of business, including interexchange service. These line-of-business restrictions were based upon the theory that, if the BOCs were allowed to enter the long-distance market, they could use their bottleneck control in the local and exchange access markets to obtain an unfair advantage in the longdistance market. The MFJ also required the BOCs to provide exchange access services, which are necessary to originate or terminate an interexchange service, that are equal in type, quality, and price among the interexchange carriers. Congress overhauled many aspects of federal regulation of communications services with the passage of the Telecommunications Act of 1996. In particular, Congress chose to modify the line-of-business restrictions established under the MFJ. In enacting the Telecommunications Act of 1996, Congress established a pro competitive and deregulatory framework designed to benefit Americans by opening telecommunications markets to competition. As a result, the 1996 Act set the stage for a new competitive regime in which carriers in previously segmented markets would be able to compete in a dynamic and integrated telecommunications market that promises lower prices and more innovative services to customers. Central to the new statutory scheme, and expressly departing from prior jurisprudence developed under the MFJ, are provisions designed to open the local services market to competition. This will lead ultimately to permit all carriers, including those that had previously enjoyed a monopoly or competitive advantage in a particular market, to provide a combined telecommunications offering that includes both local and long-distance services.
Transmission-Network Principles
59
3.2 Digital Transmission Technology 3.2.1
About the Transmission
The terms transmission and transport are used interchangeably in this text, as both are currently in use, the former being preferred in Europe and the latter in North America. In the telecommunications networks today, digital transmission is being used almost exclusively. A high-pitched voice mostly contains high frequencies while low-pitched voice contains low frequencies. A loud voice contains a high-amplitude signal while soft voice contains a lowamplitude signal. Analog signals can be combined (i.e., multiplexed) by combining them with a carrier frequency. When there is more than one channel, this is called FDM. FDM was used extensively in the past but now has generally been replaced with the digital equivalent called time-division multiplexing (TDM). The most popular TDM system is known as Tier 1 (T1). In a T1 system, an analog voice channel is sampled 8,000 times per second, and each sample is encoded into a 7-bit byte. Twenty-four such channels are mixed on these two copper pairs and transmitted at a bit rate of 1.544 Mbps. T1 in North America (E1 in the rest of the world) remains an important method of transmitting voice and data in the PSTN. E1 has 30 channels with a bit rate of 2.048 Mbps. Such a digital transmission scheme (and certainly there are modifications of it that improve efficiency, capacity, and quality) works very well with the digital-switching schemes discussed previously. Thus a talking path (i.e., a switched circuit) in the PSTN can be either analog or digital or a combination thereof. In fact, a digital signal can be transmitted over a packet-switched network as easily as in a circuit-switched network. Digitized voice is little different from data, and, therefore, if data can be transmitted over a packet network, then so can digitized voice. One of the most common applications is now known as voice over the Internet. The challenge, of course, is to get the transmitted signal to the destination fast enough (delay-related issues) because the conversation is time-sensitive. A second challenge is to get each packet, which is a small piece of a voice conversation, to the destination in the proper order. 3.2.2
Transmission MediaPhysical Layer
There are four types of media that can be used in transmitting information in the telecommunications world: 1. Copper lines (twisted pair and coaxial cables): for low- and mediumcapacity transmission over a short distance;
60
Transmission Systems Design Handbook for Wireless Networks
2. Fiber-optic transmission: for medium- and high-capacity transmission over any distance; 3. Wireless: for low- (mobile radio) and medium-capacity (microwave point-to-point) transmission over short and medium distances; 4. Satellite: for low- and medium-capacity transmission over large distances. Years ago, copper wire was the only means of transporting information. Technically known as unshielded twisted pair (UTP), it consists of a large number of pairs of copper wire of varying size in a cable. The cable did not have a shield and therefore the signal (primarily the high-frequency part of the signal) was able to leak out. Also, the twisting on the copper pair was very casual, designed as much to identify which wires belonged to a pair as to handle transmission problems. Even with these limitations, it was quite satisfactory for use in voice communications. Coaxial cable, much larger in dimensions than twisted pair, has been almost exclusively used in broadband systems for video transmission. Coaxial cable consists of a single strand of copper running down the axis of the cable. This strand is separated from the outer shielding by an insulator made of foam or other dielectrics. Covering the cable is a conductive shield. Because of the construction of the cable, obviously coaxial in nature, very high frequencies can be carried without leaking out. In fact, dozens of TV channels, each 6 MHz wide, can be carried on a single cable. Fiber optic is the third transmission medium, and it is unquestionably the transmission medium of choice today. Whereas transmission over copper utilizes frequencies in the megahertz range, transmission over fiber utilizes frequencies a million times higher. This is another way of saying that the predominant difference between electromagnetic waves and light waves is the frequency. Transmission speeds of as high as 10 Gbps have become commonplace in the industry today. At this speed, the entire 15-volume set of the Encyclopedia Britannica could be transmitted in well under one second. Of course, laying fiber, on a per-mile basis, still costs somewhat more than laying copper. However, on a per-circuit basis there is no doubt that fiber is more cost effective. Fiber comes in several forms. The two predominant ones are multimode and single-mode. The total strand diameter for both is about 125 microns (a millionth of a meter). However, the ultrapure glass that forms the core transmission medium is between 50 and 62.5 microns for the multimode fiber and about 8 to 10 microns for the single-mode fiber. It might
Transmission-Network Principles
61
appear that the multimode fiber would have a greater carrying capacity; however, just the opposite is true. With single-mode fiber, only one ray or mode can travel down the strand, making it easier to regenerate the signal at points along the span. In fact, single-mode fiber makes up the majority of todays long-distance network. The tremendous capacity of fiber certainly makes for more efficient communications; however, placing so much traffic on a single strand, for point-to-point communications, makes for greater vulnerability. Most of the disruptions in the long-distance network are a result of physical interruption of a fiber run. Called backhoe fade, the ring configuration is a solution used most often today. Wireless communication is the final option as a transmission medium. This can take several forms: including microwave (point-to-point or PMP), synchronous satellites, LEO satellites, cellular service, and PCS. In every case, however, a wireless system obviates the need for a complex wired infrastructure. In the case of synchronous satellites, transmission can take place across oceans or deserts. With microwave point-to-point systems there is no need to plant cable, and in mountainous territories (as well as downtown areas) this is a significant advantage. 3.2.3
Transmission (Backhaul) in Wireless Networks
The rapid growth in the number and diversity of WSPs cellsites has a corresponding influence on the size and complexity of their transmission (and backhaul) network. In wireless networks, backhaul is defined as the portion of the network that carries the wireless calls from cell site back to the BSC. It is then routed on to the appropriate service termination points, such as MSC and PSTN (voice) or PCN and the Internet (data). Once considered a second-tier priority in relation to RF and cell-site deployment issues, WSPs today view their backhaul and transmission networks as strategic assets since this portion of a WSPs network can run as high as 20% of an annual expense budget. The following lists some of the technological developments that are driving the transmission-network design and deployment field: • Innovations in optoelectronics leading to a near exponential expan-
sion in the amount of data that can be carried by fiber-optic cables, now reaching terabit per second speeds;
• The development of new digital wireless systems such as GSM,
CDMA, and UMTS, enabling the radio spectrum to carry far more
62
Transmission Systems Design Handbook for Wireless Networks
communications traffic at very high quality and deliver mobile Web access and video services; • The widespread adoption of extremely flexible packet-based com-
munications techniques, such as IP, that will eventually replace the large and expensive telecom switches and exchanges currently in use;
• The evolution of technologies and infrastructuresISDN, xDSL,
and CATVthat will give ordinary subscribers access to very high bandwidth services for combined video, voice, and data applications in the home;
• The introduction of intelligence into the network, allowing new
services to be developed for the needs of individual businesses or personal use and the Internet to interact with the public telephone network;
• The increasing familiarity among all types of users with the use of
the Internet and World Wide Web as a ubiquitous communications interface for information searches, carrying out financial transactions and purchases, and accessing entertainment and new media;
• The development of digital media technologies that allow music and
video to be transmitted over relatively low-speed communications links;
• The possibilities of unified messaging applications that can translate
voice, e-mail, and fax messages into the most appropriate medium for receipt by a mobile user.
3.2.4
DSX-1 Digital Interfaces (North America)
DSX-1 is a physical interface of the T1 circuit. Figure 3.1 illustrates some of the most commonly used pieces of equipment in the transmission networks. The transmultiplexer is a device for the conversion between the analog and digital worlds in transmission networks, and while rarely used today was quite common in the early 1980s during the beginning of digitalization in telecommunications. There are two standards for first-order digital transmission systems. The T1 system, developed by Bell Laboratories, is used mainly in the United States, Canada, Taiwan, Jamaica, and few other countries in the world [4]. Most of the countries around the world use the E1 system defined by European Conference of Postal and Telecommunications (CEPT) Administration.
Transmission-Network Principles
63
Digital switch DS0s 1
Digital MUX*
DS1
Digital MW radio or fiber-optics
DS1 DSX-1 DS1 Cross-connect Two 12-channel Channel 655' (digital MDF) DS1 Trans groups bank MAX (60108 kHz) Analog MUX* (FDM) MUX* (TDM-FDM) 24 PCM channels
24
Voice and 1.544 Mbps data circuits
DS1
T1 office repeater
Span line
TELCO Transmultiplexer converts analog voice channels into digital bit stream without demultiplexing
*Multiplexer
Figure 3.1 DSX-1 digital interfaces (North America).
3.2.5
North American Digital Hierarchy
Details of the T1 systems will be discussed in other chapters of this book. Here we give only a brief description and definition of what comprises T1. • There are 24 DS0 (64 Kbps) timeslots in a T1 line, providing a total
bandwidth of 24 × 64 Kbps = 1,536,000 bps (1,536 Kbps).
• Another 8,000 bps are used as the equivalent of lane markers to
allow traffic to remain in its assigned lanethese overhead bits are known as framing bits.
• Adding up the total bandwidth of the 24 DS0 channels and the 8
Kbps framing bits yields the 1,544-Kbps T1 data rate.
The North American multiplexing scheme is shown in Figure 3.2; the North American T1 digital hierarchy is shown in Figure 3.3. For the hierarchies based on the 1.544-Mbps primary rate, the principle has been established that some bits in the frame should be reserved, in particular to perform quality control of the digital paths when several digital
64
M12Rarely used today M13Most common type of multiplexer today 1 2. ..
PCM channel bank
1.544 Mbps (24 Channels)
24 DS1
DS3 1 2
M12
6.312 Mbps (96 Channels)
3 4 DS2
28 1.544-Mbps T1 trunks
1 2. .. 28
M13
1 2 3 4 5 6 7
44.736 Mbps (672 Channels)
44.736 Mbps (672 Channels)
M23 DS3
1 2 3
The North American hierarchy is standard in the United States, Canada, Taiwan, Korea (except in some new cellular links), and Japan (through DS2). Other countries use the European CEPT 30-channel hierarchy.
Figure 3.2 North American pulse-code modulation hierarchy.
3 DS3s (84 DS1s)
135 Mbps (2016 Channels)
Transmission Systems Design Handbook for Wireless Networks
Voice and data circuits
Transmission-Network Principles
Designation DS0 DS1 DS2 DS3
# DS1 signals 24/DS1 01 04 28
Bit-rate (Kbps) 64 1,544 6,312 44,736
Line code AMI AMI/B8ZS B6ZS B3ZS
65
Voice channel equivalent 1 96 24 672
Line lengths (Ft/m) 655/200 1,000/300 450/140
The AMI, B8ZS, B6ZS, and B3ZS codes are bipolar. Line codingused to ensure that enough Cable types: 100/110 Ω twisted-pair, 75 Ω coax timing information accompanies the digital Ref: ITUT G.703, G.704; Bellcore TRTSY-000499 signal to allow accurate recovery of all the bits
Figure 3.3 North American T1 digital hierarchy.
sections in tandem are involved [5]. A 64-Kbps circuit (DS0) that uses 8 Kbps for signaling is called Switched 56, Digital Data Service (DDS) or Advanced Digital Network (AND). Each carrier used to have its own name for this service. Today, this 56-Kbps service is becoming obsolete. Some important facts to remember about T1 are as follows: • Most T1 circuits today use extended super-frame (ESF) format and
B8ZS line code.
• ESF and B8ZS provide clear channel with only 2 Kbps used for
framing + 2 Kbps facility data link (FDL) + 2 Kbps (CRC-6) = 6 Kbps.
• Older T1s use 8 Kbps for framing and bit robbing for provisioning
and maintenance.
3.2.6
CEPT Digital Hierarchy
E1 is described in more detail in the following chapters. Only a short description and E1 definition are given here; more information is available in [6] and in other chapters of this book: • There are 30 DS0 (64 Kbps) timeslots in an E1 line, providing a
total bandwidth of 30 × 64 Kbps = 1,920,000 bps (1.92 Kbps).
• Sixty-four thousand bps are used as the equivalent of lane markers to
allow traffic to remain in its assigned lane.
66
Transmission Systems Design Handbook for Wireless Networks
• One 64-Kbps timeslot (TS0) is used as framing bits. • Another 64-Kbps timeslot (TS16) is used for signaling of voice fre-
quency channels.
• Adding up the total bandwidth of the 30 DS0 channels, framing,
and signaling bits yields the 2,048-Kbps E1 data rate.
The CEPT multiplexing scheme is shown in Figure 3.4. In the case of networks using a 2.048-Mbps-based hierarchy (Figure 3.5), there is in principle no basic restriction on the use of full capacity of the digital path. However, it is recognized that compatibility with recommended frame structures at the various levels of the 2-Mbps hierarchy (e.g., the use of the same frame alignment pattern) could be a preferred solution since it offers the following advantages: • Use of the same framing devices for switched and nonswitched
applications;
• End-to-end quality control performed in a unique way by the net-
work when the maintenance entity that terminates the service (e.g., the encoding device) does not belong to the network;
• The possibility of performing additional network management
functions that could be required, depending on the applications.
3.3 Plesiochronous Versus Synchronous Digital Hierarchy Traditionally, transmission systems have been asynchronous, with each terminal in the network running on its own clock. In digital systems, clocking (timing) is one of the most important considerations. Timing means using a series of repetitive pulses to keep the bit rate of the data stream constant and to indicate where the ones and zeros are located in a data stream. Since these clocks are free running and not synchronized, large variations occur in the clock rate and thus in the signal bit rate. Asynchronous multiplexing uses multiple stages; lower-rate signals are multiplexed and extra bits are added (bit stuffing) to account for the variations of each individual stream and combined with other bits (framing bits) to form a higher-level bit rate. Then bit stuffing is used again to produce even higher bit rates. At the higher asynchronous rate, these signals cannot be accessed without a complete demultiplexing procedure.
PCM 1 channel 2. .. bank 30 1st order
2.048 Mbps (30 channels) 1 M2-8 2 2nd order 3 4
E3 8.448 Mbps (120 channels) 1 2 M8-34 3 3rd order 4
E4
E2 E3 2.048 Mbps 2.048 Mbps 2.048 Mbps
1 2. M2-34 .. 16
34 Mbps (480 channels)
1 2 3 4
M34-140 MUX*
Transmission-Network Principles
E1
34.368 Mbps (480 channels)
Skip (double-step) MUX* The CEPT hierarchy is the international standard everywhere except North America (USA, Canada), Taiwan, Korea, and Japan.
67
Figure 3.4 CEPT pulse-code modulation hierarchy.
*Multiplexer
68
Transmission Systems Design Handbook for Wireless Networks
Designation E0 E1 E2 E3 E4
# E1 signals
Bit rate (Kbps)
30/E1 1 4 16 64/63 *
64 2,048 8,448 34,368 139,264
Line code
Voice channel equivalent
AMI HDB3 HDB3 HDB3 CMI
1 30 120 480 1920/1890 *
The AMI, AMI, HDB3, HDB3, & CMI andcodes CMI codes are bipolar. are bipolar. Ω Twisted Cable types: 120 twisted-pair, 75 ΩCoax coax 120W Pair, 75W (Length/type assigned for 6 dB maximum 6-dB maximum loss) loss) Ref: Ref: ITU-T ITU-T G.703, G.703, G.704 G.704
E1 400mtwisted-pair 750mcoax E3 375mcoax
* 63 E1 (1,890 channels) are mapped in SDH systems
Figure 3.5 CEPT E1 digital hierarchy.
In a plesiochronous digital hierarchy (PDH) system, the average frequency of all clocks in the system will be the same (synchronous) or nearly the same (plesiochronous) and every clock can be traced back to a highly stable reference supply. The prefix plesio, which is of Greek origin, means almost equal, but not exactly, meaning that the higher levels in the CCITT (ITU today) hierarchy are not an exact multiple of the lower level. The North American digital hierarchy starts off with a basic digital signal level of 64 Kbps (DS0) (see Table 3.1). Thereafter, all facility types are
Table 3.1 North American Data Rates Name
Rate (Kbps)
DS0
64
DS1
1,544
DS1C
3,152
DS2
6,312
DS3
44,736
DS4
274,176
Transmission-Network Principles
69
usually referred to as Tx, where x is the digital signal level within the hierarchy (e.g., T1 refers to the DS1 rate of 1.544 Mbps). Up to the DS3 rate, these signals are usually delivered from the provider on twisted-pair or coaxial cables. North American T1 service providers often refer to the signal interfaces between the user and the network as DS1 signals. In the case of user-to-user interfaces, the term DSX-1 is used to describe those DS1 signals at the crossconnect point. The CCITT digital hierarchys basic level is the DS0 rate of 64 Kbps (see Table 3.2). These signals are usually delivered from the provider on twisted-pair or coaxial cables. All E-carriers and T-carriers are based on PDH. In the late 1980s, synchronous network hierarchies were introduced. The term synchronous means occurring at regular intervals, and is usually used to describe communications in which data can be transmitted in a steady stream rather than intermittently. For example, if a telephone conversation was synchronous, each party would be required to wait a specified interval before speaking. The SONET/SDH is a newer technology in the field of digital transmission. The transmission is carried out in a synchronous mode, hence the name synchronous digital hierarchy. The most important advantage in adopting this synchronous technology is to enable the mapping of various user bit rates directly on to the main transmission signals, thus bypassing various stages of multiplexing and demultiplexing as was done in the case of earlier PDH technology. The number of new architectures and topologies made possible as a result of this new technology has rendered it the only viable alternative for adoption today. For instance, add-drop multiplexers (ADMs) have made it possible to use SONET/SDH terminals in a long chain with bit streams Table 3.2 ITU Data Rates Name DS0 E1
Rate (Kbps) 64 2,048
E2
8,448
E3
34,368
E4
139,264
70
Transmission Systems Design Handbook for Wireless Networks
added or dropped along the way in an effective manner. By closing the chain at two ends, the ring configuration is possible; this provides enhanced protection features. Other advantages include support of the network management system, easy upgrade to high bit rates, and adaptability to the existing PDH. In synchronous networks, all multiplex functions operate using clocks derived from a common source. The North American SONET system is based upon multiples of a fundamental rate of 51.840 Mbps, called Synchronous Transmission Signal, Level 1 (STS-1). The facility designators are similar, but indicate the facility type, which is usually fiber-optic cable (e.g., OC-1 is an optical carrier supporting an STS-1 signal, while OC-3 supports an STS-3 signal, and so on). Some typical rates are listed in Table 3.3. The international SDH system is based upon a fundamental rate of 155.520 Mbps, three times that of the SONET system. This fundamental signal is called Synchronous Transmission Module, Level 1 (STM-1). The typical transmission medium is defined to be fiber, but the broadband ISDN specification does define a user-network interface (UNI) STM-1 (155.520 Mbps) operating over coaxial cables. Some typical rates within this hierarchy are shown in Table 3.4. The evolution from PDH to SDH worldwide is a phased process. One of the more common problems faced at this intermediate stage is that SDHbased networks must have the flexibility to utilize existing PDH transport media.
3.4 Multiplexing and Inverse Multiplexing Multiplexing is a process in which multiple data channels are combined into a single data or physical channel at the source. Multiplexing can be Table 3.3 STS Data Rates Name
Rate (Kbps)
STS-1
51,840
STS-3
155,520
STS-9
466,560
STS-12
622,080
STS-48
2,488,320
Transmission-Network Principles
71
Table 3.4 STM Data Rates Name
Rate (Kbps)
STM-1
155,520
STM-3
466,560
STM-4
622,080
STM-16 2,488,320
implemented at any of the OSI layers. Conversely, demultiplexing is the process of separating multiplexed data channels at the destination. One example of multiplexing is when data from multiple applications is multiplexed into a single lower-layer data packet. Another example of multiplexing is when data from multiple devices is combined into a single physical channel (using a device called a multiplexer). Some methods used for multiplexing data are TDM, asynchronous timedivision multiplexing (ATDM), FDM, and statistical multiplexing. FDM involves splitting the frequency band transmitted by the channel into narrower bands. Each of these narrow bands is used to constitute a distinct channel. In FDM, information from each data channel is allocated bandwidth based on the signal frequency of the traffic. Multiple channels are combined onto a single aggregate signal for transmission. The channels are separated in the aggregate by their frequency. There are always some unused frequency spaces between channels, known as guard bands. Guard bands reduce the effects of bleedover between adjacent channels, a condition more commonly referred to as cross-talk. The other method is to allot a common channel to several different information channels one at a time (TDM). FDM was the first multiplexing scheme to be widely used, and such systems are still in use today. However, TDM is the preferred approach today. In TDM, information from each data channel is allocated bandwidth based on preassigned timeslots, regardless of whether there is data to transmit. In ATDM, information from data channels is allocated bandwidth as needed, by using dynamically assigned timeslots. In statistical multiplexing, bandwidth is dynamically allocated to any data channels that have information to transmit. Time-assignment speech-interpolation (TASI) systems represent an example of an analog statistical TDM scheme. These systems enjoyed limited use in the 1980s, and were particularly adept at sharing voice circuits; specifically PBX trunks. A TASI multiplexer is interconnected between the PBX and the trunk facilities. Usually, one analog trunk circuit is used for signaling
72
Transmission Systems Design Handbook for Wireless Networks
purposes between TASI units at each end of the link. The remaining voice trunks support analog TASI TDM voice conversations. In normal telephone conversations, the majority of time is spent in a latent (idle) state. TASI trunks will allocate chunks of voice from another channel during this idle time. If individuals were to monitor these TASI trunks, they would hear bits and pieces of various conversations. The signaling channel is used for the signaling conversion between end-point PBX units and also for the allocation of bandwidth once incoming speech energy has been detected. As digital speech processing became more common, TASI systems were created that had analog inputs and digital outputs. This type of multiplexing technique is more commonly known as digital speech interpolation (DSI). There are a few drawbacks in using TASI and DSI systems. First, users can notice a good deal of voice clipping which occurs when a little bit of speech is lost while waiting for the TASI multiplexer to detect valid speech and allocate bandwidth (more about multiplexing can be found in [7]). Clipping also occurs when there is no bandwidth present at the moment. TASI and DSI units are also very susceptible to audio input levels and may have problems with the transport of voiceband data (e.g., VF modem) signals.
3.4.1
Statistical Multiplexing
Any data communications system that has more than one asynchronous line going between common locations can benefit by installing a pair of statistical multiplexers (STATMUX). A STATMUX performs the function of combining several asynchronous data communication channels into one composite synchronous signal that is transmitted between two locations more inexpensively than the cost of individual lines. Individual users connect to asynchronous channels, and the composite communication line between the two locations is called the link. The STATMUX utilizes a different form of TDM. These multiplexers typically use a high-level data link control (HDLC)like frame for aggregate communications between units. As input-output (I/O) traffic arrives at the multiplexer (MUX), it is buffered, then inserted into the I-field of the HDLC frame. The receiving units remove the I/O traffic from the aggregate HDLC frame. Statistical multiplexers are ideally suited for the transport of asynchronous I/O data, as they can take advantage of the inherent latency in asynchronous communications. They are typically faster at transporting I/O data end to end than X.25 systems, but some of these multiplexers can also perform
Transmission-Network Principles
73
network switching functions between I/O ports. The total I/O bandwidth can (and usually does) exceed the aggregate port bandwidth (multiplexing gain). Later, many of these multiplexers incorporated switching mechanisms that allowed I/O ports to connect themselves intelligently to other destination ports upon user command. While functioning somewhat as an X.25 switch, these statistical multiplexers were usually faster, and provided more transparent I/O data-carrying capacity. Frame relay and X.25 systems are also categorized as statistical TDMs. Both systems utilize aggregate HDLC frame structures. The advantage of frame relay over X.25 is that it can support the same traffic as X.25, while facilitating bandwidth-on-demand requirements for bursty traffic. Frame relay, however, cannot adequately support voice or video traffic because of variable end-to-end delivery times (e.g., variable delay). Voice and video transmissions are of a constant-bit-rate (CBR) nature, and are not happy sitting in a queue waiting for a big LAN packet to finish transmitting.
3.4.2
3/1 Multiplexing and Subrate Multiplexing
An M13 multiplexer is usually a modular, compact unit for multiplexing up to 28 DS1s into a DS3. Compatible with North American standard interfaces, these multiplexers are capable of multiplexing up to seven low-speed DS1 signal groups (each with 4 DS1s) into one DS3 (44.736 Mbps) signal, with each low-speed signal group consisting of four DS1s. Standard features include continuous performance monitoring and extensive local and remote diagnostics, which provide off-premises restoration, thereby avoiding the cost of on-premises maintenance. Remote monitoring and control are the key features, providing remote provisioning, remote inventory, performance monitoring, and remote testing. The FMT150 is a fiber-optic multiplexer that multiplexes and transports one, two, or three DS3 signals, up to 84 DS1s, and a maximum of 2,016 voice circuits. Designed to interface with DS1 or DS3 input signals, the typical FMT150 uses 150-Mbps fiber transport plug-ins (with up to three DS3 inputs). FMT150 usually features a comprehensive maintenance system and provides complete, instantaneous monitoring and troubleshooting capabilities, including fault analysis of the remote site. It can also monitor alarms and controls at local sites. The purpose of subrate multiplexing is to fit more sub-DS0 data circuits into one DS0 (64 kb) channel. The SRDM feeds synchronous or asynchronous data circuits into a single 64-Kbps DS0 signal. For example, the
74
Transmission Systems Design Handbook for Wireless Networks
SRDM can multiplex 20 sub-DS0 signals into a DS0 channel at 2.4 Kbps, or 10 at 4.8 Kbps, or 5 at 9.6 Kbps. 3.4.3
Inverse Multiplexing
3.4.3.1 The Need for Bandwidth
The demand for large amounts of bandwidth over extended distances is driving the interest in networking technologies such as ISDN, frame relay, switched multimegabit digital service (SMDS), asynchronous transfer mode (ATM), satellite data communications systems, wireless communications systems, and others. However, making most of these services universally available requires either a new communications network infrastructure, or significant modifications to the existing one. For example, ATM offers high-bandwidth digital connections based upon fixed-size cells that can carry voice, video, and data. But universal ATM also requires the public switched telephone network to replace its time-division multiplexed switching fabric with a new ATM switching fabric and enhanced interoffice trunk facilities. Considering the value of the existing worldwide telephony infrastructure (switches, transmission systems, and embedded wiring plants), this is unlikely to happen any time soon. Therefore, while alternative transmission technologies will certainly be implemented over time to handle the growing demand for high-speed digital bandwidth, full use of the existing digital TDM infrastructure is currently required. While originally conceived as a transmission network for 64-Kbps digitized voice, it is now possible to dial up point-to-point digital connections whose bandwidth ranges from 56 Kbps to 3 Mbps and beyond. Two significant enhancements to TDM networking have made this possible. The first is newly developed software for digital TDM switches that allows dialed connections to exceed the original design channel rate of 56 or 64 Kbps, thus allowing carriers to offer dialed wideband services. The second is the use of specialized equipment that resides at the users premises to allow multiple independent digital connections to be combined to create a single, higherspeed end-to-end connection. This technique is known as inverse multiplexing, and the equipment that performs it is called an inverse multiplexer. NxT1 and NxE1 inverse multiplexing has been around since the early 1990s. Initially, bit-based inverse multiplexers (IMUXes) were developed to aggregate bandwidth from up to eight T1/E1s to provide multimegabit connections for high-speed frame relay, Internet access, videoconferencing, or private network trunks. The delay of each link may be different due to physical path length or buffering in transmission equipment. At the receiving end,
Transmission-Network Principles
75
the IMUXes reconstruct the original bit stream, using buffers to compensate for variations in delay of the individual links. Bit-based IMUXes have several properties that make them easy to use: The order of the bit stream is always preserved, the bundled link can be managed as a single entity, and the data is transparently transported regardless of protocol. However, special proprietary hardware is required at both ends of the link to implement the inverse multiplexing. 3.4.3.2 Basic Features of IMUXes
Up until now, modems have been the basis of the public network data transmission. Their advantage is that they plug directly into the ubiquitous analog lines that access the public telephony network. Their disadvantages include their susceptibility to errors and, more importantly, their limited data rates (typically up to 19.2 Kbps and lately 56 Kbps). While modems are inherently analog transmission devices, the majority of todays public network is digital. Network switches and interswitch trunks have migrated from analog to digital for ease in maintenance, supplemental service offerings, and lower transmission costs. If digital lines are used from the customer premises to the digital network infrastructure, an end-to-end digital call can be made, using the PSDN. The bandwidth of this call far exceeds the bandwidth of a modem call. Since the network is a fabric of 64-Kbps channels, a digital calls bandwidth will be a multiple of 64-Kbps. As mentioned previously, this aggregation of 64-Kbps channels can be performed by either the network itself or by inverse multiplexers located on the customers premises. Inverse multiplexing, which works at the physical layer, spreads a data stream received from any DTE device across the link and reconstructs it at the other end of the link. In other words, inverse multiplexing can be defined as the aggregation of multiple independent information channels across a network to create a single higher-rate information channel. For example, if six different independent 56-Kbps data channels are established between points A and B, inverse multplexing can be used to combine these channels to create a single 336-Kbps (i.e., 6 × 56 = 336) data stream. Likewise, 64-, 384-, and 1,536-Kbps channels can be inverse-multiplexed together. Specifically, the IMUX assures that all the channels are present and accounted for. The IMUX then segments the transmission data stream and sends it out over the individual channels (it could be leased or owned T1/E1 lines). At the receiving end the IMUX accepts the data from these channels, reordering the segments and compensating for variances in channel transit times. Depending upon the inverse multiplexing protocol in use, the IMUX may monitor the integrity of the aggregated connection. Should transmission
76
Transmission Systems Design Handbook for Wireless Networks
problems occur, the IMUX can take diagnostic action, replacing failed or failing channels with functional channels to maintain the integrity of the connection. The IMUX can also add or remove channels from the aggregated connection without terminating the connection. This allows the total amount of bandwidth between the two sites to vary according to real-time bandwidth requirements or for economies of operation. This feature is sometimes referred to as dynamic bandwidth allocation, or rubber bandwidth. The technology can be used in both dialed and leased-line environments and provides an efficient way to bridge the gap between T1/E1 and T3/E3 and alleviate the bottleneck in the broadband networks. Data-channel reordering and network-delay compensation are important features of inverse multiplexers. Because each of the channels of an inverse multiplexed call is separate and independent, there must be a mechanism within the inverse multiplexer to accept data from each of the channels and then reassemble it in the correct order. In addition, each of the channels of an inverse multiplexed call can and will likely follow a path across the network that has a different propagation time than other channels within the connection (see Figure 3.6). A call from San Francisco to New York may have one channel routed through Dallas, another through Chicago, and yet another through both Atlanta and Washington, D.C. Some channels of an inverse-multiplexed connection might go through a satellite channel, which has nearly a 300 m of propagation delay, while the remaining channels follow much shorter terrestrial paths. The digital switching matrices within each network switch add additional delay to the channel as well. Therefore, the inverse multiplexer must also compensate for the differential delays between channels. The inverse multiplexer compensates for each channels different delay Variable-delay T1/E1 circuits
Data distributed bit-wise to lines
ww
Figure 3.6 IMUX delay issues.
w w w ww w w
ww w
IMUX Outgoing w w Data reassembled in proper order y x
y x
x x x x x x x y y y y y y y
y y
y x
y x
w
x x
Incoming w
y x x x x
IMUX
Transmission-Network Principles
77
characteristics by buffering each channels data as it is received. Each channels data buffer is in essence a first-in-first-out (FIFO) buffer, and these FIFOs are unloaded in such a way that the reassembled data stream is properly reconstructed. The total end-to-end delay of an inverse-multiplexed connection is approximately that of the channel with the longest network transit time (internal inverse-multiplexer processing delays are small compared with network transit times). How is this feature used? In videoconferencing, the quality of video motion increases as the bandwidth between sites increases. A videoconferencing session might use a 384-Kbps inverse-multiplexed connection at the beginning of a course lecture, when the lecturer is on camera and moving about the stage. However, the bandwidth can be reduced when the camera focuses for an extended period on a blackboard, when the relative lack of motion requires less bandwidth and, therefore, less expense. In LAN internetworking, an inverse-multiplexed connection between two remote LANs might be established at 64 or 128 Kbps, adequate for most simple interactive activities. However, if that connection becomes saturated for a predetermined period of time by, for example, a large file transfer, the bandwidth of the connection can be automatically or manually increased to accommodate the increase in traffic, allowing file transfer to take place much faster. At the end of the file transfer, the bandwidth is reduced to the original value. The benefits are increased available bandwidth for all users and, as bandwidth is more efficiently utilized, a reduction in overall bandwidth costs. When IMUXes for the public switched digital network first appeared, standard inverse-multiplexing protocol did not exist. Therefore, each manufacturer developed its own proprietary protocol, requiring both ends of an inverse-multiplexed connection to use the same manufacturers equipment. There was no interoperability between inverse multiplexers manufactured by different vendors. Inverse-multiplexer manufacturers soon realized that, unless an inverse-multiplexer standard was implemented, inverse multiplexing would not reach its true potential as a data communications technology. Therefore, the Bandwidth-on-Demand Interoperability Group (BONDING) consortium was formed by a number of inverse-multiplexer manufacturers to develop an inverse-multiplexing standard. The resulting BONDING specification was published, implemented by vendors, tested, and publicly demonstrated. Today, nearly all inverse-multiplexing equipment implements at least a subset of the BONDING inverse-multiplexing protocol to assure interoperability between different vendors equipment. In wireless networks, bandwidth of more than 1E1 or 1T1 (unchannelized) per cell site may be required. This is usually the case in high-traffic areas
78
Transmission Systems Design Handbook for Wireless Networks
where RBS will be heavily loaded. In that case, backhaul circuits (RBS-BSC connection) are using IMUXes to divide the bandwidth into smaller, and generally more recognizable (and widely available), E1/T1 circuits. 3.4.3.3 The IMUX Metaframe
Metaframing is part of the inverse-multiplexing process. The IMUX metaframer has three major functions: 1. Creating a metaframe, larger than a DS1 frame, which aligns the received DS1 lines and permits corrections between their various delay times; 2. Providing a metaframe facility data link (MFDL) to the far end for managing logical channels and validating the receiver synchronizer; 3. Providing a metaframe parity indication of the logical channel data for an end-to-end integrity check. The IMUX supports two types of metaframes: the standard metaframe and the extended metaframe. Each is created by robbing the 17th bit in every transmitted frame. The standard metaframe length is 48 milliseconds, allowing the alignment of up to 22 milliseconds of differential delay. Using the extended metaframe (768 milliseconds in length) allows the IMUX to correct up to 45 or 125 milliseconds of differential delay, depending on the model. Again, the metaframe is created by robbing the 17th bit in every transmitted frame. This is the least significant bit in the second DS0 for T1 and the most significant bit in the second DS0 for E1. Because the normal DS1 payload is 1.536 Mbps (T1) or 1.984 Mbps (E1), the use of one bit for the metaframe leaves a user payload of 1.528 Mbps (T1) or 1.976 Mbps (E1). The MFDL, which runs at 4 Kbps within the metaframe, serves as a communication link to the far-end IMUX. The rate is constant, regardless of the number of lines in the logical channel. Each T1 or E1 contains an MFDL, but messages are sent on only one MFDL at a time. The MFDL provides the following: • Access to the far-end status and performance data; • An IP link to the far end that may be used for Telnet, SNMP, and
downloading;
• Management of the logical channel; • Identification of the T1/E1.
Transmission-Network Principles
79
Because the metaframe is carried in the payload, it will show errors when there have been impairments, even when there are intervening multiplexers. Counters like ES, UAS, and ESF are not usually end-to-end in most applications and will not increment when line errors or frame slips occur in a preceding T1 or E1 section (on the other side of a DS0 cross connect). MFE and MSYN errors may be the only indication of line problems (other than application errors). However, the presence of any line impairments that cause line errors to be reported will cause the occurrence of metaframe errors.
3.5 ATM 3.5.1
ATM Basics
ATM is the complement of STM. STM is a circuit-switched networking mechanism whereby a connection is established between two end points before data transfer commences, and torn down when it is completed. In this way, the end points allocate and reserve the connection bandwidth for the entire duration, even when not actually transmitting the data. ATM is a transmission technology that uses fixed-size packets called cells. A cell is a 53-byte packet with 5 bytes of header-descriptor information and 48 bytes of payload, or user traffic (voice, data, video, or a combination of these). Today, telecommunications companies are deploying fiber-optic cross-country and cross-oceanic links with Gbps speeds. They would like to carry, in an integrated way, both real-time traffic such as voice and high-resolution video, which can tolerate some loss but without delay, as well as non-real-time traffic such as computer data and file transfer, which may tolerate some delay, but not loss. The problem with carrying these different characteristics of traffic on the same medium in an integrated fashion is that the requirements of these traffic sources may be quite different. In other words, the data comes in bursts and must be transmitted at the peak rate of the burst, but the average arrival time between bursts may be quite large and randomly distributed. For these connections, it would be a considerable waste of bandwidth to reserve them a bucket at their peak bandwidth rate for all times, when on the average only 1 in 10 buckets may actually carry the data. And thus using STM mode of transfer becomes inefficient as the peak bandwidth of the link, peak transfer rate of the traffic, and overall burstyness of the traffic expressed as a ratio of peak to average, all go up. Terms like fast packet, cell, and bucket are used interchangeably in ATM literature.
80
Transmission Systems Design Handbook for Wireless Networks
ATM networks are connection-oriented packet-switching networks. Virtual circuits must be established between the end nodes before transmission can start and, as with any packet-switching network, routing of cells is performed at every node for each arriving cell. A virtual path identifier (VPI), an 8- or 12-bit field, together with virtual circuit identifier (VCI), a 16-bit field, contains the routing information of the cell [8]. The main idea rather than identifying a connection by the bucket number each time was just to carry the connection identifier along with the data in any bucket. This would reduce the size of the bucket so that if any one bucket got dropped en route due to congestion, less data would get lost, and in some cases, could easily be recovered. As this resembled packet switching, they called it fast packet switching with short fixed-length packets. The fixed size of the packets arose out of the desire of telecommunications companies to sustain the same transmitted voice quality as in STM networks, even in the event of some lost packets on ATM networks. The two end points in an ATM network are associated with each other via an identifier called the VCI label, instead of by a timeslot number as in an STM network. The VCI is carried in the header portion of the fast packet. The fast packet itself is carried in the same type of bucket used before, but without the label or designation for the bucket. The physical layer specification, while not explicitly a part of the ATM definition, is being considered by the same subcommittees, and SONET was standardized as the preferred physical layer. Therefore, the STS classifications refer to the speeds of the SONET link. STS-3c is 155.5 Mbps, STS-12 is 622 Mbps, and STS-48 is 2.4 Gbps. The SONET physical layer specifications outline a worldwide digital telecommunications network hierarchy internationally known as the SDH. It standardizes transmission around the bit rate of 51.84 Mbps that is also called STS-1, and multiples of this bit rate comprise higher bit rate streams. Thus, STS-3 is 3 times STS-1, STS-12 is 12 times STS-1, and so on. STS-3c is of particular interest as this is the lowest bit rate expected to carry the ATM traffic, and is also referred to as STM-1. The term SONET is the U.S. terminology for SDH. SDH specifies how payload data is framed and transported synchronously across fiber-optic transmission links without requiring all the links and nodes to have the same synchronized clock for data transmission and recovery (i.e., both the clock frequency and phase are allowed to have variations, or be plesiochronous). The intention is to allow products from multiple vendors across geographical and administrative boundaries to be able to plug and play in a standard, with the broadband ISDN network acting as a true international network. All of this rests below the ATM layer, and ATM
Transmission-Network Principles
81
cells are transported across the physical layer as payload, also called the SONET payload or the synchronous payload envelope (SPE). The physical layer is independent of the payload type, and can just as easily carry STM cells as ATM cells. Future telecommunication networks, including wireless networks, must be able to offer todays range of services as well as services with new features (e.g., variable bit rates). The requirements of modern networking involve handling multiple types of traffic (voice, video, and data), all with individual characteristics that make very different (and often opposed) demands of the telecommunication channel. The second requirement is the reliability and flexibility of the communication links. The greatest problem is that transmissions occur at statistically random intervals with variable data rates. A way of solving this problem is to take a service that takes packets on the transport layer from a higher layer and fragments them in small packets of a fixed size. The delays produced by each packet are going to be short and probably fixed, so if voice and video traffic can be assured priority handling, they can be mixed with data without diminishing any reception quality. The services are called the ATM adaptation layer (AAL) and the packets are called cells. ATM networks are connection-oriented, but the service offered is internally implemented using packet switching. In a circuit-switching network, a physical path is established from the source to the destination. In a virtual circuit network, like ATM, when a connection is established, no physical resources are allocated; rather, all the switches in between source and destination make table entries defining the route. All data will follow subsequently. When a packet comes along, the switch looks up its header information and relays it to the next switch indicated in the VCI information. This fulfills the need for variable bandwidth (i.e., if one flow needs more bandwidth, then another ATM can provide it). This is the virtue of packet switching using statistical multiplexing. Note that ATM uses packet switching to emulate circuit mode. Fixed-length packets offer many advantages because it is easier to switch packets when their length is known. There are two different cell formats, depending on where a cell is in the network: UNI and network-network interface (NNI). The UNI format is used when transmitting from host to switch and the NNI format when transmitting between switches, and there is only a minor difference between them. An overview of different ATM interfaces is illustrated in Figure 3.7. The advantage of having two different virtual connections is that it makes routing much easier between intermediate switches. This is accomplished by switching groups of VCIs on VPI level rather than on every single VCI. Only
82
Transmission Systems Design Handbook for Wireless Networks
Computer Private UNI
Private switch Private NNI
Computer Private Private switch UNI
Public UNI Carriers Public switch
Public NNI
Public switch B-ICI
Router
DXI
DSU
Carrier
Public switch
User-to-network interface (UNI): Public UNI, Private UNI Network-to-node interface (NNI): Private NNI (P-NNI) Public NNI = Interswitching system interface (ISSI) Intra-LATA ISSI (Regional Bell Operating Company) Intra-LATA ISSI (Interexchange carriers) ⇒Broadband Intercarrier Interface (B-ICI) Data exchange interface (DXI) Between routers and ATM digital service units (DSUs)
Figure 3.7 ATM interfaces overview.
the end switch will have to switch on VC level. Note that the virtual path (VP) acts as a fat pipe transporting bundles of thousands of virtual circuits. ATM cells can be transmitted over various carriers such as SDH/SONET, FDDI, or T1/E1. Traffic management is necessary to ensure that both the priorities of the user and the performance criteria of the network are met. The main threat to orderly traffic transmission is congestion and the possibility of having to start discarding cells, due to too many cells arriving at a point in the network. The network must cater to all types of traffic and deal with the different QoS requirements for the various traffic types. CBR connections, for example, are very delay and delay-variation sensitive. They require different parameters to variable bit rate (VBR) connections, such as LAN traffic. LAN traffic, due to its bursty nature, is well able to tolerate variations in end-to-end delay. It also has upper-layer software to protect it against lost or misinserted cells.
Transmission-Network Principles 3.5.2
83
Use of ATM Adaptation Layer
The AAL is the highest layer in ATM, and it is used to adapt traffic into an ATM format. In other words, the AAL maps application data into the ATM 48-byte cell payloads. The AAL function is performed at the edges of an ATM connection, and not within the network. As the AAL is used to adapt traffic to an ATM format, it is needed at the entry point to the ATM network. Once the traffic is adapted to the ATM cell format it travels across the network in ATM cells, which are switched in the ATM layer of switches along the path of the ATM connection. Once the cells reach their destination, there is a need to reassemble the traffic back into the format of the original application. The AAL is thus also used at the exit point of the ATM network. The high-level ATM protocol layer structure is shown in Figure 3.8. The ITU-T I.362 standard provides the functional descriptions for the AALs. Traffic classes are based on the following parameters: • Whether a timing relationship is required between source and
destination;
• Whether the traffic is CBR or VBR; • Whether the traffic is connection oriented or connectionless: •
Class A defines traditional synchronous data, such as that containing T1/E1 voice circuits or uncompressed broadcast video.
•
Class B covers compressed video that requires a timing relationship. End system
End system
ATM adaptation layer
Switch
ATM adaptation layer
ATM layer
ATM layer
ATM layer
Physical layer
Physical layer
Physical layer
Figure 3.8 ATM protocol layers.
84
Transmission Systems Design Handbook for Wireless Networks
• •
Class C defines bursty data such as frame relay, X.25, or large file transfer. Class D includes broadcast data, such as Service Access Point (SAP) messages in NetWare or an Address Resolution Protocol (ARP1) packet in TCP/IP.
There is also a Class X, which covers undefined or unspecified bit rate (UBR), where the user defines traffic type and timing requirements. Reference may be found, in ANSI documentation, to Class Y. In the ITU and ATM Forum documentation this is known as available bit rate (ABR). AAL5 is the ATM Forums response to the ITU-Ts excessively complex AAL3/4. AAL5 is sometimes known as simple and efficient adaptation layer (SEAL). The AAL is used at the entry and exit point of the ATM network, and all ATM switches require an AAL function. For example, signaling traffic needs to be interpreted by all switches. To interpret incoming signaling traffic, the ATM switch needs firstly to reassemble the signaling traffic. This requires an AAL function. The switch knows that traffic coming in from a given connection is signaling traffic, as a special reserved VPI/VCI value is used. Similarly, all ATM switches need to interpret operation and maintenance traffic and management traffic. The AAL provides services to application programs, taking packets from higher layers and fragmenting them so that they will fit in the cell payload. Five different types of AALs have been defined to serve the demands of the different services: • AAL1 is for real-time services with a constant bit rate. • AAL2 is used for real-time services with variable bit rates. By far the
most important benefit of AAL2 is the ability to substantially reduce bandwidth requirements for supporting voice on ATM networks, and the inherent flexibility to add features via the service-specific convergent sublayer (SSCS) structure. • AAL3 and AAL4 are for connection-oriented and connectionless transmission of non-real-time services. As there was no need for distinguishing AAL3 and 4, the two AALs merged to become AAL3/4. • AAL5 is a simplified version of AAL3. 1. A low-level protocol within the TCP/IP suite that maps IP addresses to the corresponding Ethernet addresses, ARP is used to obtain the physical address when only the logical address is known.
Transmission-Network Principles
85
In AAL0 no AAL functionality is used. The data is directly written in the payload field of the cells. AAL0 is no real AAL, as the functionality of adaptation is not needed. AAL1 is used for data streams with a constant bit rate. Clock information is transferred with the data. This layer is further split up into two sublayers: the segmentation and reassembly (SAR) sublayer and the convergence sublayer (CS). The 48-byte payload is split up into 47 bytes of data and 1 byte of header. The header contains a sequence number to detect lost packages or wrongly transmitted packages. AAL2 is responsible for the efficient transmission of delay-sensitive narrowband applications with variable bandwidth (see Figure 3.9). This means that for each transmission the QoS and the maximum cell delay or the maximum cell-loss rate must be guaranteed and the necessary bandwidth must be available. Typical applications are telephony and MPEG compressed video. In 3G wireless networks, AAL2 will be used, among other functions, to carry mixed voice and data traffic between the RBS and BSC (backhaul). AAL2 is new, and the data that would be supported by it was traditionally transmitted in AAL5 instead. A problem with the utilization of AAL5, however, is the lack of delay parameters. By contrast, AAL2 is inherently designed for the support of VBR traffic, for which timely delivery is an issue. A feature of AAL2 is the ability to accept several streams of traffic and multiplex them together. The manner of multiplexing is to accept samples and to append a small header to each sample. The primary function here is to add a channel number to identify the higher-layer stream. Once so labeled, blocks are then transferred to the ATM cell payloads. Part of this transfer is to add yet another header. The principal use of this header is to identify the start of Ideal for low-bit-rate voice Variable and constant rate voice Multiple users per VC Compression and silence suppression Idle channel suppression Payload 1
Payload 2
Payload 3
Pkt Payload 1 Pkt Payload 2 Pkt Payload 3 Hdr Hdr Hdr Pkt Payload 1 Pkt Payload 2 Pkt Payload 3 Cell Hdr Hdr header Hdr
Figure 3.9 AAL2 structure.
86
Transmission Systems Design Handbook for Wireless Networks
a flow after a short period of inactivity (one of the streams may show a blank screen for a couple of seconds and the stream may produce no output). AAL3 and AAL4 specify the connection-oriented and non-connectionoriented transmission of data packets in ATM networks. Connections can be point-to-point or PMP. Like AAL1, AAL3 and AAL4 are divided into an SAR and a CS sublayer; however, in the CS we distinguish a common part convergence sublayer (CPCS) and an application-specific part, the SSCS. AAL5 is a simplified version of AAL3 and AAL4. There is no possibility of multiplexing cells. AAL5 has significantly lower overheads than AAL3 and AAL4 and is, therefore, very widely adopted. In practice, AAL3 and AAL4 are seen as overly complex and cumbersome; only AAL1 and AAL5 are widely used. AAL1 is used for CBR traffic and AAL5 for all other: VBR, UBR, and ABR. SAAL is used for signaling protocols. SAAL guarantees the higher signaling layer a transmission of their messages with the Service-Specific Oriented Protocol (SSCOP). 3.5.3
Inverse-Multiplexing for ATM Protocol
Until recently, ATM services were limited to the core of the public network, or to enterprises with data volumes large enough to justify the T3/E3 and higher-rate facilities on which ATM was transported. Businesses with less substantial volumes of data were left to choose between dedicated T1/E1 lines and frame-relay services topping out at T1/E1 rates, and data pipes in the bandwidth gap between T1/E1 and T3/E3 were achievable only through inverse multiplexing. Recognizing the need for ATM at speeds lower than E3/T3 and STM-1/OC-3c, the ATM Forum has developed the inverse multiplexing for ATM (IMA) specification. As the name suggests, this specification offers a standard for ATM transmission at nxE1 (or nxT1) rates over inverse-multiplexed E1 (or T1) lines. Formally approved in 1997, IMA has received widespread acceptance, with numerous products implementing IMA currently on the market or in development. The IMA protocol defines a method for taking an ATM cell stream and breaking it down for transmission over multiple E1/T1 links. Where previous E1/T1 inverse multiplexing had been bit-based, with no regard for the content or structure of the original payload, IMA recognizes ATM cells and distributes them individually across the multiple links. The separate E1 or T1 links across which the ATM stream is transported are called an IMA group. IMA is very robust in its recovery features, including physical and logical layer monitoring, error detection, and correction. Physical link monitoring is based on loss of
Transmission-Network Principles
87
signal and bit error rates. For logical event monitoring, a link state machine function determines which links are stable and are maintaining data rates that meet the IMA group timing requirements. Links that are identified as part of the group are continuously monitored and can be taken in and out of the group dynamically. To properly order and account for the cells, a special IMA control protocol (ICP) cell is used to track timing information and frame sequence numbers, as well as other management information. The timing information is of particular importance, because it allows for some variance in clocking among the physical E1/T1s used in the IMA group. E1/T1 signals are delayed approximately eight milliseconds per 1,000 miles. Signals taking routes of different lengths have what is referred to as differential delay. Because the IMA protocol includes the ability to handle relatively large differential delays, an IMA group can include diversely routed E1/T1 facilities, either from a single carrier offering diverse routing or from multiple carriers. Even if the E1/T1 lines on one route are all brought down by a catastrophic failure, the lines on the other route will still generally be available, and the service can continue to run at a lower data rate until the affected lines are restored. 3.5.4
QoS in ATM Networks
ATM is a high-speed networking technique capable of supporting many different classes of traffic. CBR service is ideal for any data, text, or image transfer application that requires a fully reserved channel. VBR service is suitable for any application that can benefit from statistical multiplexing or that can tolerate or recover from potential and random packet loss. Real-time VBR service can be used by native ATM voice with bandwidth compression and silence suppression. Non-real-time VBR can be used for data transfer and frame-relay interworking. ABR provides economical support to applications that have vague requirements for throughput and delay and require a low cell-loss ratio. UBR offers a solution for less demanding applications that take advantage of any spare bandwidth and profit from it. UBR is a besteffort service with no guarantees. QoS metrics in ATM networks can be categorized into two different classes: the first one includes the call control parameters associated with connection-oriented networks and the second class is the set of information transfer parameters defined in packet networks. 3.5.4.1 Call Control Parameters
Three of the most important components of interest in connection-oriented networks are setup delay, connection release delay, and connection acceptance
88
Transmission Systems Design Handbook for Wireless Networks
probability. Connection setup delay is the time interval between the call setup message transfer to the call setup acknowledge message transfer, excluding the called users response time. This is not an ATM-specific parameter and is mainly determined by the processing delays at various signal transfer points in the network. Connection release delay is the time interval between the callrelease message transfer and receipt of the call-release acknowledge message. This is also an ATM-independent parameter. Connection acceptance probability is the proportion of the accepted calls over a long period of time, which is also referred to as the blocking probability in telephony networks. This probably is one of the most important performance metrics used in network design and allocation of network resources (such as network topology, switch capacities, link capacities, and so on). Parameters, such as call attempt rate, average call holding time, and busy-hour call attempts are used for the initial network planning. 3.5.4.2 Information Transfer Parameters
Information transfer parameters in ATM networks include cell information field bit error ratio (BER), cell-loss ratio (CLR), cell-insertion ratio (CIR), end-to-end transfer delay, cell delay variation (CDV), and skew. BER is defined as the ratio of the bit errors in the information field to the total number of bits transmitted in the information field. Assuming that errors occur randomly, the probability that there is no bit error in the 48byte (384 bit) ATM cell payload on a link is
(1 − BER ) 384 Assuming BER = 10−6, this probability is 0.99617 and increases to 0.999996 with BER = 10−9. CLR is the ratio of the number of lost cells to the total number of cells sent by a user within a specified time interval. This metric is ATM-specific and has an important impact on the quality of service provided to users. All AALs, except type 5, include cell sequence numbers to detect lost cells at the receiver. The effect of cell losses and actions taken on lost cells in ATM networks is different for different types of services. It is possible that an error occurring at the header (CIR) may not be detected by header error checking. In this case, if an undetected change in the bit pattern of a cell header corresponds to the address of another connection, then the cell is misrouted to a wrong destination.
Transmission-Network Principles
89
Cell-transfer delay (CTD) is caused by different factors like coding delay, packetization delay, propagation delay, transmission delay (5 microsec/km in fiber-optic cable), switching delay, queuing delay, and reassambly delay. CDV (jitter) consists of the constant delay component (CTD) and the random delay component that arises out of buffering within the network. The only way of controlling jitter at the receiver is at the expense of large buffers and delaying cells. Skew is defined as the difference in the presentation times of two related objects (like video stream and audio stream) and is important in multimedia applications, where it can cause delay between an image and accompanying voice or between lip motion and voice. A key consideration in the ATM standard is given to the concept of VPs. These are groups of VCs bundled together to reduce setup, switching, and control costs. Although VPs have not yet been extensively deployed, their use in networks that process a large number of switched VCs, or calls, is fundamental [9]. Many of the QoS concepts and notions developed for ATM networks explicitly depend on the network providers ability to integrate different services with different service characteristics over the same physical transport trunks via VPs. Performance management entails the periodic evaluation of ATM equipment and software. The idea is to assess the ATM system in a systematic way in order to determine how the network is performing and if error conditions are acceptable. Performance management consists of forward monitoring, backward monitoring, and reporting [10]. Forward monitoring means generating cells from one network element to a receiving network element. Backward monitoring is checking the cells at the receiving network element and reporting back to the generating network element. Reporting usually means storing the results of the monitoring activities based on the filtering of selected parameters and thresholds. An excellent source of information regarding ATM QoS, required connection availabilities, and performance management is ITU-T Recommendation I.356, B-ISDN ATM Layer Cell Transfer Performance (03/2000) and ITU-T Recommendation I.357, B-ISDN Semi-permanent Connection Availability (11/2000). 3.5.5
Definition of Availability in ATM
From a dependability point of view, a portion of an ATM semipermanent connection should have the following properties:
90
Transmission Systems Design Handbook for Wireless Networks
• The fraction of time during which it is in a down state (i.e., unable
to support a transaction) should be as low as possible. • Once a transaction has been established, it should have a low probability of being either terminated (because of insufficient data transfer performance) or prematurely released (due to the failure of a network component) before the intended end of a transaction. Availability of an ATM semipermanent connection portion is defined as the fraction of time during which the portion is able to support a transaction. Conversely, unavailability of a portion is the fraction of time during which the portion is unable to support a transaction (i.e., it is in the down state). A common availability model is used which applies to any semipermanent connection type. The model uses two states corresponding to the ability or inability of the network to sustain a connection in the available state. Transitions between the states of the model are governed by the occurrence of patterns of severely errored seconds in the ATM layer (SESATM). In order to define the availability of an ATM semipermanent connection portion, a criterion is defined for entry into the unavailable state. This criterion is applicable to any ATM semipermanent connection portion, whether the user continuously transmits cells or not. This is achieved by defining a cell transfer outcome, the SESATM. A given second is considered to be an SESATM in the following cases: • User information cells are presented during this period of time
to the connection portion and either the CLR > 1/1,024 or the severely errored cell block ratio (SECBR) > 1/32, where CLR and SECBR are computed over the considered period of time. The above CLR threshold is intended to support QoS classes in which the CLR objective is ≤ 10−5. • User information cells are not presented during this period of time to the connection portion, but the ATM connection is considered to be unable to provide acceptable cell transfer performance, because an interruption has occurred within the connection portion. This interruption prevents cells from being transmitted on the connection portion during the considered one-second period of time, should the user attempt to transmit cells. An interruption corresponds to a failure occurring within the connection portion, either of the physical layer or of the ATM layer.
Transmission-Network Principles
91
The methods for estimating the occurrence of an SESATM are taken from the set of cell-transfer parameters defined in ITU-T I.356 and the OAM facilities defined in ITU-T I.610. The onset of unavailability begins with the occurrence of 10 consecutive SESATM. These 10 seconds are part of unavailable time. A period of unavailability ends with the occurrence of 10 consecutive seconds, none of which SESATM. These 10 seconds are part of available time. The 10-second criteria are supported using a sliding window with one-second granularity [11]. A portion of a bidirectional B-ISDN connection is available if and only if both directions are available. It is recognized that in-service measurement of availability as defined above may not be practicable in many cases. Performance objectives are defined for two availability performance parameters: availability ratio (AR) and mean time between outages (MTBO). AR applies to ATM semipermanent connection portions. The AR is defined as the proportion of scheduled service time that the connection portion is in the available state. The AR is calculated by dividing the total service available time by the duration of the scheduled service time. During the scheduled service time the user may or may not transmit cells. MTBO applies to ATM semipermanent connection portions. The MTBO is defined as the average duration of continuous periods of available time, but where scheduled service time is not contiguous they are concatenated in calculating MTBO.
3.6 Voice over IP 3.6.1
H.323 Network Building Blocks
The transfer of voice traffic over packet networks, and especially voice over IP (VoIP), is rapidly gaining acceptance. Many industry analysts estimate that the overall VoIP market will become a multi-billion-dollar business within the next few years. While many corporations have long been using voice over frame relay (VoFR) to save money by utilizing excess FR capacity, the dominance of IP has shifted most attention from VoFR to VoIP. These high-speed backbones take advantage of the convergence of Internet and voice traffic to form a single managed network. This network convergence also opens the door to novel applications. Interactive shopping (Web pages incorporating a click-to-talk button) is just one example, while streaming audio, electronic white-boarding, and CD-quality conference calls in stereo are other exciting applications. But along with the initial excitement, potential users are also worried over possible degradation in voice quality when voice is carried over these
92
Transmission Systems Design Handbook for Wireless Networks
packet networks. Whether these concerns are based on experience with the early Internet telephony applications, or whether they are based on understanding the nature of packet networks, voice quality is a critical parameter in acceptance of VoIP services. As such, it is crucial to understand the factors affecting voice over packet transmission, as well as obtain the tools to measure and optimize them. VoIP services need to be able to connect to traditional circuit-switched voice networks, and the ITU-T has addressed this goal by defining H.323, a set of standards for packet-based multimedia networks. The basic elements of the H.323 network are shown in the network diagram (Figure 3.10) where H.323 terminals such as PC-based phones (left side of drawing) connect to existing ISDN, PSTN, and wireless devices (right side). The H.323 components include the following: • H.323 terminals that are end points on a LAN. • Gateways that interface between the LAN and switched-circuit
network.
• A gatekeeper that performs admission-control functions and other
duties.
• A multipoint control unit (MCU) that offers conferences between
three or more end points.
H.323 terminals are LAN-based end points for voice transmission and they support real-time, two-way communications with other H.323 entities. H.323 terminals implement voice-transmission functions and specifically include at least one voice compressor/decompressor (CODEC) that sends and receives packetized voice. Common CODECs are ITU-T G.711 (pulse code modulation [PCM]), G.723 (MP-MLQ), and G.729A (CA-ACELP). CODECs differ in their central processing unit (CPU) requirements, in the resultant voice quality, and in their inherent processing delay. Terminals also need to support signaling functions that are used for call setup, teardown, and so forth. The applicable standards here are H.225.0 signaling, which is a subset of ISDNs Q.931 signaling; H.245, which is used to exchange capabilities such as compression standards between H.323 entities; and registration, admission, status (RAS), which connects a terminal to a gatekeeper. Terminals may also implement video and data communication capabilities. The gateway serves as the interface between the H.323 and non-H.323 network. On one side, it connects to the traditional voice world, and on
Gateway H.323 terminal
H.323 terminal
Gatekeeper
Wireless
Router
PSTN
Router MCU
Transmission-Network Principles
ISDN Router
Enterprise network Gateway 93
Figure 3.10 VoIP principle diagram.
94
Transmission Systems Design Handbook for Wireless Networks
another side to packet-based devices. As the interface, the gateway needs to translate signaling messages between the two sides as well as compress and decompress the voice. A prime example of a gateway is the PSTN-IP gateway, connecting an H.323 terminal with the circuit-switched network. There are many types of gateways in existence today, ranging from support of a dozen or so analog ports to high-end gateways with simultaneous support for thousands of lines. The gatekeeper is not a mandatory entity in an H.323 network. However, if a gatekeeper is present, it must perform a set of functions. Gatekeepers manage H.323 zones, a logical collection of devices (for example, all H.323 devices within an IP subnet). Multiple gatekeepers may be present for load balancing or hot-swap backup capabilities. The philosophy behind defining the gatekeeper entity is to allow H.323 designers to separate the raw processing power of the gateway from intelligent network-control functions that can be performed in the gatekeeper. A typical gatekeeper is implemented on a PC, whereas gateways are often based on proprietary hardware platforms. Gatekeepers provide address translation (routing) for devices in their zone. This could be, for instance, the translation between internal and external numbering systems. Another important function for gatekeepers is providing admission control, specifying what devices can call what numbers. Among the optional control functions for gatekeepers are providing SNMP management information, offering directory and bandwidth management services. The MCU allows for conferencing functions between three or more terminals. Logically, an MCU contains two parts: 1. A multipoint controller (MC) that handles the signaling and control messages necessary to set up and manage conferences; 2. A multipoint processor (MP) that accepts streams from end points, replicates them, and forwards them to the correct participating end points. An MCU can implement both MC and MP functions, in which case it is referred to as a centralized MCU. Alternatively, a decentralized MCU handles only the MC functions, leaving the multipoint processor function to the end points. It is important to note that the definition of all the H.323 network entities is purely logical. No specification has been made on the physical division of the units. MCUs, for instance, can be stand-alone devices, or they can be integrated into a terminal, a gateway, or a gatekeeper.
Transmission-Network Principles 3.6.2
95
Latency and Jitter Issues
In contrast to broadcast-type media transmission, a two-way phone conversation is quite sensitive to latency. Most callers notice round-trip delays when they exceed 250 ms, so the one-way latency budget would typically be 150 ms, also specified in the ITU-T G.114 recommendation as the maximum desired one-way latency to achieve high-quality voice. Beyond that roundtrip latency, callers start feeling uneasy about holding a two-way conversation and usually end up talking over each other. At 500-ms round-trip delays and beyond, phone calls are impractical. For reference, the typical delay when speaking through a geostationary satellite is 150 to 500 ms. Data networks are not affected by such delay. An additional delay of 200 ms on an e-mail or Web page goes mostly unnoticed. But when sharing the same network, voice callers will notice the delay. When considering the one-way delay of voice traffic, one must take into account the delay added by the different segments and processes in the network. Some components in the delay budget need to be broken into fixed and variable delay. For example, for the backbone transmission there is a fixed transmission delay, which is dictated by the distance, plus a variable delay, which is the result of changing network conditions. The most important components of this latency include the following: • Backbone (network) latency. This is the delay incurred when travers-
ing the VoIP backbone. In general, to minimize this delay, it is necessary to minimize the number of router hops between end points. Some service providers are capable of providing an end-to-end delay limit over their managed backbones. Alternatively, it is possible to negotiate or specify a higher priority for voice traffic than for delayinsensitive data.
• CODEC latency. Each compression algorithm has certain built-in
delay. For example, G.723 adds a fixed 30-ms delay. When this additional gateway overhead is added in, it is possible to end up paying 32 to 35 ms for passing through the gateway. Choosing different CODECs may reduce the latency, but it may also reduce quality or result in more bandwidth being used.
• Jitter buffer depth. To compensate for the fluctuating network condi-
tions, many vendors implement a jitter buffer in their voice gateways. This is a packet buffer that holds incoming packets for a specified amount of time before forwarding them to decompression. This has the effect of smoothing the packet flow, increasing the
96
Transmission Systems Design Handbook for Wireless Networks
resiliency of the CODEC to packet loss, delayed packets, and other transmission effects. The downside of the jitter buffer, however, is that it can add significant delay. The jitter buffer size is configurable and can be optimized for given network conditions. The jitter buffer size is usually set to be an integral multiple of the expected packet interarrival time in order to buffer an integral number of packets. It is not uncommon to see jitter buffer settings approaching 80 ms for each direction. While network latency affects how much time a voice packet spends in the network, jitter controls the regularity in which voice packets arrive. Typical voice sources generate voice packets at a constant rate, so the matching voice decompression algorithm also expects incoming voice packets to arrive at a constant rate. However, the packet-by-packet delay inflicted by the network may be different for each packet. The result is that packets that are sent in equal spacing from the left gateway arrive with irregular spacing at the right gateway. There are three interesting configurations for measuring latency: 1. Measuring the latency of a device; 2. Measuring round-trip delay; 3. Measuring one-way delay. Measuring the latency of a device is important to understand how the delay budget gets spent over the network. In particular, it is interesting to measure the latency of data going through a gateway, since several user-configurable parameters, such as jitter-buffer size, affect the latency. Thus, after configuring such parameters, it is important to be able to verify that the gateway actually behaves as expected. Some products allow measuring the latency by generating controlled data through an ingress port and capturing it off an egress port. Some protocol analyzers can operate two technologies at the same time, with a synchronized timestamp that allows interport or intertechnology latency measurement. The analyzer can measure the latency through a network using a similar method. When the two end points are geographically distant, it is often less convenient to perform one-way latency measurements, because such an operation requires synchronizing the control and timestamp of two separate analyzers. Instead, many users measure the round-trip time and assume it is twice the one-way time for each direction. Round-trip measurements can be
Transmission-Network Principles
97
done using a protocol analyzer, or as a first approximation using the ping utility, generating Internet Control Message Protocol (ICMP) echo requests through the network. Since the receiving decompression algorithm requires fixed spacing between the packets, the typical solution is to implement a jitter buffer within the gateway. The jitter buffer deliberately delays incoming packets in order to present them to the decompression algorithm at fixed spacing. The jitter buffer will also fix any out-of-order errors by looking at the sequence number in the RTP frames. While the voice decompression engine receives packets directly on time, the individual packets are delayed further in transit, increasing the overall latency. Jitter is calculated based on the interarrival time of successive packets. Frequently, two numbers are given: the average interarrival time and the standard deviation. On a good network, the average interarrival time will be the interarrival time of the emitted packets, and the standard deviation will be lowpointing at a consistent interarrival time. When correct jitter measurements are desired for audio streams, it is important to take into account three phenomena: silence suppression, packet loss, and out-of-sequence errors. CODECs take advantage of periods of silence in the conversation to reduce the number of packets being sent. Typically, up to 50% bandwidth savings can be realized in this way. The RTP packet immediately after a period of silence is marked with the silence-suppression bit. Jitter calculations look at the silence-suppression bit and disregard the long gap between the packet right before the silence and the packet right after the silence period. In the event of packet loss, the interarrival time between two successive packets will also appear excessive. For instance, if three packets were sent at a time of 0, 20, and 40 ms, and the second packet was lost in transit, the interarrival time would appear to be 40 ms even if the network induced no jitter. Correct jitter measurements would discover these cases by looking at the packet sequence number and compensate for packet loss in the jitter calculation. 3.6.3
Multiprotocol Label Switching
Multiprotocol label switching (MPLS) is a layer-3 switching technology aimed at greatly improving the packet-forwarding performance of the backbone routers in the Internet or other large networks. The basic idea is to forward the packets based on a short, fixed-length identifier termed a label,
98
Transmission Systems Design Handbook for Wireless Networks
instead of the network-layer address with variable length match. The labels are assigned to the packets at the ingress node of an MPLS domain. Inside the MPLS domain, the labels attached to packets are used to make forwarding decisions. Thus, MPLS uses indexing instead of a longest address match as in conventional IP routing. The labels are finally popped out from the packets when they leave the MPLS domain at the egress nodes. By doing this, the efficiency of packet forwarding is greatly improved. Routers that support MPLS are known as label switching routers (LSRs). Although the original idea behind the development of MPLS was to facilitate fast packet switching, currently its main goal is to support traffic engineering and provide QoS [12] within IP networks. The goal of traffic engineering is to facilitate efficient and reliable network operations, and at the same time optimize the utilization of network resources. Most current network routing protocols are based on the shortest path algorithm, which implies that there is only one path between a given source and destination end system. In contrast, MPLS supports explicit routing, which can be used to optimize the utilization of network resources and enhance trafficoriented performance characteristics. For example, multiple paths can be used simultaneously to improve performance from a given source to a destination. MPLS provides explicit routing without requiring each IP packet to carry the explicit route, which makes traffic engineering easier. Another advantage is that using label switching, packets of different flows can be labeled differently and, thus, receive different forwarding (hence, different QoS). The differentiated-services (diffserv) model allows (theoretically) definition of 63 different classes of service, each of which can be mapped to the applications being used by the customer. A label-switched path (LSP) is referred to as a path from the ingress node to the egress node of an MPLS domain followed by packets with the same label. A traffic trunk is an aggregation of traffic flows of the same class, which are placed inside an LSP. Therefore, all packets on a traffic trunk have the same label and the same three-bit class-of-service field in the MPLS header. Traffic trunks are routable objects like virtual circuits in ATM and frame-relay networks. These trunks can be established either statically or dynamically (on demand) between any two nodes in an MPLS domain. A trunk can carry any aggregate of microflows, where each microflow consists of packets belonging to a single TCP or UDP flow. In general, trunks are expected to carry several such microflows of different transport types.
Transmission-Network Principles 3.6.4
99
IP-Based Wireless Networks
The definition and specification of the all-IP network is now taking place in 3GPP2, and some equipment suppliers have proposed a phased, evolutionary stepwise migration toward the all-IP network. This evolutionary approach makes it possible for operators to choose a migration pace that fills their unique needs. Packet Core Network (PCN) will serve as the cornerstone in an operators strategy to move from todays circuit-switched environment toward the Internet-enabled world of tomorrow. IP-based transport protocols are introduced in the transport and signaling legacy network. The anticipation is that the signaling will still be access specific while carried over an IP protocol. Payload protocols will most likely be evolved from current IETF protocols and initiatives, taking into account wireless requirements. IP-based transport protocols are introduced in the legacy core network. More specifically, it is proposed to carry VoIP instead of PCM-structured trunks. The protocols used in the CS network should be to the largest possible extent based on open IETF-based protocols. In the current packet data architecture there is an inherent tight connection between the CS domain and the PS domain. The current architecture requires the MSC for setting up and maintaining packet data services. Legacy voice services should evolve toward IP transport, and PS and CS services will evolve independently toward IP. Furthermore, the packet data services should evolve from the current architecture. The tasks that in GPRS are performed by the SGSN (e.g., packet session control and mobility management) are handled by the MSC in a CDMA2000 system. As a consequence, it has been shown that it is difficult to make significant changes to the session control and mobility management of packet data, because it has an immediate effect on the A interface, the MSC, and on ANSI-41. To replace this MSC-centric model so that it can serve all the foreseeable and unforeseeable features needed for multimedia services, a new entity called access control server (ACS) has been proposed. The ACS can be seen as an entity that manages the control part of the MSC for packet data without being involved in the actual switching or routing of user packet data. By introducing the ACS, the PS domain will be free to evolve independently of the CS domain, thereby enabling a smooth migration to all-IP. The multimedia call models are provided in the packet core network. The network is not only a bearer network, but also an IP multimedia service provider network able to provide multimedia services to users. Signaling gateways and media gateways are introduced in order to provide seamless
100
Transmission Systems Design Handbook for Wireless Networks
interworking between VoIP and the legacy ISUP-based networks, such as the PSTN. The shortcoming of so-called best-effort packet data is that it has limited ability to distinguish between or prioritize different types of payloads. This is sufficient for many types of traffic, such as Web browsing, e-mail download, or file transfers. On the other hand, other types of traffic, such as voice or videos have their own special requirements with respect to delay, jitter, and bit errors. For this reason it is necessary to introduce two controls: • QoS control differentiates among different types of payloads, ensur-
ing that each is handled according to its own QoS profile. • Admission control monitors the available free bandwidth and does not admit more traffic than can be handled. In order to gain from the service flexibility that IP as a service mechanism offers, the transparency of the IP protocols must be maintained. This costs spectrum in terms of packet overhead. Using the IP protocol family implies a high ratio of overhead, especially when considering narrowband services like voice. Voice CODECs generate frames of typically 15 to 30 bytes every 20 to 30 ms, whereas the Realtime Transport Protocol (RTP), IP, and Transmission Control Protocol (TCP) overhead (IPv4) is 40 bytes. Hence, we have more than 50% overhead in the VoIP. The IP header is approximately equal to 20 octets, and a speech frame is 35 octets (i.e., the IP header gives a significant overhead). In narrowband systems where transmission efficiency is important, header compression must be used. This will reduce the header to 3 or 4 octets. There are proposals for an IP header compression standard, but no standard has been accepted yet. The header compression is used on a per-link basis, which means that it does not have to be used on all links. Companies have been actively contributing to Internet Engineering Task Force (IETF) in order to standardize robust IP header compression algorithms that address these problems, and soon after IP header compression for wireless applications is fully standardized, it will be easy to implement in the base station and mobile terminal. Every host [RBS, processor in BSC, operations support systems (OSS), and so on] will have an IP address. The BSC processors have the responsibility for call control in the Base Station Subsystem (BSS). But the BSC will have no responsibility for the Operations and Maintenance (O&M) for the RBS. The OSS and the RBS will handle all O&M functionality. An IP network example is given in Figure 3.11. The IP network clouds shown could
Transmission-Network Principles
101
Internet
PSTN MSC CPS
GGSN SGSN
ATM IP
BS RBS
PDH/ SDH
TDM pipes for ATM
IP
IP RBS
BSC BS
ATM
ATM
RNC
Figure 3.11 IP-based wireless system.
be existing BSS transmission networks or they could be new Intranets built for the wireless operators data-communications usage. The basic idea is that the BSC consists of a number of processors that communicate with RBS using IP. Real-time traffic must have priority over TCP traffic, in order not to get too long delays. Some routers have RSVP and priority already implemented. Within the IETF there is a lot of activity within the field called differential services. The idea is to define something fairly simple and straightforward so that the present routers can be software upgraded. In a packet network the delay is never constant, even if the packets are delivered with priority. Therefore, the end points must have buffers to be able to cope with the delay variation. The question is how big is the variation and how big should the buffers be. If the absolute time is known, the delay variation can be kept at a minimum. A long packet can delay the sending of shorter speech packets, if the long packet started to be transmitted when the speech packet queue was empty. On a 2-Mbps link, 1,600 octets will take 6.4 ms to be transmitted and accepted. On a 384-Kbps link, 1,600 octets
102
Transmission Systems Design Handbook for Wireless Networks
will take approximately 33 ms to transmit, which is too much delay for the speech packets, so the long packets must be fragmented to a suitable size depending on the link rate. Certain requirements of the radio access network set a challenge to IP deployment. A traditional packet network cannot carry synchronization signals, and special means must be provided to base stations. Current strict delay requirements and low-bandwidth transmission links between the RBS and BSC need careful network planning to enable smooth evolution toward IP.
3.7 Complete T1 Tutorial 3.7.1
Signals in a T1 Network
The telecommunications user company must design its own facilities, purchase the correct products from various vendors, install and test the products properly, and document what it has for future reference. This section addresses engineering planning and design of a T1 line so that service can be delivered on a more predictable schedule. Details of the T1 systems, used mainly in the United States, Canada, and Japan (but also in Jamaica and some other countries around the world), are presented in the material that follows. The 1.544-Mbps bipolar PCM line signals of T1-type terminal equipment (such as channel banks, data terminals, and repeatered lines) are designated DS1, meaning digital signal, level one. At the standard cross-connect point for DS1 signals, DSX-1, the voltage level of pulses is about +3 volts and −3 volts. DS1 means that the signal meets the interface specification for DS1 signals (1.544 Mbps, bipolar +3 and 3 volt pulses, 50% duty cycle, and so on). The data pattern means that the 192 bits of payload data contain live traffic (thats 24 timeslots and 8 bits per timeslot) or some test pattern. After each 192 bits, there is a framing bit for bit number 193. That bit is either 1 or 0, depending on whether it must follow the special pattern for superframe format (SF, also referred to as D4) or ESF [13]. For details see Figure 3.12. This 193-bit frame is repeated 8,000 times per second. This is also called a frame rate of 8 kHz. Each of 24 channels is sampled and transported 8,000 times per second. If a terminal receiver cannot determine the start and finish of a frame, then receivers will always be out of frame sync with respect to transmitters, and very poor performance is noted. There is one more qualifier on the signal, and that is line code. The two line codes are alternate mark inversion (AMI) and bipolar eight zero substitution (B8ZS). This is very important to network elements that carry live traffic.
Transmission-Network Principles
103
One DS-1 frame 192 payload bits (typically 24 samples of 8 bits per sample)
193rd position for frame bit
Figure 3.12 Pulse-code modulation.
In the data communications world, all digital data is either 1 or 0, nicknamed either mark or space. If the data has a pattern of all 1s, then the first 1 is transmitted on the AMI transmission link as +3 volt pulse, the next 1 as −3 volt pulse, and the next 1 as +3 volt pulse. Internally to the electronics of equipment, the 1.544-Mbps signal might be unipolar, but everywhere accessible to the user, the signal is bipolar. DS1 signals may fall into one of three categories: SF, ESF, and possibly unframed. Sometimes test equipment will use an unframed DS1 test signal to check a transmission facility, but most terminal equipment must be selected for SF or ESF only. If terminal equipment expects to see SF and it receives ESF instead, it will probably not be able to lock its framer circuit onto the input signal and everything will stay in alarm. A T1 line provides a physical four-wire transmission path for cable carrier systems that transmit bipolar pulse streams at bit rates of 1.544 Mbps. At each CO, cable pairs connect to an office repeater. Between offices, line repeaters are located at nominal spacing of 32 dB at 772 kHz. Automatic line buildout (ALBO) equalizers in each repeater can compensate for a range of losses in the preceding cable section. Note that the overview diagram in Figure 3.13 is simplified by the use of one line in each direction to symbolize one twisted pair of wires. Other diagrams use one line to indicate both directions (two twisted pairs), and still other diagrams use four lines to indicate both pairs. The nature of the diagram dictates which symbol convention is used. The triangular symbol indicates a digital regenerator, which is somewhat related to an amplifier. Note that office repeaters use a regenerator in only one direction, and line repeaters use regenerators in two directions. A T1 channel service unit (CSU) interfaces a typical piece of customer equipment (such as a channel bank) with a public T1 transmission facility. There are several specific types of CSUs and the most common type has a DS1 signal interface on each side of the unit (the data terminal side and the
104
Transmission Systems Design Handbook for Wireless Networks
End office Line repeaters
DSX-1
End office Line repeaters
DSX-1
Mux or Channel bank
Mux or Channel bank
DSX-1
Intermediate office
Office repeaters
Office Office repeaters repeaters
Office repeaters
Figure 3.13 Transmission paths of a PCM system.
network side). In some respects, this CSU resembles an office repeater; however, additional diagnostic features are present on most CSU products. One of the most common functions is that of ones density enforcement (also referred to as ones stuffing). Customer DS1 signals containing very long strings of 0s are not allowed onto the public network (since there may be network elements that require occasional 1s for timing purposes). As a result, most CSU devices change a 0 to 1 to suppress the 16th consecutive 0, depending on specific requirements for a public facility. If the traffic has simply voice circuits, an occasional forced error of this sort is a negligible problem. If, on the other hand, the traffic is high-priority unrestricted data, then an occasional error is not acceptable. If this is the case, then the line code must be selected as B8ZS and not standard AMI. B8ZS uses selected bipolar violations (BPVs) as a means of signaling the far end that strings of 0s are in the customer data. Therefore, no errors are introduced and the system works correctly. If B8ZS line code is used, then each and every network element in the path must be provisioned for B8ZS instead of AMI (or else the element must completely ignore line codes). Note that some T1 office repeaters are aware of B8ZS and others are not aware. Note that one line code or the other will not disrupt these repeaters, but many have BPV monitors that will detect the unintentional BPVs but ignore the intentional BPVs that are part of B8ZS coding.
Transmission-Network Principles 3.7.2
105
Pulse Transmission
Pulses generated by the terminal equipment (e.g., channel bank) and repeaters are subject to distortion by attenuation and phase characteristics of the cable. In the line and office repeater units, just preceding the actual regenerator is an ALBO equalizer, which restores adequate pulse shape for detection and regeneration. Pulses generated in the terminal equipment must reach the office repeater in a predictable fashion, even if it is in the same room. In many office repeaters, the line build out (LBO) setting tells it through how many feet of cable the signal has traveled since the DSX-1 cross connect. The office repeater adapts to that attenuated signal. Most LBO settings assume the use of 22-gauge twisted-pair cable (with around 1416 pF/ft of capacitance). Note that 24-gauge cable has approximately 25% more loss than 22 gauge. By default, many office repeaters are shipped set for 0 to 133 ft of cable (assuming 22 gauge). For transmission calculations, the cable attenuation at 772 kHz is used. This is half of the 1.544-Mbps clock rate, but this is valid because the power spectrum of the pulse stream is maximum at approximately 772 kHz. 3.7.3
BERs
Pulses sent along the repeatered line are regenerated at each repeater point. The repeater looks at each timeslot and decides whether or not a pulse is present. If the logic circuit determines that there is a pulse, the repeater outputs a new pulse that is free of noise, distortion, or interference. As a result of many possible factors, a few pulses may be incorrectly regenerated. A 0 is sent instead of a 1, or vice versa. The ratio of error pulses to the total number of timeslots is called the error rate or the BER. One error in 1,000 is referred to as BER 10−3; one error in a million is BER 10−6. −6 −7 For strictly voice circuit application to T1, a BER of 10 or 10 might −8 −9 be considered acceptable performance, although a BER of 10 or 10 might be required for some data purposes. In the most common systems, once live traffic errors have entered into the transmission stream, they cannot be sorted out or corrected. Error rates tend to accumulate through an end-to-end system, although the rates tend to be low for digital systems as compared with analog systems. Voice is more forgiving of transmission errors or high BER than is data (also called nonvoice content). The reason for this forgiveness comes from the ability of the human brain to reconstruct the missing parts of conversations. If a syllable or even a word is dropped, the brain assumes the missing information from the context of the conversation and reconstructs it. This
106
Transmission Systems Design Handbook for Wireless Networks
enables people to communicate through the noise or the breakup of a marginally understandable mobile radio call. The transmission of data does not provide this reconstruction advantage. If nonvoice content is lost, it can only be recovered, if at all, through sophisticated error-correction algorithms. Such error corrections add overhead that slows the true data rate, or transmission rate, of the desired content. The greater the number of errors, the more error correction required and the slower the true data rate. 3.7.4
Overall System Length
Each repeater in the series adds a small amount of jitter to every pulse of the bit stream. A conservative limit of 200 tandem repeaters in a system ensures that the accumulated jitter will not exceed the synchronization capability of terminal equipment or higher-order multiplexers. Based on the accumulation of error rates in tandem repeater sections, and particularly in end sections (the repeater section next to the CO), the system should not include more than ten tandem span lines (nine intermediate offices). Long-range design of the span line is necessary to plan the expected cross section of the span. The selection of one-cable or two-cable operation, locations for the line repeaters, and repeater section length will depend on the future requirements of the route. Also, the cable plant must be carefully studied in terms of the number, type, and age of cables; freedom from bridge taps and branches; splicing integrity; suitability for line repeater locations; and minimum exposure to electrical and mechanical hazards. Major factors that control the design of the span include the following: • Ultimate number of systems within the cable; • Cable-pair attenuation at 772 kHz; • Cross-talk coupling loss between cable pairs; • CO noise; • Ambient temperature range.
If a maximum length section is engineered, cable loss should be measured before the final repeater location is established. This allows for changes and small errors. Note that aerial cable can be exposed to higher ambient temperatures compared with buried cable and ducted cable. Higher temperature equates to higher copper attenuation and dc resistance. For design purposes,
Transmission-Network Principles
107
the losses for aerial cable are estimated at about 5% higher than for buried cable. 3.7.5
Single-Cable and Dual-Cable Operation
In normal one-cable operation, low-level repeater inputs and high-level repeater outputs appear at the same point of the cable. As a result, near-end cross-talk (NEXT) is the limiting factor in repeatered line design. The number of systems that can be installed in a single cable is mainly controlled by the physical separation of the pairs in the two directions of transmission. Greater separation increases the coupling loss, resulting in decreased interference. A general rule is that if transmit and receive pairs are in the same cable binder group, the maximum section loss should be reduced to 15 dB to prevent cross talk. In two-cable operation, NEXT does not limit the number of systems for one cable. The choice of one-cable or two-cable operation is based on cable route, circuit requirements, availability of suitable cables, and economics. A span line includes one or more repeater sections. Typical maximum section loss is 32 dB (normal section) measured at 772 kHz. In the end section next to the CO, maximum loss is limited to 50% to 70% of the normal loss limit, or about 23 dB. In some early line repeaters, the ALBO equalizer has a range of 6 to 31 dB. In many newer line repeaters, the ALBO has a range of 0 to 35 dB, meaning the signal can be attenuated through up to 35 dB of cable loss and still become regenerated properly. In practice, repeater sections are designed with a safety factor of several dB, so spacing of 28 to 32 dB is quite common. The minimum section loss is frequently set at 9 dB because of repeater design and to attenuate reflections. In other words, loss is generally added via an XMT span pad or via a 7.5-dB equalizer in the office repeater for the end section. Equalizers help control NEXT by simulating cable attenuation characteristics, whereas the pad displays flat loss. 3.7.6
T1 Repeatered Lines
In the case of a CO-to-CO T1 repeatered line, there is normally one originating office repeater and one terminating office repeater. The convention is that the originating end powers the line with simplex readings of +V and −V, where the −V is typically 10 volts more in magnitude with respect to the +V. If both ends originate power, then there may be a simplex power loop strap set at one of the line repeaters out in the middle of the span. Note that the
108
Transmission Systems Design Handbook for Wireless Networks
voltage will appear at the span line interface even if that is open loop. Current will not flow into an open loop, however, so the voltage sensed across the 10-ohm current sensing resistor is the best indication of the validity of the simplex current. In one-cable operation, errors can be caused by NEXT between cable pairs in opposite directions of transmission. Physically separating the different groups of pairs as much as possible is preferred. Refer to Figure 3.14. Shortening the repeater spacing will reduce the level differential and the sensitivity to NEXT. One single T1 system in a cable has only itself with which to interfere (transmit signal with respect to receive signal). Twenty systems in a cable makes many more cases for NEXT. Not only does one single system have itself for interference, but 19 other systems are also producing NEXT. If a particular cable use is expected to grow over time to its maximum capacity, then the correct safety margin must be calculated into the design. As stated earlier, in the end section next to the CO, maximum loss is limited to 50% to 70% of the normal loss limit (32 dB), or about 23 dB. The 23-dB limit includes the loss in the tip cable and in the office wiring to the line-terminating shelf for office repeaters. If exchange service is mixed in the same cable with T1, then an extra safety factor of 8 dB should be allowed to tolerate impulse noise. This brings down the end section to a maximum of 15 dB. An end section of 15 dB represents approximately 3,700 feet of cable, depending on the exact type. If this cable distance is too short to be
System 1 West-East
Transmit West-East System 2 West-East System 2 East-West Transmit East-West System 1 East-West Twisted copper pairs D shield Cable sheath
Figure 3.14 Typical cable layout.
Transmission-Network Principles
109
economically tolerable, then consideration should be given to leaving out the exchange service pairs to gain back the 8-dB safety margin. Office wiring typically goes from the cable vault directly to the office repeater shelf, from the office repeater shelf to the automatic protection switch (if equipped), and from there to the channel bank, radio MUX, or other terminal equipment. Along the way, there may be one or more DSX-1 cross-connect panels. Western Electric ABAM cable is a traditional selection to connect all of these elements. ABAM is a designation for 22-gauge, 110-ohm, insulated, twisted-pair cable. Use a loss figure of approximately 0.4 dB per 100 feet of office cabling, assuming 22 American wire gauge (AWG), remembering that 14 to 16 pF/ft is the normal cable capacitance. Cable with higher capacitance will give problems on the longer cable runs. CAT 5 LAN cable was not intended for T1, but it has low capacitance, so it makes a good substitute. In some rural service areas, lightning strikes aerial cable so frequently that one extra measure is applied at the cable vault. The outside cable might be 22 gauge and the tip wiring to the line terminating shelf might be 22 gauge, but one short section of higher gauge (smaller diameter) cable is added at the cable vault. This 2550 ft length of 24- or 26-gauge cable is called fuse cable. As its name implies, it acts as a fuse element that will open when huge currents from lightning appear. This technique is effective in keeping lightning surges out of the CO, but it has side effects. The 24- or 26-gauge insert may be short, but it interjects one extra attenuation factor in cable loss estimation. Furthermore, when the fuse cable opens up, that pair must be abandoned. If route junctions are present along the cable, then it is advisable to adjust repeater section spacings to place repeater housings at the junctions. Otherwise, there is greater possibility of signal-level differences from adjoining cable branches, which would contribute to cross talk. If a repeaterless junction is necessary, T1 span pads may be used to equalize levels between branches. Manufacturers of repeater housings typically offer several configurations depending on such factors as the number of line repeater units in the nest, whether there are attached stub cables, the length of such stubs, and the presence or absence of lightning protectors. It often seems as though each repeater-housing vendor has chosen to describe T1 pairs by different methods. It is important to be aware that in a single direction, the transmit pair from one repeater becomes the receive pair 6,000 ft away. The east-west transmit is separate from the west-east transmit.
110
3.7.7
Transmission Systems Design Handbook for Wireless Networks
Order Wire
The order wire is simply a loaded voice frequency pair that runs through the T1 cable and appears inside each line repeater housing. When one technician is troubleshooting at the remote housing and one technician is at the CO, it is very convenient to have a place to plug in a technicians telephone (butt-in phone). The order wire is typically terminated to an automatic ringdown line at the office repeater area of the CO. If the copper order wire is not implemented (as it is usually the case with the new wireless network) then the technician at the remote cell site must have a two-way radio or cellular telephone with plenty of battery and talk time. 3.7.8
Lightning
Lightning is an electrical discharge pulse in the atmosphere, which averages 20 kA or more in current. Commonly, these discharges are viewed as a flash from cloud to ground, although it can also work in reverse. Once ionization of the atmosphere occurs, this becomes luminescent, conductive plasma (a lightning bolt) sometimes reaching 60,000°F. Lightning can deliver a tremendous discharge of energy at any grounded object. Lightning strikes are somewhat predictable over a geographical region. The isokeraunic level is the number of thunderstorm days per year. This isokeraunic number varies from over 100 along the Gulf Coast of Florida to less than 5 in the Pacific Northwest. Nevertheless, virtually all areas in North America are subject to lightning strikes to some degree. Exterior equipment, rooftop equipment, and equipment connected to aerial and buried copper cables are subject to possible damage. In low-lightning areas, protected-type line repeaters are used along aerial cable and unprotected-type line repeaters are used along buried cable. In contrast, in areas of heavy lightning activity, it is quite common to use protected-type line repeaters regardless of whether the cable is aerial or buried. Aerial cables are especially susceptible to lightning strikes. The huge energy pulse from lightning momentarily raises the potential of ground at that strike point and then travels along the copper pairs to the CO where it finds a lower ground potential. At the CO, primary protection consists of three-element gas tube protectors, typically installed at the well-grounded protection frame or near the cable vault. In some COs, gas tubes are installed at the top of the relay rack with office repeaters. Primary protectors must be present and grounded properly. Many items of transmission equipment, such as office repeaters, have secondary protection in the form of solid-state surge limiters, but they are not effective if primary
Transmission-Network Principles
111
protection is bad. When the lightning voltage causes the primary gas tube to conduct, it represents a short circuit to the span-powering regulator. In a typical T1 transmission line, span power is fed only from the CO. In some cases, additional power comes from the far end unit. During normal operation, the 60-mA simplex current flows normally, the gas tubes sit idly, waiting for a lightning strike, and the DS1 traffic moves along. In the instant of a lightning strike (somewhere mid-span), the lightning acts as a huge current pulse that raises the ground potential at that strike location. If the cable is not properly grounded everywhere, the lightning can enter the copper pairs and flow toward the best ground point, which might be toward the nearest CO or, in the other direction, toward the remote end. The lightning might be in the form of a metallic voltage appearing from tip to ring on a pair. It might be in the form of a metallic voltage appearing from tip to ground or from ring to ground or it might be in the form of a longitudinal current surge. If a metallic voltage appears at a three-element gas tube protector (called the primary protector), the tube will fire either tip to ground, ring to ground, or tip to ring. This assumes that the gas tube is both working and grounded correctly. Some companies, however, fail to periodically test their gas tubes with a gas tube checker. If measured with a simple meter, gas tubes appear to be an open circuit whether they are working or not working, making it unreliable. If the gas tube has an improper voltage rating, it will not work correctly. If the voltage rating is too high, lightning voltage can seep in before it fires (thereby stressing equipment). If the voltage rating is too low, the normal dc voltage applied, at the simplex power feed end of the span, is enough to set it off prematurely, or at least to hold the gas tube in glow mode after the strike (thereby forcing a failure situation after the lightning strike). When the gas tube fires on schedule, it is effectively producing a short to ground. If there is some span power feed repeater nearby, this acts as a dead short on its current loop, which causes a big current surge. Various T1 products have built-in secondary surge protectors to withstand this secondary surge, called the current surge. But if the primary surge protector fails, it will most likely burn out the equipment or anything around it. In some cases, this surge protector is in the form of extra series resistance to limit current peaks. Otherwise, this protector is in the form of a fuse that will open up at a high current point. In yet other cases, the protector is a combination resistor and fuse. As a general rule, however, two-element protectors are not recommended for T1 circuits. Two-element protectors are suitable for POTS on a two-wire circuit. Due to the nature of T1 and its simplex current
112
Transmission Systems Design Handbook for Wireless Networks
loop, equipment may be damaged through the use of anything other than three-element gas tubes. Some telephone operating companies have the policy that T1 is only placed on new cables, dedicated for T1. This is because the headaches of rehabilitating old exchange cable can be severe. The job of eliminating every last bridge tap is difficult on some older cables. Impulse noise from ordinary analog subscriber loops can become a problem when mixed side-by-side with T1. 3.7.9
Simplex Power Design
Office repeaters are located at the CO where noise-free 48-VDC power is available. Line repeaters seldom have any local source of ac or dc power. Instead, the power for line repeater electronics is fed down the copper pairs from the office repeater. Inside a line repeater, there are two digital regenerators designated side 1 and side 2. Line repeaters are powered by dc current flow through a loop formed from the simplexes of the two cable pairs associated with side 1 and side 2. For proper operation, the line repeater must have current flowing through it in the right polarity. In this case, the repeater represents an equivalent resistance of 100 to 120 ohms on a common 60-mA simplex loop. Note that in most cases the loop must be completed at the office repeaters for this to be a valid loop, and office repeaters must have some type of switch or jumper specifically for this purpose. In a few cases, however, the simplex loop is not made this way and the power is fed all the way into some CPE terminal equipment where the loop is made there. However, these cases are not common. After this simplex power loop is engineered and installed correctly, a voltage drop can be measured across side 1 of a line repeater (about 7 VDC). Similarly measure from the span-receive side of the office repeater to the span-transmit side (about 712 VDC). In many far-end office repeaters, T1 channel service units, smart jacks and network interface units (NIUs) that receive power from the span line, this simplex arrangement presents a voltage drop of 11 to 12 VDC. This is one method of verifying that the current has the correct polarity. If it is wrong, then the voltage measurement is only 0 to 1 VDC. The positive current flows in the same direction as the PCM signal direction. In the classic T1 repeatered line, the simplex power planning must account for the equivalent dc resistances of office repeaters, line repeaters, attenuation pads and equalizers, and the copper conductors themselves (Figure 3.15). The longer the span line, the more repeaters must be in series; hence, more voltage must be applied at one end to feed the current loop. In
Transmission-Network Principles Network interface unit (NIU)
Office repeater
C.O. switch
Current + regulator −
Frm Gnd 48VDC Gnd
113
Optional CPE power
Simplex x loop selected
CPE
Frm Gnd Gnd 48VDC Optional
Figure 3.15 Shortspan power feed.
a short-length repeatered line, this voltage might be only 20 to 30 VDC, but as the length is stretched out to 10 to 15 mi, this might become 130 VDC. As it gets extremely long, simplex current might be fed from both ends, with the simplex looping back both ways in the middle. If these dc calculations are made and there are no further engineering guidelines, it is assumed that 22-gauge cable has an equivalent dc resistance of about 18 ohms per 1,000 ft. Of course, 24- and 26-gauge cable have much higher resistance. Assume that each line repeater is 120 ohms and one unpowered office repeater is 170 ohms. There are small additional resistances for LBO networks and other pads in the circuit, and these must be added into the calculation. If you know this total equivalent resistance in the loop and you know the current is 60 mA, then applying Ohms law will result in the minimum necessary voltage. Note that many of the most modern automatic span powering repeaters simply need local 48 VDC and they will develop the necessary voltage to regulate 60 mA into the simplex loop up to a maximum loop of 4,000 to 4,200 ohms. However, in many high-rise building installations, the T1 span line is rather short, perhaps from the basement equipment room to the tenth floor, and therefore no line repeaters are needed. Calling it the repeaterless T1 span line, we might see an office repeater at the near end and a smart jack or NIU at the far end, only 3,000 cable feet away. In this short-span-line case, the
114
Transmission Systems Design Handbook for Wireless Networks
voltage necessary to drive the simplex current loop is only 20 VDC, but the current loop must still be connected at the ends for the current flow to be correct. Transmission people will often refer to simplex power, loop current, and simplex voltage, using these terms interchangeably to mean about the same thing. Obviously, the voltage applied to a loop resistance yields a loop current. It is important to keep in mind that equivalent resistance through cable is based on one twisted pair acting as parallel resistance (resistance of one wire divided by 2). However, the simplex loop current must pass out the cable length and then pass back the same length (× 2). This effectively makes equivalent resistance the same as one wire for the one-way length. 3.7.10 Data Error Rates
Pulses sent along the repeatered line are regenerated at each repeater point. The repeater looks at each timeslot and decides whether or not a pulse is present. If the repeater logic determines that there is a pulse, the repeater puts out a new pulse free of noise, distortion, or interference incurred in the preceding repeater section. Owing to degradation factors, a small number of pulses may be incorrectly regenerated; that is, a pulse will be transmitted where it was not present or vice versa. The BER is the ratio of transmitted time periods (pulse or no pulse) that are received incorrectly at the end point to the total number of time periods. The total error rate is the arithmetic sum of the error rates of the individual repeater sections. Because of the effect of impulse noise from CO equipment, end sections (the section nearest the CO) are the principal source of errors and, therefore, are shortened to increase the signal-to-noise ratio at the office repeater. Between the terminal ends, a maximum error rate of 1 in 106 (BER = 10−6) will result in good voice communications. Pulse errors cause transients in the individual voice channels, but at this rate they are not noticeable to the average listener. On the other hand, while a higher error rate of 1 in 105 (BER = 10−5) will result in acceptable voice transmission, audible clicks are noticeable. Many data systems are less tolerant of bit errors, and facilities are engineered to meet error rates of 1 in 108, 109, or 1010 wherever possible. Any bit error on the span line results in a BPV, which can show up in a T1 span line. A BPV occurs as a result of bad wiring connections anywhere. A normal DS1 signal uses alternate mark inversion as the line code. In other words, if the terminal data stream is 11111111, then the line code is transmitted as marks (pulses) of alternating polarity, so we would see +1, −1, +1, −1, +1, −1, +1, −1. If a BPV has been created, we would see two consecutive pulses of the same polarity. In this case, we might see +1, −1, +1, +1, −1,
Transmission-Network Principles
115
+1, −1, +1. B8ZS is a different line code that intentionally sends and receives BPVs in a specific pattern to carry a special meaning related to 64-Kbps clear channel unrestricted data. Most basic transmission line elements such as line repeaters are completely transparent to any type of framing (SF, ESF, or otherwise). In contrast, most pieces of DS1 terminal equipment, such as channel banks, are very sensitive to proper framing format. Between these two types of elements exist intermediate elements, such as higher-order multiplexers and automatic protection switches. These elements may or may not be sensitive to framing format and some even convert from one framing format to another. A few also have the ability to autoconfigure depending on the signal format that is first received. 3.7.11 Voltage and Temperature Factors
Normal simplex loop current is 60 mA for modern line repeaters. Some older repeaters use more current (100, 120, or 150 mA). Occasionally, problems with ac power induction can be overcome by increasing the loop current within the tolerance of the repeaters. However, in the absence of 60-Hz induction, most simplex loops are set between 55 mA and 65 mA. Copper cable tends to have more resistance at higher ambient temperatures (it will take more voltage to drive the constant current through the loop), so knowing the expected cable temperature extremes for a locality helps set a strategy for optimizing loop current for best performance. Aerial cable is exposed to much higher temperatures than buried cable, and this must be calculated into the design. The interesting situation is presented when there are no line repeaters. If the total length of the facility is less than 3,000 to 4,000 ft, then probably no mid-span T1 line repeaters are necessary. In that case, the dc polarity might accidentally be applied backward at the CO and there is nothing on the copper line to fail until it gets to the NIU. So, if the NIU does not work on day one, it must first be established that dc is going correctly (this takes a dc voltmeter across the span side, and tests for a 68-VDC voltage drop), then move to tracing the high-frequency signal. A DS1 signal at 1.544 MHz is pretty unique and can be traced from point to point with the right kind of full-featured DS1 test set, not a receive-only DS1 monitor. It is important to remember that simplex current is only seen from the office repeater or NIU to the outside. Once DS1 signals are inside the CO, inside from the office repeater, then simplex is no longer present. Inside, the DS1 signals are +3 volt and −3 volt pulses with no simplex.
116
Transmission Systems Design Handbook for Wireless Networks
3.7.12 T1 Engineering, Installation, and Documentation
Once the installation has been completed, the installation crew must supply as-built drawings for files. These may be the original untouched engineering drawings or they may contain notations from the crew for changes made during installation. Examples include setting option switches differently to achieve better performance, using different cable pairs from the plan, and installing plug-in cards into specific slots. In some operating companies, this is done with paper drawings, while in others it is handled exclusively with electronic files. Every installation team and maintenance technician must have a good T1 bit error rate test (BERT) instrument. In many cases, a single BERT is adequate, but in others having one BERT at each end of a facility is an advantage. The best BERTs feature is to have two receive ports instead of one. The second port becomes handy for troubleshooting synchronization problems. In most cases of troubleshooting, all that is needed is a good BERT. A really complex case of finger pointing over the exact DS1 waveform can only be solved by using a good high-frequency dual-trace oscilloscope. With an oscilloscope, use two calibrated high-impedance probes set for A-B, or differential mode. Tie the two probe grounds together at the logic ground of the device under test. That way it can be checked if there is a dc offset voltage present. The new generation of BERTS can also show the DS1 waveform on the small LCD screen and perform detailed analysis based on applicable standards. The transmission acceptance test procedure (ATP) on T1 circuits will be described in later chapters of this book. 3.7.13 Troubleshooting and Problem Classification
Within the telephone industry there exists a standard terminology problem. In some places, the abbreviation T&R stands for tip and ring, which identifies one wire from the other wire in a two-wire pair; elsewhere T&R stands for transmit and receive. Frequently, these are not single wires. In DS1, these are a transmit pair and a receive pair. Of the transmit pair, there is a tip wire and a ring wire. Sometimes on a schematic they are named TT and TR (for transmit tip and transmit ring) and then RT and RR (for receive tip and receive ring). On still other equipment these are named T and R on the transmit pair and T1 and R1 on the receive pair. Again, recall that the transmit signal at the near end becomes the receive signal at the far end. In some places transmit is abbreviated as XMT or Tx, and receive is abbreviated as RCV or Rx. Unfortunately, there is no consistency.
Transmission-Network Principles
117
Another such set of terms is east and west. In the early days of longdistance circuits, AT&T Long Lines developed standard terminology to explain which long-haul equipment at one town was connected to other equipment located in the next town. All transmission directions were arbitrarily called either east-west or west-east. Therefore, it is quite common to be tracking a signal from one office repeater shelf marked east, meaning this is facing the span to the east. It might then go to another office repeater shelf marked west, meaning this is facing the span to the west. It is often more foolproof to simply draw lots of arrows on a circuit diagram, indicating the direction of all signals. Backbone microwave systems connecting cities use the same nomenclature. Exact symptoms are extremely important, since an intermittent problem is much harder to fix than a constant one. If the problem is intermittent, then its pattern of occurrence must be made reproducible. In other words, if it fails once per day every day, then that is significant. If it fails only on the hottest day of the year, something else is indicated. It is more difficult if it is an intermittent noise problem on a DS1 circuit, and in general, with an intermittent signal, it is important to keep testing it until it fails consistently. To a large extent, on DS1 signal it either works or it doesnt work. Rarely is there a marginal noisy problem except with grounding problems. If there is a clean BER with one legal DS1 test pattern, but a dirty BER with another legal pattern, often this is a clue to a bridge tap on the copper pair. As a general rule, BPVs are generated from two sources: a poorly performing T1 line repeater or bad copper T1 line pairs, or it is a problem within the last few feet of DS1 jumper wires, such as a single broken wire (not a broken pair) at a wire-wrap post. BPVs can also be caused by a shiner, a wire where the plastic insulation has been sliced off to reveal the shiny tinned copper conductor. Shiners can easily make an intermittent short to a grounded shelf or anything else. Thermal intermittent problems are hard to reproduce without a lab temperature chamber. It may help to have a can of freeze spray, but this is not usually a solution. Fortunately, with modern low-power components, thermals are not now common. An easy test is to tap an intermittent shelf with the handle of a screwdriver to see if some intermittent component will produce a bit error. Many systems and products include complete documentation, covering acceptance testing, while individual circuit packs may have their own segmented card-level tests. Once everything is hooked up end to end, a real-world test is necessary to verify that the entire system is fully ready to support live traffic.
118
Transmission Systems Design Handbook for Wireless Networks
3.7.14 Switch Options, Line Codes, and Framing
With DS1 signals inside the CO, it is necessary to verify where the framing format is SF (also erroneously called D4) or ESF. In some cases, it may not matter (they automatically detect and set SF/ESF), but it is helpful to know (line repeaters and office repeaters generally dont care). It is important to verify whether the line coding is AMI or B8ZS. 3.7.15 Fractional T1
In the early 1990s, fractional T1 (F-T1) became very popular and many operators began offering this service. In this way, customers could lease a portion of the T1s 1.544-Mbps bandwidth at a fraction of the cost of a full T1. Compared with F-T1 service, full T1 links were still more cost-effective if the bandwidth was almost fully used or if the links were fairly short. For most applications today, F-T1 local loops are rarely used, as they are neither economical nor practical. 3.7.16 T1 (J1) in Japan
The basic format for T1 transmission facilities in Japan is similar to North American ANSI standard and called J1. However, the CMI line coding used in Japan is different from the one used in North America [14]. Rather than using 50% duty cycle (half width) mark pulses, continuous marks are transmitted, inverting the voltage after each mark. The spaces also have a transition, but in the middle of the pulse period. This ensures that there are always sufficient transitions to keep synchronization, regardless of whether the signal is all 0s or all 1s. Japanese digital hierarchy (J1) is shown in Table 3.5. Table 3.5 Japanese Digital Hierarchy
Designation
Bit Rate (Mbps)
Number of Voice Channels
DS1
1.544
24
DS2
6.132
96
DS3
32.064
480
DS4
97.728
1,440
DS5
400.352
5,760
Transmission-Network Principles
119
3.8 Complete E1 Tutorial 3.8.1
Introduction to E1 Networks
The E1 (CEPT) digital transmission system is the lowest level of the European digital communication hierarchy. An E1 facility provides full-duplex transmission at 2.048 Mbps. Bandwidth is divided into 32 multiplexed (byte-interleaved) 64-Kbps channels. Depending on the framing format, one or two of the 64-Kbps channels are used for framing and other overhead functions. The remaining bandwidth (1.920 or 1.984 Mbps) can be used for either voice or data information. For digitized voice applications, the information bandwidth typically consists of 32 multiplexed 64-Kbps channels. For the transmission of data, the information bandwidth may be channelized as for voice, or it may carry from one to as many as several hundred multiplexed signals on an unchannelized basis. 3.8.2
Customer Premises Equipment
Various types of equipment (CPE) are employed at the customers location. Data terminal equipment (DTE) provides the source for the transmitted signal and the destination for the received signal. DTE includes such equipment as the following: • Multiplexers; • PBXs; • Channel banks; • Front-end processors; • Computers.
Other equipment interfaces the DTE to the public E1 network and provides various termination and interface functions, including the following: • Electrical interface; • Surge and lightning protection; • Signal regeneration and pulse-density assurance; • Keep-alive and yellow signal; • Loopback to the line controlled from the network.
120
Transmission Systems Design Handbook for Wireless Networks
Specifications for E1 interfaces are defined in International Telephone and Telegraph Consultative Committee (CCITT, an international standards group) recommendations. For instance, E1 signal levels are defined in CCITT Recommendation G.703. Note that CCITT is now called ITU, but the recommendation designations have not changed. 3.8.3
Signal Characteristics
The E1 signals are bipolar pulse trains and the data is assembled using a TDM scheme. When voice is transmitted over E1 networks, it is digitized, using PCM. An E1 signal is divided into timeslots of 488 nanoseconds each, or 2,048,000 timeslots per second. The presence or absence of a pulse in each timeslot encodes data (or digitized voice). A pulse, if it exists, will have onehalf the duration of the timeslot and amplitude of three volts. When a pulse is present, the timeslot data is a 1; when no pulse is present, the timeslot contains a 0. E1 signals employ alternate mark inversion (AMI) line coding, in which consecutive pulses are expected to be of opposite polarity. Consecutive pulses of the same polarity, called a bipolar violation or code violation, indicate a transmission error, except in special density-preserving codes described under HDB3. An example of AMI line coding is illustrated in Figure 3.16. 3.8.4
Transmission Facilities
E1 signals are transmitted primarily over standard twisted-pair copper wire. Signal loss on the wire is approximately 5 or 6 dB per 1,000 feet. Repeaters are typically employed every 6,000 feet along the transmission facility to compensate for the signal losses and to ensure an adequate signal level at the network interface (the termination of the E1 line at the customer premises). Figure 3.17 illustrates a typical E1 network interface. E1 signals may also be transmitted via satellites, digital microwave radios, fiber-optic systems, and coaxial cable modems. In the carrier networks, signals may be multiplexed into even higher-speed signals. Bits
0
0
Signal
Figure 3.16 AMI line coding.
0
1
1
0
1
1
Transmission-Network Principles Customer premises equipment
DTE
DCE
121
E1 line
3000 ft. Repeater
6000 ft. Repeater
Network interface (NI)
Figure 3.17 One end of a typical E1 link.
3.8.5
Pulse Density
In order to interpret and regenerate an E1 signal, repeaters and other network equipment must be able to determine timeslots based on the pulses in the received signal. Since pulses occur only when 1s are transmitted, signals with too many consecutive 0s can cause timing problems. There is no specific pulse-density requirement for the E1 environment. Ones density is automatically maintained by high-density binary 3 (HDB3) coding, as described in the next section. Because there are different national standards for connecting CPE equipment to public E1 networks, often there is a need for two different line interfaces. One is designed for direct connection to an E1 line; the other is designed for connection to an E1 line through an approved networktermination unit. 3.8.6
HDB3
HDB3, which is used in E1 networks, replaces strings of four 0s with special sequence (i.e., fixed codes containing intentional bipolar violations). In other words, HDB3 is a coding method that does not allow more than three consecutive 0s. This code is recommended by the ITU-T G.703 for the 2-, 8-, and 34-Mbps systems [14]. Equipment receiving data containing the special sequences automatically translates them back into strings of four 0s. The special sequences in HDB3 coding are 000V and 100V. In these sequences, V represents a bipolar violation. The choice to use 100V or 000V is made so that the pulses violating the bipolar rule take on alternate +1 and −1 levels. Sequence 100V is used when there have been an even number of 1s since the last special sequence; 000V is used when there have been an odd number of 1s since the last special sequence. Special sequences
122
Transmission Systems Design Handbook for Wireless Networks
follow each other if the string of 0s continues. The choice of the first special sequence is arbitrary. An example of HDB3 coding is illustrated in Figure 3.18. 3.8.7
E1 Framing Synchronization
In E1, data is grouped into frames of 256 bits. Each frame consists of 32 8-bit timeslots, and 8,000 frames are transmitted each second. This provides for the following E1 transmission rate: 8,000 × 256 = 2,048,000, i.e., 2.048 Mbps 8,000 × 8 = 64,000 (i.e., 64 Kbps) and 32 × 64,000 = 2.048 Mbps A 64-Kbps timeslot has the same data rate as a 64-Kbps DS0 in the T1 North American digital hierarchy. Both evolved from the restrictions imposed by digitized voice. Framing information is carried in timeslot 0 (TS0), while signaling information, if used, is carried in timeslot 16 (TS16). The remaining 30 timeslots carry user information. A group of 16 frames constitutes a multiframe. Figure 3.19 illustrates the basic E1 frame structure. 3.8.8
E1 Framing Formats
E1 defines two primary framing formats: • TS0 framing formats; • TS16 framing formats. 4 consecutive zeros Input bit stream
1 0 1 1 0 0 0 0 0 1 0 0 0 0 0 0 0 0
HDB3-coded bit stream
1 0 1 1 1 0 0 V 0 1 0 0 0 V 1 0 0 V
HDB3-coded levels
− 0 + − + 0 0 + 0 − 0 0 0 − + 0 0 +
Figure 3.18 HDB3 line coding.
Transmission-Network Principles 1
2
TS0
FR 0
3
TS1
4
5
6
8
8-bit timeslot 3.91 µs
TS16
TS2
FR 1
7
123
FR 2
TS31
FR 15
256-bit frame 125 µs
16-frame multiframe 2 ms
Figure 3.19 E1 framing structure
The TS16 multiframe format was developed to provide signaling information to a public switched E1 network. In a switched network, individual 64Kbps timeslots can be routed independently through the network. When TS16 is not used for signaling, that timeslot carries data. As the name implies, TS0 framing formats use timeslot 0 (TS0) to provide the framing pattern that allows equipment receiving the E1 signal to synchronize on the pattern and correctly interpret data. There are two basic TS0 framing formats: • TS0 framing format (without CRC-4); • TS0 multiframe format (with CRC-4).
In the TS0 framing format (without CRC-4), frame synchronization is maintained via the frame-alignment signal (FAS), a bit pattern (X0011011) present in TS0 (bits 2 through 8) in every other frame. Frames that contain the FAS pattern are designated as word frames; frames not containing the FAS pattern are designated as not-word frames. Frame synchronization does not require all 8 bits of TS0 in every frame. Therefore, the TS0 bits have been defined by CCITT for other uses. To use the TS0 bits effectively, CCITT defined a multiframe format. A multiframe consists of 16 consecutive frames numbered 0 to 15. While the FAS distinguishes word frames from not-word frames, the multiframe is found by looking for the 001011 pattern, known as the multiframe-alignment signal (MAS), in bit position one of the not-word frames 1, 3, 5, 7, 9, and 11. This pattern is interleaved with the CRC-4 bits and the E bits.
124
Transmission Systems Design Handbook for Wireless Networks
As a further refinement, each multiframe is divided into submultiframes, which consist of 8 consecutive frames: frames 0 to 7 and frames 8 to 15. The submultiframe construction, sometimes referred to as the CRC multiframe, is used when CRC-4 error checking is employed. It is the network (or receiving) equipments job to synchronize on the FAS and MAS patterns to identify frames and multiframes properly. An out-of-frame (OOF) event is declared whenever there are three consecutive framing errors. As described in CCITT Recommendation G.732, a framing error is defined as an incorrect bit in one of the seven framing bits in theTS0 word, or an error in bit 2 of TS0 not-word. Three consecutive words or notwords containing errors results in OOF. When CRC-4 is not used, bit 1 of TS0 carries the international spare bits instead. Without CRC-4 coding, only two types of frames can be identifiedframes containing the FAS and frames not containing the FAS. Since the TS0 Multiframe formats are defined in CCITT Recommendations G.704 and G.706, the E1 NSM describes its framing options as G.704 and G.706. G.704 refers to a framing format without CRC-4 capabilities, and G.706 refers to a multiframe format with CRC-4 capabilities. 3.8.9
Spare Bits
The CCITT has set aside a number of spare bits in TS0 that can be used to transmit additional information between end points. These bits include the following: • International spare bits (Si); • Remote-frame alarm bits (Yf ); • National bits (Sa4Sa8).
The following sections provide a brief description of the location and application of these spare bits. 3.8.9.1 International Spare Bits
The international spare bits (Si) are transmitted in bit 1 of TS0. The last two Si bits in each CRC multiframe have been redefined per CCITT Recommendation G.704. When CRC-4 is enabled, these two bits represent CRC-4 error-indication bits (E bits). These bits are transmitted and received in two consecutive frames that do not contain the FAS pattern. In the receive
Transmission-Network Principles
125
direction, a 0 in the E-bit location indicates a far- (or remote-) end block error (FEBE or REBE). In the transmit direction, E-bit locations identify received errored submultiframes by setting the binary state of one E bit from 1 to 0 for each errored submultiframe. The delay between the detection of an errored submultiframe and the setting of the E-bit indicating the error state is less than 1 sec. If CRC-4 is not enabled, the Si bits are differentiated as either odd or even, depending on their frame. The odd Si bits can be used as one 4-Kbps data link, and the even Si bits can be used as a different 4-Kbps data link. 3.8.9.2 Remote-Frame Alarm Bits
Bit 3 of each not-word frame has been defined as a remote-frame alarm bit (also known as a remote or distant FAS). This bit is set to 1 to indicate loss of frame alignment with the received signal. 3.8.9.3 National Bits and Performance Monitoring
The strategic importance of communications continues to grow at a rapid pace. In todays marketplace, corporations look to their high-speed digital networks to give them a competitive advantage. With so much depending on their networks, managers are becoming more and more quality conscious, and when it comes time to implement a new E1 circuit, the quality of the service is just as important as the cost. Because quality is of such concern, users need better ways to track the performance of E1 circuits offered by a service provider. A corporation that manages its services closely is more likely to get the most out of its network and is also likely to get a quicker and more congenial response from a carrier when problems do arise. A user who relies on the carrier to monitor circuit quality is literally putting the business in the carriers hands. In the United States, AT&T originally tariffed T1 in a format known as D4, which did not provide for in-service monitoring of the T1 circuits. As the requirement to meet contracted performance levels for T1 became a more significant factor, AT&T developed the ESF to incorporate performancemonitoring-management functions in the overhead bandwidth (used solely for frame synchronization in the D4 format). In ESF, 2 Kbps of the overhead is dedicated to logic-error monitoring, and 4 Kbps is reserved as a two-way communication channel. This maintenance channel is called, as discussed in previous chapters, the facility data link, or FDL. For companies using E1 circuits, CCITT has defined certain national spare bits in the frame-alignment timeslot (TS0). The performance monitoring employs one of these bits to create a
126
Transmission Systems Design Handbook for Wireless Networks
4-Kbps data link (the Embedded Operations Channel) which is analogous to the ESF FDL. Five national bits (Sa4Sa8) are transmitted and received in bit positions 4 through 8 of TS0 of the FAS not-word. Each national bit can be a 4-Kbps data link to the far end. These data links are sometimes used to transmit system control and status information. The E1 NSM uses one of the national spare bits (Sa4Sa8) to transmit performance data and to communicate with the far-end unit. This 4-Kbps data link serves the same purpose as the FDL in T1. The E1 NSM allows the user to determine which national bit position will serve as the data link. Traditionally, customers have relied on their multiplexers for E1 performance monitoring. The typical performance-monitoring method used by multiplexers involves dedicating a channel (864 Kbps) to network management. This management channel is used for various transmissions, including configuration parameters, routing information, and bandwidth-use statistics. In most cases, error-rate monitoring is performed only on this management channel and is extrapolated to give an approximate error rate for the entire E1. Assuming a full 64-Kbps management channel, error-rate monitoring is performed only on about 3% of the bandwidth, leaving 97% essentially unmonitored. With an 8-Kbps maintenance channel, error monitoring occurs on less than 0.4% of the bandwidth. For users interested in maintaining or improving service quality to make sure they get maximum efficiency from their networks, the installation of performance-monitoring units (PMUs) is recommended and offers numerous advantages: • The PMU monitors the full E1 bandwidth full time, without dis-
rupting service or decreasing usable bandwidth. • The same performance data can be accessed by both carriers and customers. • The location of PMUs in the network is ideal for gathering and reporting performance data. Because PMUs are installed at the customer premises, at the interface between the customer and carrier portions of an E1 circuit, they provide a critical point of demarcation. Looking in one direction, the PMU monitors carrier performance from the users point of view. Looking in the other direction, the PMU monitors the customers input into the carriers network. Because the entire bandwidth is monitored for a range of format and logic
Transmission-Network Principles
127
errors, virtually all performance failures and degradations are detected and isolated to the customers or carriers portion of the network. And because both the customer and carrier can read the same performance data, the nature of the problem is equally clear to both parties. 3.8.10 Global Framing Formats
As a general rule, outside of the United States and Canada (T1), most countries use E1 as a primary rate service. Japan has a variation of the T1, and Germany and France impose a slight variation of the E1 and, thus, have unique formats different from the rest of the market. Only one element remains constant and universal, and that is DS0. However, the form of PCM encoding differs between T1 (mu-law companding) and E1 (A-law companding). The term companding is a contraction of the words compressing and expanding. Companding is the process of compressing the amplitude ranges of a singal for economical transmission and then expanding it back to its original form at the receiving end. The ITU-T companding standard is based on A-law used in the conversion between analog and digital signals in PCM systms. Mu-law is used in Japan and North America. Therefore, sometimes required E1/T1 conversion involves both the compression law and the signaling format.
References [1]
Clark, M. P., Networks and Telecommunications: Design and Operations, Second Edition, New York: Wiley, 1997.
[2]
Gunn, H. J., Principles of Traffic and Network Designs, Geneva, IL; abc TeleTraining, Inc., 1986.
[3]
Bellcore, CCS/SS7 Computer Based Training, Reference Guide, NJ, 1993.
[4]
G. van Bosse, J., Signaling in Telecommunication Networks, New York: Wiley, 1998.
[5]
Winch, R. G., Telecommunications Transmission Systems, New York: McGraw Hill, 1993.
[6]
ITU-T Recommendation G. 702, Digital Hierarchy Bit Rates.
[7]
Auki, P., et al., RATES: A Server for MPLS Traffic Engineering, IEEE Network, March/April 2000
[8]
Onvural, R. O., Asynchronous Transfer Mode Networks: Performance Issues, Norwood, MA: Artech House, 1994.
128
Transmission Systems Design Handbook for Wireless Networks
[9] Doverspike, R., and I. Saniee, eds., Heuristic Approaches for Telecommunications Network Management, Planning and Expansion, Norwell, MA: Kluwer Academic, 2000. [10] Black, U., QoS in Wide Area Networks, Upper Saddle River, NJ: Prentice Hall, 2000. [11] ITU-T I.357, B-ISDN Semipermanent Connection Availability, November 2000. [12] Aukia, P., et al, RATES: A Server for MPLS Traffic Engineering, IEEE Network, March/April 2000. [13] Larus Corporation, Transmission Engineering Tutorials: T1 Repeatered Lines, 1996. [14] Flanagan, W. A., The Guide to T1 Networking, Fourth Edition, New York: Telecom Library, 1990.
4 Wireless-Network Architecture 4.1 2G Wireless-Network Architecture A simplified diagram of todays 2G wireless-network architecture is shown in Figure 4.1. Shown here are three RBSs, the BSC, and the MSC and connection with the PSTN. In addition to that, there are a number of other nodes not shown here, like the voice mail system (VMS), SMS, HLR/VLR, and so on. Regardless of whether the access transmission network (connection between BSC and RBS or backhaul) is leased or owned by the operator, RBS can be connected to the BSC via microwave links, fiber-optic, or wireline (usually copper) systems. The BSC provides the connectivity between the MSC/PSTN and the radio network. It performs the radio call management functions and radio network management functions and its capacity is usually given in Erlangs. BSC blocking probability is defined to be the probability that a new request for service is rejected at the BSC due to lack of resources. Resources are understood to be card processing, transmission link capacity, countable resources such as channel cards, and so on. BSC blocking probability does not include blocking due to limitations of the air interface or call admission control, and the BSC blocking probability is usually specified to be 0.5%. BSC and MSC are usually colocated and, therefore, shown as one node. It is important to note that the physical layer of the transmission network, copper, microwave, or fiber optic will not change from 2G to 3G wireless networks. The only change will be more stringent requirements on its 129
130
Transmission Systems Design Handbook for Wireless Networks
MW
RBS 3
MW radio MW radio T1/E1 leased lines
RBS 2
PSTN T1/E1 leased lines
BSC/MSC T1/E1 leased lines RBS 1
Figure 4.1 Example of 2G wireless network.
quality, availability, and reliability as well as higher capacities required to carry the traffic. There are today several vertically oriented, single-service networks capable of delivering similar services. We have wireline and wireless networks as well as pure data/IP networks as well as cable TV/CATV networks. These are basically separate networks that build on different principles and practices to ensure the reliability of a single service with different approaches to network management, guaranteed service levels, and so on. For over 100 years, classic telephony networks have been optimized to carry real-time voice traffic between fixed points in the network. The classic telephony network supports an integrated service concept, involving the following: • One service typevoice; • One subscription; • One user terminal.
This leads to a vertical industry orientation, where the operator offers everything from subscriber access to service creation and service delivery
Wireless-Network Architecture
131
across a wholly owned network infrastructure optimized for a particular service category. Each vertically integrated network incorporates its own protocols, nodes, and end-user equipment and terminals. This means that the telephony (voice) and data service domains are still more or less kept separate. The rapid convergence between telecommunications and datacommunications will lead to a convergence of these purpose-built networks into ATM/IP-based multiservice, or next-generation, networks that can provide reliable and real-time communications. This network convergence raises some fundamental issues of network characteristics and how to bridge the inherent value of reliable circuit-switching technology with more besteffort-oriented packet-switching technology.
4.2 3G Wireless-Network Architecture 4.2.1
Directions in 3G Developments
Multiservice switching is the foundation of next-generation networks. The need for wireline- and wireless-network infrastructure to support future data-communication and IP services is obvious. However, not only is the type of traffic changing, the quantity of traffic is also growing rapidly. Both of these new factors are of critical importance in the development of future networks. Another significant fact is that the major source of revenue for most operators around the world is still voice services. And it is likely to remain so for the coming yearswhile the main source of growth will be data-based services. This means that networks must be optimized to carry packet-oriented traffic at the same time that they are delivering a reliable and high-quality voice service; that is, the technologies are converging. So far the networks have been shaped by the concept of one network, one service, one subscription, and one user terminal. This changed with the introduction of the Internet. The future networks will be multiservice networks, able to carry a full range of services, from voice communications and simple file transfers to high-speed Internet and real-time, broadband multimedia services. These services will be accessible via different access networks and a number of different terminals. Using packet technology as the infrastructure for telephony services implies that packet networks must meet stringent requirements; QoS is a key issue. IP is today not a mature bearer technology for efficient large-scale telephony QoS solutions as regards delays and latency in voice streams. The performance available from todays IP environment cannot provide carrierclass telephony services, but that may change very soon.
132
Transmission Systems Design Handbook for Wireless Networks
The only technology available today that can meet those requirements is ATM. ATM seems to be the only appropriate technology for use in core multiservice networks, to carry both IP/data services and voice. ATM is able to act as a connectivity layer for both traditional voice and IP/data services. ATM was developed and standardized for both telephony and data, and the large incumbent telecom operators were involved in the standardization process. Therefore, most operators see ATM as the only safe migration path to new generation networks. In other words ATM has now found its role as a network infrastructure technology capable of carrying all the different transport needs of the future networks, including IP. The fundamental structure of the next generation multiservice network is based on a shared bearer network providing multiple services based on ATM. Figure 4.2 shows the layered backbone network model. The bearer network is accessed through so-called media gateways (MGWs) or hybrid switches. The MGWs control the traffic in and out of the bearer network and handle interconnections to other networks, different access networks or to traditional switches and routers. In this new network infrastructure, telephone calls will be converted to ATM switched virtual circuits while separate telephony servers control the setup of the calls through the network. The telephony servers thus provide the equivalent intelligence of todays telephony exchanges, but are not involved in the actual connections as such; the calls are set up end to end
MSC server
HLR
Wireless
MGW
Internet, intranets
MGW
ISDN/PSTN wireless
MGW
Connectivity backbone network ATM, IP Servers handling control Media gateway handling connection
Figure 4.2 Layered backbone network model.
Wireless-Network Architecture
133
across the ATM network using a switched virtual circuit between the edge devices. This separation of the switching and connectivity functions in the network is a key to the evolution path toward a single bearer network based on ATM. Telephony subscribers may be connected to the ATM bearer network via the MGW from traditional switches, standard access nodes, or if it is an enterprise customer, PBX. The telephony server does not only control phone calls, it also contains all the required telephony intelligence for all MGWs in the domain. The data access to the bearer network may range from LANs and ATM switched leased lines, to cellular networks and data transmitted from or through these. All access, voice as well as data, terminates in an MGW, which ensures that traffic gets onto the ATM bearer network. A single telephony server with several stand-alone MGWs (multiservice switches) would satisfy regulatory requirements on geographically spread points of presence. The operators would not have to deploy a traditional countrywide circuit-switched network, but would be able to simultaneously offer the full range of telephony and data services from day one. As network management is very expensive and complicated, another main benefit is the fact that only one network has to be managed instead of several different networks. Today there are major developments in virtually all areas of network access and infrastructure. In access, new digital subscriber line (DSL) and wireless technologies are opening up bandwidth and giving users more and more network capacity, instant access, and on-line services. In the core network (CN), new switching and transport technologies such as ATM and WDM are expanding the capacity and the flexibility of the networks. While the focus for initial wireless 3G deployment is voice services, it is expected that wireless 3G will also become a key Internet access technology. Although expected 3G data rates of up to 2 Mbps will never compete with wireline services, the trigger that will fuel exceptional 3G growth is the promise of true mobility, Internet-connected data, and information services anywhere, any time, from a variety of handheld platforms. Global standards for 3G systems are still evolving, but high-level architectures and overall network characteristics are beginning to solidify, accelerating the need for new testing strategies and capabilities for a key component of 3G, the terrestrial radio access network (RAN). RAN infrastructure will support 3G functions including access, roaming, transparent connection to the PSTN and the Internet, and QoS management for data and Web connections. The concept of the 3G wireless network is shown in Figure 4.3. The high-QoS characteristics of classic telephony networks must now migrate onto horizontally oriented next-generation networks that can support
134
Transmission Systems Design Handbook for Wireless Networks
MSC server MGW
PSTN
MGW
Internet
Backbone network
MGW Radio network Radio base station
MGW
RN server RBS (IP/ATM)
Figure 4.3 Concept of 3G wireless network.
multiple services based on ATM and the IP protocol. Similar developments have been under way in the enterprise market and took off around 1995 and 1996 with the widespread adoption of the TCP/IP protocol for intranet technology. For public network operators and service providers, the heavy increase in data traffic is leading to a bottleneck in narrowband networkswireline or wireless or both. Operators have to offer circuit-switched and packet-switched services, and must expand even further to offer multiservice and multimedia networks (Figure 4.4). It is also essential for the operator to optimize the network resourcesfor example, to use one transmission network that will be suitable for all services. Operators need to ensure that their investments will address their transmission requirements well into the future. The future lies with packet-based transmission technologies. Almost all networks of the past and today are vertically integrated. Vertically integrated networks are single-service networks where the operator offers anything from subscriber access to service creation and delivery. Services are offered across a wholly owned network infrastructure, optimized for a particular service category. Each vertically integrated network incorporates its own protocols, nodes, and end-user equipment and terminals. They operate different principles to ensure the reliability of a single service. Vertically
Wireless-Network Architecture
Wireless voice
Differentiation + usage
Wireless Internet
135
Wireless + multimedia 3G Real-time Internet Mobile video
Convergence Mobile office Ubiquitous (mobile e-mail) Personal coverage Mobile WWW Good voice mobile Wireline quality Bluetooth Coverage multimedia Capacity Mobile image services Cost Mobile e-com VAS WLAN ~384 (512) wide area Tariffing/prepaid ~115/384 Kbps (2 Mbps) locally ~64 Kbps ~10 Kbps Time 02/03+ 01/02 99/00 Figure 4.4 Wireless evolution.
integrated networks have different approaches to network management, guaranteed service levels, etc. Vertical integration means that telephony and data service domains are kept more or less separate. The rapid convergence between telecommunications and datacommunications will lead to a convergence of these purpose-built networks into ATM/IP-based multiservice, or next-generation, networks. This network convergence raises some fundamental issues with respect to network characteristics and how to bridge the gap between the inherent value of reliable circuit-switching technology and best-effort packet-switching technology. In order to survive in a converged communications marketplace, operators need to draw on truly open systems that invite competition in horizontal layers. The QoS characteristics of classic telephony networks must now migrate to horizontally oriented next-generation networks that can support multiple services based on ATM and IP, since the 3G network will ultimately be an open IP platform supporting a wide range of new global services [1]. IP telephony (IPT) has quickly emerged as a serious alternative to traditional circuit-switched telephony. Not just because it offers a more cost-effective service, but because the underlying technology offers a wealth of new business opportunities. The evolution of IPT will not only depend on its integration into successful business operations, but
136
Transmission Systems Design Handbook for Wireless Networks
also on the evolution of the underlying technology. One of the most limiting barriers for this type of service offering is, without a doubt, lack of interoperability. If different vendor offerings and different carrier networks cannot interoperate, this will limit the possibility for end-user connectivity. A number of all-IP wireless-network architectures have been proposed and planned for 2004 and beyond and it is too early to say exactly how the evolution of the wireless network will go from there. In 3G networks and beyond, bandwidth flexibility is a key issue and involves a flexible decentralized provision of bandwidth to a single user as the need varies, but also cost-effective bandwidth provision to a large number of users with different bandwidth requirements in the same network. 4.2.2
Horizontally Layered Network Architecture
In the horizontally layered network architecture (Figure 4.5), functionality and nodes are arranged in layers according to their specific areas of use. The layered concept of the network architecture of the 3GPP specifications comprises three distinct layers: 1. Application layer; 2. Network control layer; 3. Connectivity layer. 4.2.2.1 The Application Layer
The application layer is where the end-user applications reside. In modern networks, applications are implemented in mobile terminals and in dedicated application servers in the network. The application servers are often complemented with content servers, which host service-related databases or libraries, such as video-clip libraries or news history databases. Concepts such as the virtual home environment (VHE) and open service architecture (OSA) were developed in the 3GPP to allow operators to provide unique services. Operators benefit from being able to differentiate themselves from one another by providing unique services, thus securing for themselves a higher position in the value chain. They also have the option of developing these services themselves or of obtaining them from third-party software houses and they can even get external service providers to run them. This flexibility allows the operator to choose from a huge portfolio of services that it can offer its subscribers. The application layer interfaces with the
Application servers
External servers
Content servers
Application layer Control servers
Network control layer Access network
GCP Connectivity layer Media gateways
ISDN, PSTN, Internet
137
Figure 4.5 Horizontally layered network architecture.
Wireless-Network Architecture
APIs
Control servers
138
Transmission Systems Design Handbook for Wireless Networks
network control layer via a defined set of open application program interfaces (API). By using open APIs, application developers can make use of the features of standardized service capabilities to design new services and applications. 4.2.2.2 The Network Control Layer
The network control layer incorporates all the functionality needed to provide seamless, high-quality services across different types of networks. The different networks can be seen as a set of domains, each of which houses control servers that are specific to a given network. Generally speaking, the network control layer houses several different kinds of network servers. The servers are responsible for controlling mobility management, the setup and release of calls and sessions requested by end users, circuit-mode supplementary services, security, and similar functions. These domains can be owned by various individual operators or by a single operator. 4.2.2.3 The Connectivity Layer
The connectivity layer is a pure transport mechanism that is capable of transporting any type of information via voice, data, and multimedia streams (Figure 4.6). Its backbone architecture incorporates core and edge equipment. The core equipment transports aggregated traffic streams between the different nodes at the edges of the backbone. As a rule, core equipment is a backbone router or backbone switch that handles traffic streams either according to very simple classification principles or to routes that the network operator has predefined by means of traffic engineering. Edge equipment collects customer-specific data and statistics for accounting and billing purposes and provides the single bit-pipes that guarantee an appropriate QoS. The edge equipment is usually an MGW, which operates under the full control of the nodes in the network control layer. In addition, an MGW allows the bit streams to be processed, thus providing coding and decoding of speech streams, canceling echo, bridging multiple party calls, and converting transport protocols. The nodes in the network control layer also control these manipulations. This exertion of control down to the bit-stream level allows the variety of services and applications implemented by the different network control domains to be achieved via a common connectivity layer. At the same time, the services and application are independent of the transport technology applied, which may be mixed or
GCP
ATM ATM
MGW
GCP
IP router ATM
ATM switches
GCP ATM Connectivity layer
MGW ATM
ATM
MGW
139
Figure 4.6 The connectivity layer.
ATM
ATM
MGW
Wireless-Network Architecture
GCP
140
Transmission Systems Design Handbook for Wireless Networks
vary over time as the network evolves. Connectivity-layer solutions can be based on either ATM transmission or IP transmission. It may also use a QoS-enabled IP-backbone network running IP over SDH, carrying packetand circuit-mode communications. 4.2.3
3G Core Network
The 3G CN supports circuit- and packet-switched services. It contains the hardware and the software needed to provide end users with multimedia applications. The CN spans both the control and connectivity logical layers. One of the new nodes required in the 3G wireless network is the MGW. The MGW performs functions such as speech coding and decoding, echo cancellation, conference-call bridging, tone and announcement generation, setup and release of user data bearers, and QoS IP routing and switching, including QoS handling and packet retunneling. The MGW will also contain interface functions for different transport standards, for example between an IP- or ATM-based CN and an external STM network. For volume-based charging support, the MGW keeps track of the volumes sent and received (for packet-based services), and it performs some security functions (for example, for packet mode services). Most MGW resources are shared between packet- and circuit-communication service, or can easily be reconfigured from one communication mode to the other. The MSC server handles control-layer functions related to circuitmode communication services at the WCDMA or CDMA2000 RAN and PSTN/ISDN borders, and performs MGW control, ISDN services control, mobility management, authentication, data collection and output, services switching function (SSF), Internet dial-in services (RAS), and element management. In addition to these functions, the MSC server houses the interworking and gateway functionality necessary to act as an SMS-IWMSC and SMS-GMSC for the Short Message Service. The HLR is a network database for mobile telecommunications in general. The HLR holds all mobile-specific subscriber data and contains a number of functions for managing this data, controlling services and enabling subscribers to access and receive their services when roaming within and outside their home PLMNs. The HLR communicates with GSNs, MSCs, and other network elements via the MAP protocol. The HLR is a real-time mobile telecommunications node of GSM, CDMA2000, and UMTS systems. The HLR is vital for the operation of a network, as it holds all subscriber data, and it is equally important for the call setup in the network, as well as for the control of the roaming subscribers.
Wireless-Network Architecture 4.2.4
141
Universal Mobile Telephone System
UMTS is the accepted 3G standard for GSM operators. UMTS requires paired 5-MHz RF channels, four times as wide as the paired 1.25-MHz channels required for CDMA2000. For this reason, UMTS is sometimes referred to as WCDMA. By migrating to UMTS, operators will gain access to additional spectrum as well as the greater capacity and expanded functionality of the new technology. UMTS incorporates a more efficient variable vocoder (CODEC). In common with CDMA2000 1x, this vocoder will increase the voice capacity of a given amount of spectrum. Outside of the Americas, UMTS is being deployed on the 1,900-MHz (uplink) and 2,100MHz (downlink) frequencies. Because of this, some operators, primarily those in the Americas who now use the 1,900-MHz frequencies for PCS, would be unable to migrate to UMTS. Allocation of other frequencies for UMTS may or may not be possible. The well-publicized failures of U.S. operators to acquire frequencies at 700 MHz (occupied by TV broadcasters), 1,700 MHz (occupied by the military), or 2,5002,600 MHz (occupied by educational broadcasters) provide examples. 4.2.5
3G in GSM Networks
New 3G GSM networks will require new radio and CN elements as well as a new air interface. This will require new BSSs, which will include radio network controller (RNC) and Node B. The RNC will include support for connection to legacy systems and provide efficient packet connection with the CN packet devices (SSGN or equivalent). The RNC performs radio network control functions that include call establishment and release, handoff, radio resource management, power control, diversity combining, and soft handoff (handoff ). A Node B is equivalent to a base station in the 2G network, but also incorporates support for the 3G air interfaces. New cell-planning methods will be needed to support the new frequency allocations for 3G and the radio interface changes. More 3G base stations will be needed than are necessary in a comparable 2G coverage area. This gives an advantage to GSM 1800 and 1900 network operators whose cells already cover a smaller coverage area than those for GSM 900 networks. GSM 900 network operators will need to fill in coverage in between existing cell sites. The 3G CN will be an evolution from GPRS or equivalent 2.5G CN systems. Upgrades to the mobile and transit switching systems to deliver packets will also be needed. A new piece of network infrastructure for 3G is also the MGW, which resides at the boundary between different networks to
142
Transmission Systems Design Handbook for Wireless Networks
process end-user data such as voice coding and decoding, convert protocols, and map quality of service. The connectivity layer also provides access to backbone switches and nonmobile networks such as cable television. In some vendor solutions, MGWs are controlled remotely by the MSC and GSN servers by means of the Gateway Control Protocol (GCP). The ITU is working to ensure that the GCP is an open standard protocol. Existing network operators can then upgrade their MSC and GSNs to implement 3G or alternatively to implement a new stand-alone MGW that is controlled from the server part of an upgraded 2G node. 4.2.6
3G CDMA Network
The ITU manages the 3G umbrella standard known as IMT-2000. This standard endorses five different modes of RF interface and three major types of terrestrial infrastructure known as RAN. The intention is for any of the RF modes to work with any of the RAN types. The two major types of RAN are UMTS WCDMA and IS-2000 (also known as CDMA2000). UMTS originated in Europe and WCDMA in Japan, but these are now almost identical, having been converged into a single specification under the 3GPP. IS-2000 is predominantly North American and is defined by the 3GPP2 organization. More recently, a third RAN concept has been added, providing direct access to and from IP-based networks. Third-generation systems are ultimately expected to migrate toward IP as part of the global trend toward carrying all traffic types over packet networks. However, the current UMTS WCDMA specification explicitly defines ATM as the transport layer in the RAN. While some IS-2000 RANs also use ATM, the 3GPP2 specifications allow the manufacturer to choose the underlying transport layer. Given the planned migration paths of various 2G systems to 3G, 70% of 2G subscribers worldwide are expected to eventually migrate to some type of the CDMA version of 3G. The main network elements and interfaces referred to in the 3GPP specifications (see Figure 4.7) include the terrestrial components of the 3G system, referred to collectively as the RAN: • User equipment, also called mobile station, subscriber unit, or sim-
ply handset, including mobile cellular telephones, handheld PDAs, and cellular modems connected to PCs;
• Node B, usually called the RBS, providing gateway services between
the handset/RF interface and the RNC, via Iub interface;
Wireless-Network Architecture
Uu
(Handset) User equipment
Iub Node B
Radio network controller
ATM Radio access network
Iu
ATM
143
Mobile switching center
PSTN
Core network interface Internet core IP, WDM, PoS, ATM.
Figure 4.7 Main interfaces in 3G wireless network.
• RNC (or BSC), connecting to and coordinating as many as 150 base
stations (the RNC manages activities such as handing over active calls between base stations);
• CN interface, referring to other terrestrial CN infrastructure con-
nected to the RAN through Iu interface, such as the Internet and PSTN (the gateway device for this activity is usually called a mobile switching center or mobile multimedia switch).
An extensive set of protocols for communication within the RAN, to and from the user equipment, and between other networks has been developed by 3GPP. These protocols sit above AAL2 and AAL5. Together, they implement control-plane functions (e.g., signaling required to establish a call) and user-plane functions (e.g., voice or packet data). Wireless access in 3G network design constitutes one of the most important and driving requirements for the application of ATM extended with AAL2 switching (more information on ATM and AAL can be found in [2]). The following description is the high-level data flow for voice and data call in CDMA2000 BSC and toward other parts of the RAN showing practical application of AAL2 and AAL5. New terminology used here aside from the well-known backhaul includes sidehaul and fronthaul. Fronthaul is a connection between BSC and MSC, and sidehaul is a connection between two BSCs.
144
Transmission Systems Design Handbook for Wireless Networks
Figure 4.8 shows the data flow for a land-to-mobile voice call in the 3G wireless system architecture. Voice traffic arrives at the BSC from the MSC (1). The fronthaul module receives PCM-encoded DS0s from the MSC and converts this traffic into AAL1 for transmission to the echo cancellation and vocoding module (2, 3). It also performs the required functions on the traffic channel, and converts the resulting data stream to the modified AAL2 format for transfer to the SEP module (4, 5). Modified AAL2 is a proprietary format used within the BSC. It is very similar to AAL2, with the exception that the channel identifier (CID) bit of the common part sublayer (CPS) packet is not used. This results in a simpler internal format, and therefore processing complexity and delays are both reduced. It should be noted that modified AAL2 is only used internally within the BSC, while all external AAL2 interfaces use standard AAL2. The selector element processing (SEP) module performs the selector element processing functions and sends the resulting modified AAL2 traffic channel to either the backhaul module for transmission to the appropriate RBS or to the sidehaul interface module for inter-BSC handoff (6, 7). The appropriate interface module converts the modified AAL2 data stream to the standard AAL2 format for transmission to the RBS or BSC (8). In the case of backhaul transmission to the RBS, the interface module also provides the necessary formatting. In the case of sidehaul transmission, the interface module converts the modified AAL2 format into AAL2. T1/E1 MSC
1
Radio-packet
Fronthaul
PDSN
AAL1 2 3 AAL1
Echo cancellation Modified vocoding AAL2 4 Modified AAL2 7
ATM
Sidehaul interface module AAL2 packet 8 over T1/E1 Second BSC
Figure 4.8 Voice call in 3G wireless network.
6 Modified AAL2 multicast AAL2U
SEP
5 7
Modified AAL2 Backhaul inteface module AAL2 packet 8 over T1/E1 RBSs
Wireless-Network Architecture
145
Figure 4.9 shows the data flow for a land-to-mobile packet call in the BSC architecture. Packet traffic arrives at the BSC from the PDSN (1). The interface module receives the IP/AAL5-formatted data stream over an OC-3 link. The interface module converts this traffic into AAL5 packets for transmission to the SEP module (2, 3). The SEP module performs the required selector element processing. This converts the output to the AAL5 format and sends the resulting AAL5 traffic channel to either the backhaul module for transmission to the appropriate RBS, or to the sidehaul module for inter-BSC handoff (4, 5). The appropriate backhaul module converts the AAL5 data stream to the standard AAL5 format for transmission to the RBS or BSC (6). In the case of transmission to the RBS, the backhaul module also provides the necessary formatting. In the case of sidehaul transmission, the interface module converts the modified AAL2 format to standard AAL2. 4.2.7
3G Traffic Classes
When defining the UMTS QoS classes, also referred to as traffic classes, the restrictions and limitations of the air interface have to be taken into account. It is not reasonable to define complex mechanisms as they have been defined in fixed networks, due to different error characteristics of the air interface [3]. T1/E1 Radio-packet
Fronthaul
MSC
IP/AAL5 over OC-3
PDSN
1
2 AAL5 Modified AAL2/5 4 multicast
Echo cancellation vocoding Modified AAL2/5
ATM
5
Sidehaul interface module 6
AAL2/5 packet over T1/E1
Second BSC
Figure 4.9 Packet call in 3G wireless network.
SEP
AAL5 3 5
Modified AAL2/5 Backhaul interface module 6
AAL2/5 packet over T1/E1
RBSs
146
Transmission Systems Design Handbook for Wireless Networks
The QoS mechanisms provided in the wireless network must be robust and capable of providing reasonable QoS resolution. There are four separate QoS classes: 1. Conversational class; 2. Streaming class; 3. Interactive class; 4. Background class. The main distinguishing factor between these QoS classes is how delay sensitive the traffic is. Conversational class is meant for traffic that is very delay sensitive, while background class is the most delay-insensitive traffic class. Conversational and streaming classes are mainly intended to be used to carry real-time traffic flows. Conversational real-time services, like video telephony, are the most delay-sensitive applications, and those data streams should be carried in conversational class. The interactive background and classes are mainly meant to be used by traditional Internet applications like the Web, e-mail, Telnet, FTP, and news. Due to looser delay requirements as compared with conversational and streaming classes, both provide better error rate by means of channel coding and retransmission. The main difference between interactive and background class is that interactive class is mainly used by interactive applications (e.g., interactive e-mail or Web browsing), while background class is meant for background traffic (e.g., background download of e-mails or background file downloading). Responsiveness of the interactive applications is ensured by separating interactive and background applications. Traffic in the interactive class has higher priority in scheduling than background-class traffic, so background applications use transmission resources only when interactive applications do not need them. This is very important in wireless environments, where the bandwidth is low as compared with fixed networks. 4.2.7.1 Conversational Class
The most well-known use of this scheme is telephony voice. But with Internet and multimedia, a number of new applications will require this scheme, for example voice over IP and videoconferencing tools. Real-time conversation is always performed between peers (or groups) of live (human) end users. This is the only scheme where the required characteristics are strictly given by human perception.
Wireless-Network Architecture
147
A real-time conversation scheme is characterized by the transfer time being low because of the conversational nature of the scheme, and at the same time, the time relation (variation) between information entities of the stream must be preserved in the same way as for real-time streams. The maximum transfer delay is given by the human perception of video and audio conversation. Therefore, the limit for acceptable transfer delay is very strict, as failure to provide low enough transfer delay will result in unacceptable lack of quality. The transfer delay requirement is therefore both significantly lower and more stringent than the round-trip delay of the interactive traffic case. Real-time-conversation fundamental characteristics for QoS are as follows: • Preserved time relation (variation) between information entities of
the stream; • Conversational pattern (stringent and low delay). 4.2.7.2 Streaming Class
When the user is looking at (listening to) real-time video (audio), the scheme of real-time streams applies. The real-time data flow is always aiming at a live (human) destination. It is a one-way transport. This scheme is one of the newcomers in data communication, raising a number of new requirements in both telecommunication and data communication systems. It is characterized by the fact that the time relations (variation) between information entities (i.e., samples, packets) within a flow must be preserved, although it does not have any requirements on low transfer delay. The delay variation of the end-to-end flow must be limited to preserve the time relation (variation) between information entities of the stream. But as the stream normally is time aligned at the receiving end (in the user equipment), the highest acceptable delay variation over the transmission media is given by the capability of the time-alignment function of the application. Acceptable delay variation is thus much greater than the delay variation given by the limits of human perception. Real-time streamings fundamental characteristic for QoS is the preservation of time relation (variation) between information entities of the stream. 4.2.7.3 Interactive Class
When the end user (either a machine or a human) is on-line requesting data from remote equipment (e.g., a server), this scheme applies. Examples of human interaction with the remote equipment are Web browsing, database
148
Transmission Systems Design Handbook for Wireless Networks
retrieval, and server access. Examples of machines interaction with remote equipment are polling for measurement records and automatic database inquiry. Interactive traffic is the other classic data-communication scheme that on an overall level is characterized by the request response pattern of the end user. At the message destination there is an entity expecting the message (response) within a certain time. Round-trip delay time is therefore one of the key attributes. Another characteristic is that the content of the packets will be transparently transferred (with low BER). Interactive traffic fundamental characteristics for QoS include the following: • Request response pattern; • Preserve payload content. 4.2.7.4 Background Class
When the end user, typically a computer, sends and receives data files in the background, this scheme applies. Examples are background delivery of e-mails, SMS, download of databases, and reception of measurement records. Background traffic is one of the classic data-communication schemes that on an overall level is characterized by the fact that the destination is not expecting the data within a certain time. The scheme is thus more or less delivery-time insensitive. Another characteristic is that the content of the packets is transparently transferred (with low BER). Background traffic fundamental characteristics for QoS include the following: • Destination does not expect data within a certain time; • Preserved payload content.
4.3 3G Transmission Networks 4.3.1
Replacing TDM with ATM in Transmission Networks
Wireless networks are leading the evolution of the information and communications society toward the mobile information society (MIS). This means that subscriber numbers are continuing to increase as mobile penetration reaches new heights. Also, multimedia communications and other packet-based traffic will gradually increase their role and finally predominate in mobile networks. This development has already started with modest data volumes over current
Wireless-Network Architecture
149
mobile (and wireless in general) networks; a rapid increase in data applications and traffic is expected soon. New technologies and technical solutions enable higher data volumes right now in existing networks. In GSM networks, HSCSD and GPRS greatly expand these networks capabilities to handle data traffic; they also enable new and user-friendlier applications thanks to the higher bit rates available. This development will continue with still higher bit rates over the air interface in the new 3G WCDMA- and CDMA2000-based networks and 1xEV-DO (also called HDR). These increasing data traffic volumes mean that the share of the packet-based traffic in the total traffic mix in the mobile network is increasing, that the same time as total traffic volumes are also rising rapidly. Evolution of the circuit-switched networks into packet-based networks will take some time, and should be done in well-planned and managed steps, so that the efficiency of the mobile network is preserved during the change over phase. In many cases, basic mobile voice services are also growing quickly due to growth in the number of subscribers, which also contributes to the overall traffic increase and continues to require economic solutions for this type of traffic. Therefore, the well-planned steps are vital to manage mobile operators cash flows and to make full use of existing investments. It is in the interest of a mobile-network operator to direct his future transmissionnetwork strategy toward this expected increase in the penetration of advanced data services. Transmission is an important element in any wireless network, affecting both the services and service quality offered, as well as the costs of the wireless operator. Optimization of transmission solutions is thus certainly worthwhile from the operators business point of view. In current mobile networks, transmission has been optimized for the narrowband circuitswitched traffic and this type of traffic will continue to dominate for some years. However, as stated above, packet-based information over the mobile network will show rapid growth and any reasonable network development plan must take this into account and plan for a smooth and economic transition and evolution path for the transmission network. So, in broad terms, the transmission network must continue to provide well-engineered and economically optimized solutions for the growing volumes of circuit-based traffic, while at the same time develop the strategy and the readiness to cope with the even faster growing data traffic of the future. This type of transmission solution is needed in all parts of the mobile network, both in access networks with many points and low-capacity links, as well as in CNs with high traffic volumes. This means, for example, that in a wireless network, a transmission solution is needed that provides for efficient transport of large number of
150
Transmission Systems Design Handbook for Wireless Networks
voice channels and that can evolve to also carry packet-based traffic, either ATM or IP or both. The solutions might be similar or different in different parts of the network and even the role and share of the different traffic types (TDM, ATM, and IP) might be different, but the transmission network must support them all in a planned and managed way. In any mobile network, there are different transmission needs, typically divided into two main application areas with their own characteristics: 1. The access transmission network, which connects the base stations to the closest network control or network hub point, and called here the access transmission network or backhaul. It would include the radio base station, BSC (or RNC), and BSC/MSC. 2. The core transmission network, which connects the control (or hub) points to the mobile network switching centers, and called here the core transmission network. The radio network will be connected to the CN by a backbone network (access and core transmission network), allowing wideband access and interconnection of subscribers. The 3G backbone network can use any transport technology, but is certain to be based on packet technologies, such as ATM and IP. The backbone network is built as a mesh of IP routing or ATM switching nodes interconnected by point-to-point links. Technologies such as IP over ATM may be used that uses ATM switching to multiplex IP traffic. This IP over ATM architecture supports voice traffic alongside IP. Many vendors prefer a pure end-to-end IP approach, whereas others prefer an ATM/IP hybrid to guarantee quality of service. Alternatively, IP over SONET/SDH could be a different backbone network solution that would eliminate the ATM layer by establishing point-to-point links between IP routers directly over SONET/SDH rings that run over a dense-wavelengthdivision multiplexing (DWDM) layer. This enables terabits per second (Tbps) of aggregate network bandwidth. When transporting voice over a packet-based network, overhead is added to each voice packet. The amount of overhead depends very much on what the protocol stacks for the user data look like. Figure 4.10 sketches the different possible protocol stacks. So if speech is transported over an AAL2/ATM connection, 5 octets are needed for the ATM cell header and 3 octets for the AAL2 minicell header. Assuming the AMR CODEC is used (WCDMA systems), the speech frame size can vary between 10 and 40
Wireless-Network Architecture
Voice RTP UDP IP LAC-U GTP UDP IP AAL5 ATM SDH Voice over GPRS
Header size 12 8 20 13? >=97 20 8 20 >=8
Voice RTP UDP IP PPP AAL5 ATM SDH Voice over IP on ATM AAL5
Header size 12 8 20
>=49
12 >=8
Voice RTP UDP IP PPP SDH Voice over IP
Header size 12 8 20 >=41 12
151
Voice AAL2 ATM SDH
Header size 3 5
Voice over ATM AAL2
Figure 4.10 User-plane protocol stack alternatives.
octets, depending on the quality of the radio link. If for an average connection the speech frame size is about 21 octets, two speech frames will fit into an ATM cell. So the overhead is 5.5 octets per speech frame in average. Bandwidth increase factor is 1.26. If IP is used directly on SDH, the overhead is calculated as about 41 octets per speech frame. By applying IP header compression, the overhead can be reduced to about 13 octets per speech frame. Bandwidth increase factor is 2.95 (1.62 for header compression). If IP is transported over ATM, the overhead is about 49 octets per speech frame plus the ATM cell header. Assuming again AMR CODEC, the IP datagram varies between 59 (49 + 10) and 89 (49 + 40) octets in size. This means that two ATM cells always have to be used, leaving empty space in the second ATM cell. So 106 (2 × 53) octets are always sent per speech frame. Bandwidth increase factor is 5.05. If the speech is sent via GPRS access, the calculation is even worse. When the transmission is packet based, silence suppression can be applied. This means that typically only 40% of the time does data need to be transferred. The rest is silence, and no bandwidth is used. This is common for ATM and IP transmission. IP is a best-effort transport, so IP as such does not guarantee any QoS. Of course, platforms and systems such as the value-added service centers, gateways, billing systems, customer service elements, IN systems, and the
152
Transmission Systems Design Handbook for Wireless Networks
like will also need to be upgraded. High-bandwidth over-the-air applications (data, video, etc.) also require high-capacity transmission systems, much more so than 1G and 2G networks. Increasing subscriber numbers and share of data traffic creates significant growth in the transmission capacity needed for long-term network evolution. The implication of this is an increasing role of fiber and high-capacity microwave systems (SDH/SONET) in future transmission networks and their physical-layer implementation. There are more and more data traffic and applications that create bursty traffic. Thus, in order to provide cost-effective transmission solutions for data traffic, packet-switching solutions introducing dynamics to traffic handling must be provided. Also, investments made in fiber- and microwaveradio-based transmission links are very important as new evolution phases can be built on top of existing infrastructure. Wireless networks of the next generation will also require new, more advanced solutions for the core and access transmission networks. Data traffic is inherently variable, and transporting it over the TDM network is inefficient. In a TDM-based approach, timeslots are dedicated for connections regardless of whether information is actually being sent. In a multiservice network, the underlying network can be physically subdivided into multiple networks, one for each service (i.e., voice, data, private lines, etc.). Using an ATM-based infrastructure, much more efficient use of transmission network is possible, since ATM allocates bandwidth on demand based on immediate user needs. In 2G wireless networks, deterministic multiplexing is applied, when each connection is characterized by a constant bandwidth (e.g., one timeslot). The minimum needed bandwidth over the physical link is then simply the sum of the constant bandwidths of the connections. Since the traffic characterization is not probabilistic, statistical gain is not available. Third-generation wireless networks use packet-switched (ATM) systems and statistical multiplexing. When several connections from VBR sources are multiplexed together, a statistical multiplexing gain is obtained (Figure 4.11), because there is a certain probability that traffic bursts on different connections do not appear at the same time. It is possible to maintain the same blocking probability with less bandwidth if statistical multiplexing is used instead of deterministic multiplexing. The price for it is that the QoS (packet delay and loss) will not be ensured in a deterministic, but in a probabilistic, fashion. Statistical multiplexing of data traffic can occur side by side with the transmission of the delay and loss-sensitive traffic such as voice and video [4]. Like voice telephony, ATM is fundamentally a connection-oriented tele-
Wireless-Network Architecture
153
X Mpbs 1
Y Mbps
X Mpbs 3
2
Z Mbps
Y Mbps
Due to the multiplexing gain, here: Y < X + X; Z < Y + Y;
Figure 4.11 Multiplexing gain.
communications system. That means that a connection must be established between two points before data can be transferred between them. An ATM connection specifies the transmission path, allowing ATM cells to self-route through an ATM network. Being connection-oriented also allows ATM to specify a guaranteed QoS for each connection. Since voice telephony is a real-time application, delay, among other quality measurements, is the most important factor that affects the quality of voice. According to ITU-T Recommendation G114, an end-to-end delay of 0 to 150 ms is acceptable for most applications. A delay of 150 to 400 ms is acceptable assuming that the administrators are aware of the transmission time impact of the transmission quality of user application, but any delay over 400 ms is unacceptable for general network planning purpose [5]. AAL2 has been designed and used to reduce the packing delay for narrowband services. The idea is to multiplex voice packets from several sources onto one ATM cell so that the time to fill a cell can be reduced significantly. IP over ATM and use of IP routers is a basis for 3G wireless network transmissions. An IP router is a packet-switching device used to connect several different networks to form one common network based on IP
154
Transmission Systems Design Handbook for Wireless Networks
networking technology. Based on its understanding of the network of which it is a part, the router decides how each packet is going to be forwarded, but it also must be able to differentiate between high-priority packets and lowpriority packets and make the right decision and avoid internal congestion at the same time. IP routers for wireless networks must efficiently be able to handle small packets of data, low-speed links, delay-sensitive traffic, synchronization, a large number of nodes, and continuous on-line connections. These demands come from the nature of wireless traffic, where low-priority packets cannot block the way for high-priority voice packets. Routers in wireless networks must also be able to provide radio base stations with a high-quality synchronization signal that is distributed via transmission links between routers and base stations, assuming that a global positioning system (GPS) is not used. Routers in 3G wireless networks are an integral part of products, such as RNC for WCDMA, BSC for CDMA2000, and MGWs for packet data services. 4.3.2
Importance of AAL2
Third-generation wireless networks will integrate multimedia services, 2G voice services, and TCP/IP networks [6]. As mentioned earlier, data traffic is inherently variable, and transporting this type of traffic over an underlying TDM network is inefficient and thus expensive. An ATM-based approach can take advantage of the statistical nature of data traffic in addition to the constant rate of voice in order to provide a more bandwidth-efficient solution. Of course, ATM over the fiber-optic media has to have the same traditional transmission-network functionality like multiplexing, grooming, add-drop multiplexing, and protection while offering additional services like guaranteed performance, virtual private networks, prioritized rerouting, adjustable statistical multiplexing levels, security screening, customer network management, and so on. With an ATM-based transmission infrastructure, provisioning circuits, changing their bandwidth, and monitoring them is possible through the use of a standard network management system. In deterministic multiplexing, each connection is allocated its peak bandwidth. In ATM, statistical multiplexing is used where the amount of bandwidth allocated in the network to the VBR source is less than its peak, but greater than the average bit rate. So, the sum of the peak rates of connections multiplexed can be greater than the link bandwidth as long as the sum of their statistical bandwidths is less than or equal to the provisioned link bandwidth. The bandwidth efficiency due to statistical multiplexing increases as the statistical bandwidths of connections get closer to their
Wireless-Network Architecture
155
average bit rates and decreases as they approach their peak rates. In general, though, statistical multiplexing allows more connections to be multiplexed in the network than deterministic multiplexing, therefore allowing better utilization of network resources. Generally speaking, efficiency gain due to statistical multiplexing is a factor of a number of different connection characteristics and network parameters. For example, depending how bursty the sources are and the length of those bursts, the efficiency gain due to statistical multiplexing may or may not be significant. In 3G wireless networks, core and access transmission networks are ATM based, so the calculations of the required transmission links will be a lot different from 1G and 2G circuit-switched wireless networks. The AAL performs functions required by the user, control and management planes, and supports the mapping between the ATM layer and the next higher layer. The functions performed in the AAL depend upon the higher-layer requirements. In short, the AAL supports all of the functions required to map information between the ATM network and the non-ATM application that may be using it. The transport of voice traffic over ATM networks has been a fundamental principle of its design from the start. However, its deployment has been problematic, and telephony over ATM is still an area that is being developed. Normally we regard voice over ATM as the transport of voice over emulated circuits (a replacement for a PRI or T1 carrying voice calls). ATM has always been able to carry voice and data over the same wires, since that is what it was designed to do. In an uncompressed format (standard 64Kbps PCM), the traffic is CBR and is presented to the network over a circuit using AAL1. When compression is deployed it is possible to use AAL5, or the AAL2, which is specifically designed to work with compression hardware [7]. But until recently, voice transport via ATM relied primarily on AAL1 circuit emulation, which adds 12% to 15% overhead to every voice circuit. Moreover, AAL1 lacks bandwidth-saving features like voice compression and silence suppression. With AAL1, an emulated T1 (1.544-Mbps) circuit requires 1.74 Mbps of ATM bandwidth and is not really applicable for costsensitive real-world applications. The new AAL2, on the other hand, was designed specifically for costeffective voice transport. AAL2 is used in 3G wireless networks as a backhaul connection between RBSs and the BSC. A new adaptation layer was required to provide the flexibility for network operators to control delay on voice services and to overcome the excessive bandwidth needed by using structured circuit emulation. AAL1 simply cannot be extended to meet these new ATM
156
Transmission Systems Design Handbook for Wireless Networks
networking requirements. AAL2, as specified in ITU-T Recommendations I.363.2 (1997), I.366.1 (1998), and I.366.2 (1999), carries the specific mandate to provide efficient voice-over-ATM services. Developed by the ITU and adapted by the ATM Forum in February 1999, AAL2 as defined in ITU-T Recommendation I.363.2 includes the following capabilities: • VBR-real time (VBR-rt) support. While AAL1 supports only CBR
transmission, AAL2 supports both CBR and a traffic class called VBR-rt, which is a better fit for voice calls and other applications that send information at a variable rate.
• Statistical multiplexing. Unlike AAL1 circuit emulation, which
reserves a fixed amount of bandwidth for each circuit, AAL2 can allocate unused bandwidth to other traffic on demand.
• Cell sharing. AAL2 can pack several short packets from different
sources into one ATM cell, letting multiple connections share the same bandwidth with less overhead.
• Variable packet fill delay. To let service providers balance delay
against efficiency, AAL2 supports variable settings for packet fill delay, the time allotted to stuff packets into cells before putting them on the wire.
• Voice optimization. AAL2 includes specific bandwidth-saving fea-
tures like voice compression, silence detection and suppression, and idle voice channel suppression.
Voice communication by nature is half duplex; one person is silent while the other speaks. There are also pauses between sentences and words with no speech in either direction. By taking advantage of these two characteristics, it is possible to save bandwidth by halting the transmission of packets during these silent periods. This is known as silence suppression or digital speech interpolation (DSI). The extra bandwidth saved from the silent period of one voice channel can be used by other connections if using AAL5or AAL2-based connections. With AAL1 (circuit emulation-based services), these savings can only effectively be used by ABR or UBR services since the connection admission control mechanism will have allocated the bandwidth required for CBR QoS. This technique can improve bandwidth utilization by as much as 40%. In order to create a natural-sounding conversation, background noise can be generated at the far end to recreate a realistic environment.
Wireless-Network Architecture
157
AAL2s VBR service handles voice-over-ATM far more efficiently than the CBR service of AAL1, with its inefficiently utilized, permanently allocated bandwidth. Before AAL2, users wanting to implement voice-overATM had to live with AAL1s limitations or adopt a proprietary solution (increasing network efficiency but negating interoperability). The new standard means that ATM switches can extend to voice the benefits of ATMs statistical gain. Access connections using this scheme can transport voice circuits over the same facilities as data circuits, minimizing the use of precious bandwidth. AAL2 is designed to make use of the VBR ATM traffic classes (with higher multiplexing gains), providing bandwidth-efficient transmission of low-rate, short and variable packets for delay-sensitive applications. AAL2s structure lets network administrators take traffic variations into account in the design of an ATM network optimized to match traffic conditions. AAL2 also enables multiple user channels on a single ATM virtual circuit and varying traffic conditions for each individual user or channel. Its structure also provides for the packing of short-length packets into one (or more) ATM cell and the mechanisms to recover from transmission errors. Compared to AAL1 and its fixed payload, AAL2 handles a variable payload within cells and across cells. This provides a dramatic improvement in bandwidth efficiency over either structured or unstructured circuit emulation using AAL1. It has been proven that AAL2 provides good mechanisms for fine adjustment of packetization delay and for achieving high bandwidth efficiency for low bit rate and VBR application. The cost is some overhead in the form of headers. This overhead is not important at all when compared with that for techniques, such as partial filling of cells (AAL2 behaves at least as well as partial filling), but may be significant when users have a choice among several AALs. For example, for transmitting long packets (longer that 45 octets), the user may chose either AAL2 (through it segmentation SSCS) or AAL5 (explicitly specified for long packet data). It is known that for small packets (up to a few hundred octets) AAL2 is more efficient than AAL5, while for long packets AAL5 is definitely recommended. In this case the user should select the AAL based on its data-generation patterns. A very realistic application for AAL2 switching has been identified for 3G of wireless networks, in the frame of IMT-2000 standardization, to support the functionality known as soft handoff. The ATM using AAL2 for narrowband services described in this specification fulfills an urgent market need for an efficient transport mechanism to carry voice, voice-band data, circuit-mode data, frame-mode data, and fax traffic. Voice transport will include support for compressed voice and
158
Transmission Systems Design Handbook for Wireless Networks
noncompressed voice together with silence removal. Reference [8] describes the procedures and signaling required to support the efficient transport of narrowband services across an ATM network between two interworking functions (IWFs) to interconnect pairs of non-ATM trunks. It specifies the use of ATM virtual circuits with AAL2 to transport bearer information and ATM virtual circuits with AAL2 or AAL5 to transport CCS. The virtual circuits used may be PVCs, SPVCs, or SVCs. The specification supports the transport of common channel signaling (CCS) information as well as channel-associated signaling (CAS) information. ATM trunking using AAL2 provides both switched and nonswitched services to the narrowband network. 4.3.3
QoS Concept
From an engineering planning prospective, the initial choice of network topology can seriously impact initial investment and future flexibility. Although much has been learned in terms of network topology (system design), LOS and path-loss (coverage) issues, and QoS optimization issues, much is still in the process of being developed, tested, deployed, and tested again in the field. It is already proven that the delivery of raw bandwidth over wireless media without acceptable QoS will not result in market acceptance. Without clearly understanding QoS in the context of a wireless broadband access system, it is unlikely that the underlying architecture and the resulting hardware and software design will result in a successful system. The issues of quality delivery are somewhat more complex for wireless broadband access systems than for wireline systems. Data-delivery problems include slow peripheral access, data errors, dropouts, unnecessary retransmissions, traffic congestion, out-of-sequence data packets, latency, and jitter. In addition, wireless access introduces high inherent BER, limited bandwidth, user contention, radio interference, and TCP traffic-rate management. QoS mechanisms must address all of these concerns. In data networking, quality usually implies the process of delivering data in a reliable and timely manner. The definition of reliable and timely is dependent upon the nature of the traffic being addressed. These terms may include references to limitations in data loss, data retransmission, and packet order inversions, as well as data accuracy expectations and latency variations (jitter). QoS is a complex concept, requiring a complex mechanism for implementation. A casual user doing occasional Web browsing, but no FTP file downloads or real-time multimedia sessions, may have a different definition of QoS than a power user of large databases or financial files, frequent
Wireless-Network Architecture
159
H.323 video conferencing and IP telephony. For example, in a wireless system, QoS mechanisms must cope not only with considerations particular to the wireless environment, but with wireline-networking considerations as well. In ATM networks, traffic descriptors are usually used only as a rough guide, and many service providers systematically practice overprovisioning and allocating a lot more bandwidth than necessary [9]. Although not a very efficient system, the perceived QoS is satisfactory in most cases. The migration from circuit-switched to ATM and a packet-switched network has also affected QoS mechanisms. An IP-centric wireless system for packet-switched network traffic requires a new approach to provide optimal QoS performance. The use of QoS as the underlying guide to system architecture and design constitutes the fundamental differentiation between wireless broadband access systems designed with traditional circuit-centric or ATM-centric approaches and IP-centric wireless broadband access systems. Queuing is the commonly accepted tool for manipulating data communications flows. Data packets must be queued for packet headers to be examined or modified, for routing decisions to be made, or for data flows to be output on appropriate ports. However, queuing introduces a delay in traffic streams that can be detrimental and can totally defeat the intent of queuing. Excessive queuing can delay time-sensitive packets beyond their useful time frames or increase the round trip time, producing unacceptable jitter or causing the data transport mechanisms to time-out. Therefore, queuing must be used intelligently and sparingly, without introducing undue delay in delaysensitive traffic. To achieve high-quality (often referred to as toll-quality in circuitswitched networks) voice transmission, the absolute amount of transmission delay, as well as the variation in that delay, called jitter, must be kept low. In a wireless environment where TDMA, FEC, and other such techniques are necessary, queuing must be used only to enable packet and radio-frame processing. However, in the case of real-time flows, the overall added delay in real-time traffic should be held to below about 20 to 25 ms. This amount of delay is not perceptible to the human ear. Several additional factors contribute to delay in data networks, including cell-packing delay, coding and compression delay, queuing delay, and freeze-out delay. Cell-packing delay, or packetization delay, is simply the amount of time it takes for the sending device to fill packets before they can be sent. Coding and compression delay represents the amount of computational time that is needed for the sampling, quantization, and compression of the signal. This delay can become significant if a large amount of compression is being done to the signal. Queuing delay occurs in networks experiencing congestion and represents
160
Transmission Systems Design Handbook for Wireless Networks
how long packets must wait in queue at a bottleneck node in the network. The use of queue management as the primary QoS mechanism in providing QoS-based differentiated services is a simple and straightforward method for wireless broadband systems. Wireless systems are usually more bandwidthconstrained, however and are therefore more sensitive to delay than their wireline counterparts. So QoS-based differentiated services must be provided with mechanisms that go beyond simple queuing. In a network not experiencing congestion, the most significant type of delay should be freeze-out, or serialization delay. This delay is due to the fact that packets take a finite amount of time to transmit, and during this time, the channel is not available to transmit any other information. For example, when two voice channels are packetized and sent over the same data link, each has to wait to send its packets while the link is busy transmitting the other channels packets. Freeze-out delay can be roughly quantified as the size of packets divided by the speed of the channel (in bps). Freeze-out delay by itself provides a minimum estimate of one-way delay for packetized voice traffic; in reality, actual delay would be higher due to the additional delays mentioned before. TCP controls transmission rates by sensing when packet loss occurs. Because TCP/IP was created primarily for the wireline environment with an extremely low inherent BER (today it is on the order of −9 1 × 10 or better for fiber optics), TCP assumes any packet loss is due to network congestion, not error. Therefore, TCP assumes that the transmission rate exceeds the capacity of the network and slows the rate of transmission; however, packet loss in the wireless link segment is due primarily to the high inherent BER, not congestion. With a range of data flows, each having different bandwidth, latency, and jitter requirements, the IP-centric wireless system must be able to manage QoS mechanism parameters over a wide range in real time. The QoS mechanism must be able to alter system behavior to the extent that one or more data flows corresponding to specific applications be transparently switched on and off from the appropriate end users. This approach is in contrast to other QoS mechanisms that seek to achieve high QoS by establishing circuit-centric connections from end to end without regard for the underlying applications actual QoS requirements. By providing an applicationspecific QoS mechanism, scarce wireless bandwidth can be conserved and dynamically allocated where needed by the QoS mechanisms associated with each application type. Mobile quality of service (M-QoS) has been defined in order to augment traditional QoS requirements found in wired ATM networks [10]. This wireless QoS refers to the QoS parameters associated with the wireless
Wireless-Network Architecture
161
links, such as link delay, bit error rate, and channel reservation, as well as the performance parameters associated with the handoff blocking probability and cell loss. 4.3.4
ATM Physical Layer
Bandwidth is the key resource in both circuit- and packet-based networks. In circuit-based networks, a fixed amount of bandwidth is dedicated to a call in progress. Since 55% to 60% of the call consists of silence, circuit-based networks do not make optimum use of bandwidth. The principal advantage of packet-based networks is that they use bandwidth much more efficiently. Unlike circuit-based networks, packets from many different sources share a circuit, allowing for efficient use of fixed capacity. ATM seems to be the technology of choice for many different 3G wireless networks. The physical layer is at the lowest level of the ATM stack. It takes the full cells from the mid-layer and transmits them over the physical medium. The ITU-T originally defined only two speeds that should be supported by ATM (i.e., 155.52 Mbps and 622.08 Mbps); however, over time a number of additional speeds and interfaces have evolved, going as low as E1/T1 and as high as the Gbps range. The physical layer itself is subdivided into two sublayers: the transmission convergence (TC) sublayer and the physical medium dependent (PMD) sublayer. These two sublayers work together to ensure that the optical or copper interfaces receive and transmit the cells efficiently, with the appropriate timing structure in place. ATM, being an international transmission technology, has to be able to work with a variety of formats, speeds, transmission media, and distances that may vary from country to country. The standardization of the physical layer interfaces enabled just such connectivity. Single-mode fiber, multimode fiber, coaxial pairs, and shielded and unshielded twisted pairs are today all standardized for use in the ATM environment. The TC sublayer takes care of header error check (HEC) generation and verification, cell scrambling and descrambling, cell delineation, and decoupling. The HEC is a one-byte field in the ATM cell header, which protects the header from errors. The PMD sublayer covers bit timing, line coding, the physical connectors, and signal characteristics. The UNI documents detail the physical media types allowed at the user interface and the details differ for the public and private UNIs. For example, Category 5 twisted pair is permitted at the private UNI, but not at the public UNI. The original objective for the physical layer was operation over the SDH and SONET only. When the ATM Forum V3.0 and V3.1 specifications were
162
Transmission Systems Design Handbook for Wireless Networks
ratified, however, other interfaces were included. These interfaces were the DS3 and a 100-Mbps interface based on the transparent asynchronous transmitter/receiver interface (TAXI) fiber distributed data interface (FDDI) standard. DS1 operates at 1.544 Mbps and is approved for ATM over twisted pair at a distance of up to 3,000 feet. DS3 operates at 44.736 Mbps on coaxial cable up to 900 feet. STS-1 (51.84 Mbps), STS-3c (155.52 Mbps), and STS-12 (622.08 Mbps) operate over single-mode fiber up to 15 km. E1 (2.048 Mbps) and E3 (34.368 Mbps), together with the Japanese standard J2 (6.312 Mbps), are standardized for coaxial cable with no distance specified. Recently, a definition for a 2.5-Gbps physical interface (an SDH interface) was completed. This definition describes how cells are mapped to this higher-speed transport. 4.3.4.1 ATM in SONET/SDH Fiber-Optic Networks
The integration of SONET/SDH and ATM involves more than offering interfaces that allow connectivity between the technology networks. The high-speed transmission attributes of SONET/SDH and the switching and bandwidth-management capabilities of ATM complement each other to form the foundation of broadband networking. SONET/SDH provides ATM with access to a high-speed infrastructure; conversely, ATM offers the high-speed traffic that takes full advantages of a SONET/SDH infrastructure. Key to integration of ATM and SONET/SDH is the ability to monitor and react to failures, ensuring survivability. SONET/SDH rings provide protection against the failures within the core transmission network, but should also extend to protect broadband services against link and node failures. The integrated ring provides facility-layer reliability (link) complementing ATM reroute, which provides ATM-layer reliability (node) in addition to link reliability. The ATM service platform should be able to intelligently monitor the SONET/SDH payload overhead. Upon the detection of a failure, the ATM services should be able to provide SONET/SDH 1+1 automatic protection switching as defined by ITU and Bellcore specifications. The ATM switch needs to provide one-for-one redundancy on the ports, and the complete switchover must take place within the specified 50-ms QoS parameter defined by service providers. If SONET/SDH switchover is unable to overcome the network failure, ATM reroute will reestablish the failed connection over a new route. These inherent SONET/SDH switchover-protection capabilities and ATM reroute capabilities are critical to new broadband services. Both SONET/SDH and ATM have operations and maintenance information in their headers, but use the information in complementary ways, with
Wireless-Network Architecture
163
SONET/SDH checking for errors on a span-by-span basis, while ATM considers performance on an end-to-end basis between different switches. ATM will send 50 cells per SONET/ATM frame, each with its own operations, administration, and maintenance functionality, making it possible for the ATM layer to detect a fiber cut much sooner than the SONET/SDH layer. This way, ATM can enhance the speed with which SONET/SDH detects problems. As per ITU-T I.630, ATM Protection Switching, the individual VP/VC protection-switching concept was developed to apply primarily to the situations where server-layer protection-switching does not exist. It is useful to protect only a part of VPs/VCs that need high reliability. The rest of the VPs/VCs remain unprotected. This helps to reduce the necessary bandwidth for protection and can be used for protection against ATM-layer defects as well as physical-layer defects. The ATM protection-switching architecture can be a 1+1 type or an m :n type. In the 1+1 architecture type, a protection entity is dedicated to each working entity with the working entity bridged onto the protection entity at the source of the protected domain. The traffic on working and protection entities is transmitted simultaneously to the sink of the protected domain, where a selection between the working and protection entity is made based on some predetermined criteria, such as server defect indication. In the m :n architecture type, m dedicated protection entities are shared by n working entities, where m ≤ n typically. The bandwidth of each protection entity should be allocated in such a way that it may be possible to protect any of the n working entities in case at least one of the m protection entities is available. When a working entity is determined to be impaired, it first must be assigned to an available protection entity followed by transition from the working to protection entity at both the source and sink of the protected domain. It is noted that when more than m working entities are impaired, only m working entities can be protected. In general, if lower-layer (e.g., SDH or optical) protection mechanisms are being utilized in conjunction with ATM-layer protection mechanisms, then the lower layers should have a chance to restore working traffic before the ATM layer initiates protection actions. The objective here is to avoid unnecessary protection actions and any issues of contention. 4.3.4.2 ATM in Microwave Radio Networks
ATM was originally developed for use in high-reliability fiber-optic SONET/SDH networks and not for more difficult media like radio. Over
164
Transmission Systems Design Handbook for Wireless Networks
the last few years more and more transmission systems, especially those in wireless networks, are using ATM over the microwave networks. These radio systems carrying packetized traffic such as ATM (or frame-relay) have to be designed in a way that takes into account behavior of this kind of traffic. Because ATM is primarily designed for an essentially error-free environment, in the wireless arena the sources of errors and their consequences on ATM traffic and its QoS are being studied today [11]. Although important in any network, error bursts are expected to be very significant sources of degradation in the microwave network. ATM is designed for low-BER links, and the radio links with just a moderate BER can cause unacceptably high cell-loss and misinsertion rates [12]. By definition, a misinserted cell is a received cell that has no corresponding transmitted cell on the considered connection. Cell misinsertion on a particular connection is caused by defects on the physical layer affecting any cells not previously associated with this connection. Since the mechanisms that cause misinserted cells have nothing to do with the number of cells transmitted on the observed connection, this performance parameter cannot be expressed as a ratio, only as a rate. In a process of dimensioning microwave point-to-point systems for ATM traffic, there are a number of issues to be considered. Since bit errors in the microwave system typically appear in multiples and spread less than the ATM header length, the singlebit-header correction feature may not improve cell-loss rate as much as predicted and intended. The latest research shows that the BER is degraded approximately one decade from the microwave radio system to the ATM CBR virtual circuit due to the cell loss. In wireless networks, broadband terminals as well as smaller and denser cells will increase the total capacity needs enormously. Many of todays 1E1/T1 links will be increased to STM-1/OC-3 and higher capacities and will require high-capacity SDH/SONET microwave radios even at spur links to the last cell site in the network. Aside from the capacity, these microwave radios need a very sophisticated error-correction technique to satisfy ATM transport layer requirements. −9 Normally, in a fiber-optic system, BER should be 10 measured at the −10 ATM CBR virtual circuit. The same quality corresponds to BER = 10 in −6 the microwave radio system. Systems with BERs worse than 10 are considered unavailable. Although still under research, the following facts should be kept in mind: −9
• To achieve 1 × 10 user BER from the ATM network, 1 × 10
required on the SDH radio link.
−10
is
Wireless-Network Architecture
165
−6
• When BER is above 1 × 10
(low BER) end users quality is very low (1 cell lost per second). AIS generation at low BER should be considered both for the purpose of rerouting and disconnecting. −5 • For ATM NMS operation, 8 × 10 is required on the SDH microwave radio link. 4.3.5
Traffic Modeling and Simulation Tools
The transmission network consists of the elements required for the transport of call and signaling information between the major nodes in the service network. The core transmission network covers elements required for the transport of calls and signaling between nodes in the CN. Typically, the major nodes are POIs, MSCs, BSCs, TSCs, STPs, HLRs, and ESNs. The access transmission network consists of the elements that are required for the transport of call and signaling information between the RBSs and BSCs in the service network. Due to recent and ongoing efforts of both regional and global standardization and research processes, there is a wide consensus regarding the basic architectural aspects of 3G mobile communications systems, including applying ATM as switching and multiplexing technology [11]. In order to utilize transmission facilities efficiently, traffic simulation models and tools will be used in the process of wireless-network planning. All three aspects of network planning (RF, core, and access) will be based on statistical methods and traffic-simulation tools. Multiplexing of different traffic streams has consequences on the dimensioning, since there will in many cases be a nonnegligible gain compared to just adding the capacity demands of individual traffic streams. The reason is that the variation of the individual streams is usually smoothed out in the aggregated traffic stream. It is often convenient to split up this socalled multiplexing gain into traffic levels: • Multiplexing gain on a call level for all services (only call arrivals are
considered, often modeled as a Poisson process); • Multiplexing on a transaction level (for services with discontinuous transmission, such as interactive services, voice with DTX); • Multiplexing on a packet level (as with VBR packet services). For constant bit rate services with continuous transmission, multiplexing gain can be obtained on a call level only. In order to include multiplexing
166
Transmission Systems Design Handbook for Wireless Networks
gain on subcall levels (for example, with VBR services) when dimensioning, it must be considered that arrivals of transactions or packets (depending on the level considered) often exhibit a pattern that deviates significantly from the Poisson process. Inverse multiplexing for ATM is a method that makes it possible for several physical links to carry a single ATM stream. The main advantage is increased robustness. The traffic is distributed on all physical links and in case of a failure on one physical link, the traffic is distributed over the remaining physical links. No traffic will be lost if the remaining capacity is sufficient. Another important factor is that larger links result in an increased potential for statistical multiplexing gain. In order to perform the calculation, the following data is needed: • Service characteristic description; • Traffic-related data; • GoS requirement.
The service description contains parameters representing the basic behavior of the service. The traffic mix is responsible for setting how intensive the given services are used in the RBS. This data is most likely based on the outcome of the cell-planning procedure. The GoS requirement reflects the percentage of the users that will not get service because of the limited resource on the transmission and switching network. There are two types of services that can be defined and characterized: circuit- and packet-type services. Circuit-type services are for services with definite bandwidth demand and holding time. These types of services are delay sensitive and demonstrate CBR or VBR with well-defined characteristics (like voice). The parameters that describe the circuit-type services are equivalent bandwidth and overhead. The equivalent bandwidth for CBR is the bit rate itself. For VBR it is somewhere between the peak and the average bit rate and it is very dependent on the traffic characteristics. The overhead of the service takes into account the amount of extra data volume that must be transferred on the transport network excluding the ATM overhead, which is included later on, when the logical links are mapped onto the physical ones. The overhead can consist of protocol overheads like frame protocol, MAC, RLC, and so on. Calculation of the overhead should take into consideration what kinds of overheads are already added in the equivalent bandwidth.
Wireless-Network Architecture
167
Packet-type services can be used to describe best-effort services that are not delay-sensitive services. For best-effort services, well-defined resources are not guaranteed within a certain time, and only an average throughput over a longer period of time (at least an hour) can be guaranteed. Mean BHT, burstiness factor, and overhead characterize this type of service. The average throughput is given by the mean BHT in bytes per hour. The burstiness factor describes how bursty the traffic is; that is, what portion of a given bandwidth can be utilized by the traffic (in an average level) to be able to handle the deviations from the average bit rate. Its value is between 0 and 1, where burstiness equal to 1 means nonbursty traffic (a typical value is 0.6 or 0.7 in the case of traffic generated by IP applications). The smaller the burstiness factor, the more bursty the traffic, requiring higher bandwidth. The overhead parameter has the same role as in the circuit-service type. The other group of information is traffic data and GoS requirement specific to an RBS. The traffic data is given by the traffic mix parameter, which is a list of services offered by the RBS and their level of usage. In case of circuit service, the latter one is an Erlang value describing the traffic volume. For packet service it is the average number of simultaneously attached users (as defined in WCDMA). In both cases they are meant to be a busyhour value. The GoS requirement is responsible for adjusting the percentage of the calls that cannot be served because of transmission resource shortage. The corresponding parameter name is blocking. It can be set for each RBS and affects only the circuit-type services. Trunk calculation for RBS-BSC backhaul interface is based on RBS capacity and can be defined in terms of the number of sectors and the number of simultaneous calls it can support. Let us assume that under full configuration, the CDMA2000 RBS will support 4 RF carriers, equating to a maximum of 12 RF sectors. This requires that 8 E1 (or 2 E1s per RF carrier) backhaul connections between the RBS and BSC. The actual number of E1/T1 spans, however, depends on traffic demand. Due to the high efficiency of ATM-based backhaul, the maximum number of 8K traffic channel per E1 span is 180 while the number of 13K traffic channels per E1 span is 125. These are typical CDMA2000 numbers that may differ from supplier to supplier. The RBS traffic load can be acquired from RF planning in terms of Erlangs. Then, the traffic is converted to the number of channels by Erlang B formula. Based on above maximum number of channels per E1 span, the number of E1 spans of RBS-BSC can be derived accordingly. For example, under mobility environment, 1 carrier and 3-sector RBS may require 1T1 or 1E1 backhaul to the BSC.
168
Transmission Systems Design Handbook for Wireless Networks
The interface of BSC-BSC is to support inter-BSC soft handoff which carries packet-based voice traffic channels. The inter-BSC handoff traffic Erlangs can be determined by the formula below: Handoff traffic (in Erlangs) = a × l × T where l = the handoff arrival rate at border cell, which is the function of mo-
bile speed, border cell radius, and the number of calls at border cell (air capacity); T = the mean usage duration, which is the average handoff call duration;
a = the ratio of the rate of system border crossings and the rate of cell
border crossing.
Based on the above equation, we are able to calculate handoff traffic in Erlangs. Again, the number of E1/T1 spans is calculated in the same way as the backhaul interface (RBS-BSC). 1 The interface of BSC-MSC is based on the IOS standard to support telephony service. The trunk group size depends on the voice and circuitswitched data traffic load on the BSC. Basically, the BSC traffic Erlang can be calculated from Erlang/sub and the amount of subscribers within the BSC. Then, it is converted to the number of trunks and E1s/T1s by Erlang B formula. Usually, trunk size is dimensioned under the load of 70% to protect the system from overloading. 4.3.6
2G and 3G Coexistence
The ATM Forum specification for circuit emulation service (CES) defines the means for ATM-based networks to employ AAL1 to emulate, or simulate, synchronous TDM circuits over the asynchronous infrastructure of ATM networks [2]. The circuit emulation (CE) function enables existing TDM circuits to be mapped over ATM. CE thus enables operators to migrate an existing TDM network to ATM while preserving the investment in TDM equipment. The reader should note that CE products are available for all major circuits, for example the American T1 standard and the 1. International Organization for Standarization.
Wireless-Network Architecture
169
European E1 circuit standard. CE is broken into two versions, structured and unstructured, but both versions of CE use AAL1 CBR connections. In structured CE, the ATM network recognizes the internal structure of the circuit and is able to recover this structure at the receiving end. Circuit structure refers to the timeslots. Structured T1/E1 supports N × 64 Kbps (fractional T1/E1). This means that particular timeslots may be mapped to different virtual circuits and hence to different destinations. Several timeslots from a source circuit may be mapped to one virtual circuit. E1 contains 32 timeslots per frame, and the first timeslot, timeslot 0, is used for framing. As framing is irrelevant within the ATM network this timeslot (timeslot 0) is often terminated within the first ATM switch and regenerated at the destination ATM switch, which produces an E1 output. Additionally, some means of recovering the original signal clocking must be available. Which method is to be used at the destination is communicated across the network in a CE call setup request, or is set by the administrator. All timeslot 1s from each frame are mapped to an ATM cell and, hence, to an ATM VC. In mapping the source information into cells in this manner, an important issue is encountered, that of latency. An AAL1 1-byte header will be included in each cell payload; thus, to use the ATM bandwidth as efficiently as possible we should wait for 47 timeslot 1s. In waiting this long to fill the ATM cell we may reach a point where the QoS of the application is compromised. To get around the latency issue we may choose to pad out the cell, which is to send a cell only partially filled with voice samples. The level of padding will depend on several factors, including how far the call has to go overall. Different strategies may be employed for different types of calls. For example, a call that is recognized as a long-distance or international call may be heavily padded by the CE function in order to minimize latency. That call will inherently have a long end-to-end delay. Similarly, it may be acceptable to map local calls to ATM cells without any padding to maximize bandwidth usage. Structured CE also specifies support CAS, commonly used by PBXs to indicate off-hook and on-hook conditions. Since the structured mapping of the individual timeslots does not convey TDM framing information end-toend, the CAS information is encoded and transported separately, requiring additional overhead. In unstructured CE, the network does not attempt to recognize the internal circuit structure. Rather, it simply transmits the entire circuit across the network. This unstructured service emulates a 1.5- or 2-Mbps data leased line. On E1 circuit, a 376-bit (or 47-byte) chunk of the source signal is taken. These 47 bytes have the AAL1 one-byte header added to make up the full 48-byte payload. An ATM cell header is added, making a full ATM cell.
170
Transmission Systems Design Handbook for Wireless Networks
These cells are then sent on a CBR connection. CE IWF provides timing to the TDM equipment in a synchronous mode, or accepts timing in an asynchronous mode. Timing transfer is critical for many legacy TDM networks, especially T1/E1 multiplexer networks. CE is most likely the method that will be used to combine existing TDMA networks with the new ATM-based 3G networks. On the other hand, existing transmission networks are based on PDH as defined in Recommendation G.702. ATM is considered the suitable technique to support B-ISDN. The SDH will form the basis of transport of the ATM cells, but during the transition period, there is the need to transport ATM cells using existing PDH transmission networks. Recommendation ITU-T G.804 provides the mapping to be used for this transport of ATM cells on the different PDH bit rates for both 1.544- and 2.048-Mbps hierarchies [13]. These mappings cover both the 1.544-Mbps and 2.048-Mbpsbased hierarchies and are used in conjunction with the frame structures defined in Recommendation G.832. The detailed requirements on how to map ATM on a fractional physical link will be in accordance with ATM Forum Document af-phy0130.000, ATM on Fractional E1/T1 (September 1999). 4.3.7
Transmission-Network Architecture
4.3.7.1 Transmission-Network Objectives
The main objectives of the transmission network are to connect all the points of interest, satisfy the capacity demands and provide reliable service using microwave, copper, fiber-optics or satellites. During the wireless-network build-out, it is important to establish a transmission plan that will include all the present requirements as well as future expansion (number of cell sites, RBS type, and future capacity requirements). Transmission network design typically involves a trade-off between network reliability and speed of deployment and price. An example of the small, three cell-site, RF network where mixed backhaul (transmission) media are usedleased T1/E1 lines and microwaveis shown in Figure 4.1. In wireless networks, the term backhaul (and sometimes access transmission or access transport network) is used to describe RBS-BCS connectivity exclusively. The terms core transport and core transmission are usually used to describe network connectivity between other network nodes. The core transmission network is the connection between both MSCs and BSCs and MSCs and the PSTN. BSC-MSC connectivity refers to the ability of the BSC to support the reliable transfer of signaling messages with the MSC for call (e.g., voice, fax,
Wireless-Network Architecture
171
and packet data) setup and teardown, mobility management, radio resource management, and transmission facilities (terrestrial circuit) management. BSC-BSC connectivity involves the ability of the source BSC to support signaling messages with the target BSC for direct RBS to RBS soft or softer handoff, access handoff, access probe handoff, and channel assignment into soft or softer handoff in CDMA networks. The supported signaling protocol will allow for efficient resource allocation and deallocation (inter-RBS connection setup and teardown) and call connection control between a source and a target RBS during soft handoff. Various types of transmission-network topologies are shown in Figure 4.12. Star and tree formats are examples of linear transmission-network architecture and used for small- to medium-size wireless networks. The size of the network is assessed based on the number of cell sites and the required backhaul capacity. The small wireless network shown in Figure 4.1 is an example of this type of transmission-network architecture. There is no network protection in this case and a problem on any of the E1/T1 links will affect one or more (in case of daisy-chain sites) cell sites. Added service reliability can be achieved with automatic rerouting. Many successful mobile operators protect transmission by using automatic traffic rerouting, assuring additional reliability in normal situations, such as when access microwave radio links suffer cutoff due to poor weather conditions or possible fiber-
Star Ring
Switch/NOC location BSC MSC
Tree Hub site Spur (end) site
Figure 4.12 Transmission network topology.
172
Transmission Systems Design Handbook for Wireless Networks
optic cable cuts or any other human error. With a flexible rerouting transmission system such as T1/E1 trunk rerouting, backup capacity can pass via physically separate routes, as the problem is not likely to interrupt both routes simultaneously. The base station trunk is the entire physical transmission link between two base stations or sites or between a base station and its base station controller, typically T1/E1 or nxT1/E1 links. In case of traffic failure, trunk rerouting switches all traffic in the main trunk simultaneously to the backup trunk. Large base stations comprising a number of circuits are switched simultaneously for minimum service downtime. Rerouting can be arranged for all sites or only critical sites, such as base stations that are labeled as higher priorityfor example, hub sites. Hub sites are those sites that collect traffic from more than one other site (typically three to four other sites) and carry that traffic toward the BCS location or fiber-optic ring hub site. For a larger transmission-network it is recommended that a ring configuration be used as a high-capacity backbone carrying traffic to the switch location. Additional fibers in the fiber-optic network or cross-polarization in the case of a microwave network can be used to further increase (double) the capacity of the ring. The ring configuration shown in Figure 4.13 has the BSC and MSC not colocated but interconnected through two SDH highcapacity networks. It is usually recommended that the BSC and MSC be colocated and placed close to the point of presence of the PSTN to simplify interconnect. Ring architecture is considered a reliable communication facility since it provides automatic protection from the following: • Site hardware (batteries, towers, antenna system) failures; • Radio or MUX equipment failures; • Propagation failures in the microwave network; • Cable cuts in the fiber-optic network.
A ring topology also provides basic user features such as simple operation, fault location, and maintenance, and it provides alternate routing of E1/T1 traffic automatically and no loss of E1/T1 traffic due to signal failure. Each E1/T1 circuit must be dedicated completely around the ring, and reuse of the same E1/T1 in the opposite direction is not possible. For ultimate reliability, both directions can be 1+1 hardware protected. In microwave systems
Wireless-Network Architecture RBS
173
RBS RBS N × 2 Mpbs
RT RT RT
ECC STM-1 RT
RT
RT
N × 2 Mpbs
N × 2 Mpbs
ECC
BSC
RBS RBS
RT
RBS
WS
Q3
RT
RT
STM-0
RBS
RT STM-1 ring
STM-1 ECC
TMN OS SDH transport network
Mobile MSC exchange
RBS RT
RT
RBS BSCBase station controller RBSRadio base station BRTRadio terminal
Figure 4.13 Microwave ring topology.
additional protection (e.g., space or frequency diversity) at lower frequencies may be required against short-term multipath outages. All the sites that belong to the ring are considered hub sites and have to be planned so that during the deployment stage they are completed first in order to provide connectivity and protection for the rest of the network (spur links). In PDH networks, additional hardware with built-in intelligence to assess the T1/E1 quality and switch circuits, if needed, will be required. This hardware has to be added at every site and it is useful for small networks. SONET/SDH have incorporated several protection and switching techniques from their inception. These include linear APS, path-switched rings, line-switched rings, and virtual rings. These techniques provide the ability for a network to detect the problem (under 10 ms) and heal itself automatically in the case of failure with a restoration time under 50 ms. Self-healing schemes use fully duplicated transmission systems and capacity for alternate routing of todays TDM or STM circuit facilities. The restoration capacity and the associated transmission systems are essentially unused, except in the rare occasions of network failure.
174
Transmission Systems Design Handbook for Wireless Networks
Although expensive and relatively complex to implement, the dualhomed ring architecture is the choice for high-capacity digital service providers. This architecture uses a drop-and-continue feature that ensures that traffic is available to pass between adjacent rings at two separate nodes or offices. If an entire office is lost, the receiving ring equipment will select traffic from the other office or node. Although this architecture looks expensive, due to network survivability it offers a high potential for cost reduction in the long run. ATM features dynamic bandwidth allocation and ATM switches which together with SONET/SDH transmission will provide support for the emerging broadband multimedia services and existing legacy low-bandwidth telephony and data services. ATM can provide fault detection and traffic rerouting much faster than the existing 50-ms switching time requirement in the SONET/SDH networks. ATM-based capacity management and dynamic reconfiguration have the potential to significantly reduce the transmission facilities required for network survivability, providing economic justification for ATM deployments in large networks. 4.3.7.2 Cluster Topology
The cluster topology is a time-saving, cost-efficient, flexible, reliable, and future-proof way of building and implementing the transmission network for wireless systems. This topology is applicable to large wireless networks. In this context, the transmission network refers to the access network connections from the BSC to the RBSs. The basic idea is to group the RBSs into several independent subnetworks, or clusters. Each cluster has a separate connection to the BSC or to an intercity transfer point. Grouping the RBSs into several clusters creates the initial cluster topology network. The number of clusters and the number of RBSs in each cluster are dependent upon the total number of RBSs and their configuration. Some of the more important factors governing the design are listed below: • The total number of timeslots generated by the RBSs in a cluster
must not exceed the capacity of the link connecting the cluster to the BSC.
• The cluster size must support a flexible, uncomplicated, and effi-
cient network topology inside the cluster. A too large cluster will result in a large and inflexible topology.
Wireless-Network Architecture
175
• In a microwave link network, the number of clusters connected to
one BSC must not be too high, as this will result in high concentration of microwave equipment at the BSC site, which may cause interference problems as well as tower stability problems (too many antennas).
• In general, a cluster size of 10 to 25 RBSs is recommended.
The RBSs in each cluster can be connected to a cross-connect node placed at a central hub-site in each cluster. The cross-connect node consolidates the traffic from the RBSs on the link to the BSC in order to minimize the required link capacity toward the BSC. Figure 4.14 gives an example of a cluster. The topology inside the clusters can be of any type: cascade star, tree, ring, or a combination of these. Each cluster is then directly connected to a hub site in the central cluster, which is either the BSC site or an intercity (backbone) transfer point connecting to the BSC cluster in another city, as shown in Figure 4.15. The RBSs in this central cluster can be connected directly to the BSC or the intercity transfer point. They can also be connected through a cross-connect node to the BSC or the intercity transfer = RBS site = RBS site with = cross-connect = node
To BSC or intercity transfer point
Figure 4.14 RBS cluster.
176
Transmission Systems Design Handbook for Wireless Networks City A (no BSC)
Intercity connection
City B (BSC)
= RBS site = RBS site with = cross-connect = node = BSC
Figure 4.15 Intercity connection.
point. In a microwave link scenario, the links from the clusters toward the central cluster should be hardware protected as 1+1 to guarantee high availability. If required, considering the hop length the links should operate in the lower frequency bands to reduce fading due to rain. Operating in the bands below 10 GHz completely eliminates fading problems caused by rain. In the cluster topology, it is both easy and cost efficient to create a high level of redundancy. Redundancy is achieved by connecting several clusters in ring structures (Figure 4.16). A ring is built simply by adding a link between the cross-connect nodes in two of the outer clusters. All links in a microwave ring should be nonprotected 1+0 links. However, in the initial network the links from the outer clusters to the central cluster were 1+1 protected. Therefore the standby radios from the 1+1 links can be used for the new connections between the outer clusters required to create the ring, but also for expansion in other parts of the network, thus minimizing investment cost for new equipment. As shown in Figure 4.16, there are two possible paths from each cluster. Should the primary path be down due to a link failure, there is always a secondary path to the BSC. In a ring structure, all links in the ring must be able
Wireless-Network Architecture
177
= RBS site = RBS site with = cross-connect = node = BSC Figure 4.16 Redundancy configuration.
to transport traffic from the other cluster(s) in the ring. Therefore, the existing links may have to be upgraded to higher capacity, unless the links were dimensioned for a future ring structure already in the design of the initial network. More complex redundancy schemes can be constructed by connecting rings to each other to form a meshed network. These connections should, as before, be made between the cross-connect nodes in the clusters. The cross-connect nodes in each cluster are also excellent points of connection to a fiber-optic network. The cluster topology is an ideal topology when introducing one or more remote BSCs. A remote BSC is a small and compact stand-alone node that can be introduced to add BSC capacity to high-density traffic areas. A remote BSC should, however, not be introduced initially but at a later stage when the traffic distribution becomes clear and when the need for such a node can be analyzed. The idea is to invest as the network grows. In the cluster topology, a remote BSC is introduced simply by replacing one of the cross-connect nodes with a remote BSC. Furthermore, by introducing a remote BSC in a cluster, the required capacity on the link from the cluster in question to the central cluster (MSC) is substantially reduced. The reduction
178
Transmission Systems Design Handbook for Wireless Networks
is a consequence of the trunking efficiency of the remote BSC. Thus, there is no need for large rearrangement of the transmission network when a remote BSC is introduced. 4.3.7.3 Quality, Performance, and Availability
In todays wireless networks, with converged voice and data, performance degradation may be as dangerous and costly as hardware failures. A degraded transmission network can result in unacceptable signal transmission quality, loss of information, and dropped connections. High availability does not mean just preventing catastrophic failures but also preventing quality and performance degradation. BER, errored seconds (ESs), and one-way delay are usually parameters of interest that will define the quality of the transmission network. Some wireless technologies are more sensitive to delays in the T1/E1 links than others. For example, cdmaOne and CDMA2000 due to the soft handoff have much more stringent requirements on the network delays and synchronization than TDMA-based wireless technologies. High availability of the wireless network is an end-to-end network goalthe network management system (NMS) can help identify critical resources, traffic patterns, and performance levels. NMS can also be used to configure error thresholds, set corporate policies, and provide reports showing end-to-end results. Transmission-network survivability is usually measured by its long-term availability or average network uptime. Most operators expect their network to be continuously available (or at least as little downtime as possible) to minimize potential loss of revenue. The survivable wireless network has an infrastructure of transmission facilities and reliable network elements that are used to manage them. High network availability at the transport level may be achieved using millisecond restoration schemes provided by self-healing network configurations such as SONET/SDH rings or fast facility protection (FFP). DACS in combination with SONET/SDH ring configurations will ensure network availability and survivability. An FFP network comprises two physically diverse routes with identical transmission systems (route diversity). Each route carries half of the working traffic and half of the restoration traffic. The restoration traffic on each route is the duplicate of working traffic on the other route. If the media on these routes are different (i.e., one is fiber optic and the other is, for example, microwave), we talk about the media diversity. These highly reliable solutions dont come cheap, and in many cases a compromise must be made between the cost of the network and its deployment time and network reliability.
Wireless-Network Architecture
179
Regardless of the transmission-network medium and topology, hardware redundancy is an option when designing the transmission network. Protection types usually used are 1+1, where one card or module serves as a protection for another one, or 1+N, where one card or module protects N other units. Linear 1+1 protection switching means that identical payloads will be transmitted on the working and protect fibers or working and protect frequencies in case of the microwave system. Liner 1+N protection switching assumes that there is one protect fiber or frequency for N working fibers or frequencies. A rule of thumb is that if all the hardware is protected with 1+1 and 1+N configuration, fewer spare parts are needed. In case of hardware failure, protection will kick in and the operator will have sufficient time to order replacement parts from the supplier. A ring configuration could provide protection against hardware failures as well, so additional hardware protection might not be required. This is something that transmission engineers must decide, and it is a decision that will be based not only on technical but also budgetary requirements.
References [1]
Bos, L. and S. Leroy, Toward an All-IP-Based UMTS System Architecture, IEEE Network, January/February 2001, pp. 3645.
[2]
McDysan D., and D. Spohn, ATM Theory and Applications, New York: McGraw Hill, 1998.
[3]
ETSI, UMTS-QoS Concept and Architecture, ETSI TS 123 107 V4.0.0, December 2000.
[4]
Malis, A.G., Reconstructing Transmission Networks Using ATM and DWDM, IEEE Communications Magazine, Vol 37, No. 6, June 1999, pp. 140145.
[5]
Liu, C. et al, Packing Density of Voice Trunking Using AAL2, Globecom 99 General Conference, 1999.
[6]
Eneroth, G., et al, Applying ATM/AAL2 as a Switching Technology in 3G Mobile Access Networks, IEEE Communications Magazine, June 1999, Vol 37, No. 6, pp. 112122.
[7]
McDaysan, D., and D. Spohn, ATM Theory and Applications, New York: McGrawHill, 1998.
[8]
ATM Forum, ATM Trunking AF-VTOA-0113.00, February 2000.
[9]
Roberts, Jim W., Traffic Theory and the Internet, IEEE Communications Magazine, January 2001, pp. 9499.
Using
AAL2
for
Narrowband
Services,
180
Transmission Systems Design Handbook for Wireless Networks
[10] Hac, A., Multimedia Applications Support for Wireless ATM Networks, Upper Saddle River, NJ: Prentice Hall, 2000. [11] Zorzi, M., R.R. Rao, On the Impact of Burst Errors on Wireless ATM, IEEEE Personal Communications, Vol.6, No. 4, August 1999, pp. 6576. [12] Lankl, B., and M. Salerno, ATM Traffic and Its Impact on Radio System Design, Sixth European Conference on Fixed Radio Systems and Networks, Bergen, Norway, June 1998. [13] ITU-T G.804, ATM Cell Mapping into Plesiochronous Digital Hierarchy (PDH), February 1998.
5 Theory and Principles of Fiber-Optic Transmission 5.1 Basics of Fiber-Optic Transmission The three primary components of a fiber-optic link are the optical transmitter, the optical receiver, and the fiber-optic cable. In the transmitter, the input signal modulates the light output from the semiconductor laser diode, which is then focused into the fiber-optic cable. This fiber carries the modulated optical signal to the receiver, which then reconverts the optical signal back into the original electrical (analog or digital) signal. Fiber-optic communications offers several advantages over metallic systems. One of them is that any form of outside electronic, magnetic, or radio frequency interference does not distort the transmitted signals. Therefore, optical cables are completely immune to lightning or high-voltage interference. Furthermore, optical fibers will emit no radiation, which suits them for todays tougher standards for interference, also known as electromagnetic compatibility (EMC). Because optical signals do not require grounding connections, the transmitter and receiver are electrically isolated and free from ground loop problems. With no chance of terminal-to-terminal ground potential shifts, plus safety from sparking and shock, fiber optics is increasingly the choice for many processing applications where safe operations in hazardous or flammable environments is a requirement. Digital computing, telephone, and video broadcast systems require new avenues for improved transmission. The high signal bandwidth of optical 181
182
Transmission Systems Design Handbook for Wireless Networks
fibers means increased channel capacity. Also, longer cable runs require fewer repeaters, because fiber-optic cables have extremely low attenuation rates. This suits them for broadcast and telecommunications use. Compared with conventional coaxial cables with the same signal carrying ability, the smaller diameter and lighter weight of fiber-optic cables mean relatively easier installation, especially in crowded duct areas. A singleconductor fiber-optic cable weighs about 6 lbs per 1,000 ft. A comparable coaxial cable weighs 80 lbs per 1,000 ftabout 13 times more. Electronic bugging depends on electromagnetic monitoring; fiberoptic systems are immune to this technique. They have to be physically tapped to extract data, which decreases signal levels and increases error ratesboth of which are readily detected. Some of the features of fiber-optic systems include the following: • All dielectric: low signal radiation, secure transmission, EFI and EMI
immunity, lightning immunity; • Optical signal: no ground loops, spark hazard, suitability for operation in flammable areas; • Low attenuation: greater distance and fewer repeaters; • Small size: less duct space, fewer additional ducts installed; • High bandwidth: future signal capacity expansion (especially today
with DWDM).
The cross section of an optical fiber is shown in Figure 5.1. Core refers to the light transmission area of the fiber, either glass or plastic. The larger the core, the more light that will be transmitted into the fiber. The function of the
Cladding Core Coating
Figure 5.1 Cross section of optical fiber.
Theory and Principles of Fiber-Optic Transmission
183
cladding is to provide lower refractive indexes at the core interface in order to cause reflection within the core so that light waves are transmitted through the fiber. Coating is a multilayer of plastic applied to preserve fiber strength, absorb shock, and provide extra fiber protection. These buffer coatings are available from 250 µm to 900 µm. Fiber size is commonly referred to by the outer diameter of its core, cladding, and coating. For example, 50/125/250 indicates a fiber with a core of 50 µm, cladding of 125 µm, and a coating of 250 µm. The coating is always removed when joining or connecting fibers. A micron (µm) is equal to one-millionth of a meter, and 25 µm are equal to 1/1,000 of an inch (a sheet of paper is approximately 25 µm thick). Fiber types can be identified by the types of paths (or modes) that the light rays travel within the fiber core. There are two basic types of fiber, multimode and single-mode. Multimode-fiber cores may be either step index or graded index. Step-index multimode fiber derives its name from the sharp steplike difference in the refractive index of the core and cladding. In the more common graded-index multimode fiber, the light rays are also guided down the fiber in multiple pathways. The effect of this grading is that the light rays are speeded up in the outer layers, to match those rays going the shorter pathway directly down the axis. The result is that a graded index fiber equalizes the propagation times of the various modes so that data can be sent over a much longer distance and at higher rates before light pulses start to overlap and become less distinguishable at the receiver end. Graded index fibers are commercially available with core diameters of 50, 62.5, and 100 µm. The core of the single-mode fiber is extremely small, approximately 5 to 10 µm. The single-mode fiber allows only a single light ray or modes to be transmitted down the core, virtually eliminating any distortion due to the light pulses overlapping. The single mode has a higher capacity and capability than either of the two multimode types.
5.2 Design Principles 5.2.1
Bandwidth and Attenuation
When selecting components for a fiber-optic system, there are two optical fiber factors that affect transmission performancebandwidth and attenuation. Bandwidth is the measure of the data-carrying capacity of the fiber. The greater the bandwidth, the greater the information capacity. Bandwidth is expressed in a frequency-distance form (MHz-km) and can be limited by
184
Transmission Systems Design Handbook for Wireless Networks
transmitter, receiver, or dispersion of the optical fiber. Fiber can affect bandwidth due to the fact that different wavelengths travel at different speeds along the fiber. To avoid this chromatic dispersion, lasers with narrow optical bandwidths are used with fiber that has low dispersion. In addition to physical changes to the light pulse that result from frequency or bandwidth limitations, there are also reductions in the level of optical power as the light pulse travels to and through the fiber. This optical power loss or attenuation is expressed in decibels per kilometer at a specified wavelength. Light is an electromagnetic wave, and short wavelengths are in the ultraviolet spectrum, while microwaves, radar, television, and radio operate in the longest wavelength areas. In between the ultraviolet and the microwave spectrums, there are fiber-optic wavelengths, which are in the infrared spectrum. Just as the speed of light slows when traveling in transparent materials, each infrared wavelength is transmitted differently within the fiber; therefore, attenuation, or optical power loss, must be measured in specific wavelengths for each fiber type. Transmission loss or attenuation varies with wavelength ( l). Wavelengths are measured in nanometers (nm)billionths of a meterwhich represent the distance between two cycles of the same wave and are calculated as l=1 f where f is the frequency of the signal. Losses of optical power at the different wavelengths occur in the fiber due to absorption, reflection, and scattering. These occur over distance depending on the specific fiber, its size, purity, and refraction indexes. The amount of optical power loss due to absorption and scattering of optical radiation at a specified wavelength is expressed as an attenuation rate in decibels of optical power per kilometers. Fibers are optimized for operation at certain wavelengths. For example, less than 1 dB/km loss is attainable in 50/125 µm multimode fiber operating at 1,300 nm, and less than 3 dB/km (50% loss) is attainable for the same fiber operating at 850 nm. Two wavelength regions, 850 or 1,300 nm, are the areas most often specified for fiberoptic transmission but lately optical fibers have also been optimized at the 1,550 nm region for single-mode transmission systems. Without protection, an optical fiber is subject to losses of optical power caused by microbending; they are minute fiber deviations caused by lateral forces. Different types of protection for the fiber are available to minimize
Theory and Principles of Fiber-Optic Transmission
185
microbending. Step index fibers are relatively more resistant to microbending losses than graded index. When its freezing cold outside, fiber-optic cable enclosed in innerduct systems that are not buried beneath the frost line (for example, innerduct strung under a bridge) could be at risk. Innerduct is a smaller, usually PVC duct used for the protection of the fiber-optic cables within the larger cable ducts housing a large number of different types of cables. Exposed innerduct could collect water which if ice can exert crushing pressure on the fiber-optic cable inside the conduit. That pressure can cause microbends in the fiber and degrade the signal, especially at higher speeds, or it can even break the fiber. This freezing phenomenon was discovered in the early 1990s by a large telephone company, which noticed that its fiber-optic network experienced problems on cold nights in certain parts of the network. As it warmed up, the signal would be restored to its usual quality.
5.2.2
Optical Power Budgets and Distance Calculations
The key to network distance is the optical power budgetthe amount of light available to make a fiber-optic connection. This chapter will explain how to determine the maximum fiber-optic distances attainable using media converters in various network environments. A simple calculation is used to determine how much fiber-optic light, measured in dBm, is available. It tells us how many decibels over 1 mW of electrical, or in this case optical, power are available. The first step in calculating the optical power budget is determining how much light is available for the electronic devices themselves. Two measurements are needed from the manufacturer of the equipmentminimum transmit power and minimum receive sensitivity. Minimum transmit power represents the worst-case transmit power for a device guaranteed to provide at least that much power. It is important to note that some vendors will list an average minimum transmit power. Average minimum transmit power does not guarantee that a product will perform at that products minimum transmit power. The second piece of information required is the minimum receive sensitivity. This figure represents the minimum amount of light required by the receiver to operate correctly. Again, the actual minimum should be used, not an average of minimums. With minimum transmit power and minimum receive sensitivity data, it is possible now to calculate the available light as a worst-case scenario. In reality, the available light will probably be higher than this calculated value:
186
Transmission Systems Design Handbook for Wireless Networks
available light = minimum transmit power − minimum receive sensitivity Minimum receive sensitivity is usually a negative number, such as −33 dBm. Subtracting a negative number is the same as adding its absolute value. For example, if a device has a minimum transmit power of −10 dBm and a minimum receive sensitivity of −33 dBm, the available power will be
(−10 dB) − (−33 dB) = 23 dB − 10 dB + 33 dB = 23 dB Also, when connecting devices from different vendors, or different models of products from the same vendor, the available power calculation needs to be computed in both directions with the smaller of the two used for the rest of the calculations. For example, assume connecting two devices labeled Device 1 and Device 2. Device 1 has a minimum transmit power of −3 dBm and a minimum receive sensitivity of −32 dBm, and Device 2 has a minimum transmit power of −1dBm and a receive sensitivity of −31 dBm. The available power going from Device 1 to Device 2 would be calculated by −3 dBm − (−31 dBm) or 28 dB. The available power going from Device 2 to Device 1 would be calculated by −1 dBm − (−32 dBm) or 29 dB. There is less light available in the Device-1-to-Device-2 direction, so we will use that figure for our calculations; if that half of the link works, then so will the other (worst-case scenario). See Table 5.1.
Table 5.1 Fiber-Optic Link Budget Device 1
Device 2
Min Tx Power
−3 dBm
−1 dBm
Min Rx Sensitivity
−32 dBm
−31 dBm
Budget 1
Budget 2
Device 1 Tx
−3 dBm
−1 dBm
Device 2 Rx
−31 dBm
−32 dBm
Available Power
28 dB
29 dB
Theory and Principles of Fiber-Optic Transmission
187
From available power we must subtract out all of the losses including cable attenuation, connector losses, and splice losses. Cable attenuation is often the most significant loss factor and should be determined by getting the exact number off of the cable being installed or using the manufacturers worst-case number of the type of fiber you planned to install. This number will range from 0.22 to 0.5 dB/km. Multiply this factor by the number of kilometers in the installation. A fiber with 0.4 dB/km of loss will lose 16 dB over a 40-km distance. Also, it is important to remember that fiber does not come in 40-km spools; therefore a 40-km installation will have several splices. Each splice will typically introduce 0.1db of additional loss, and the fiber installers should be able to provide a guaranteed worst-case number. The number of splices should be multiplied by this number. Connectors are another source of light loss. A typical long-haul installation will have six connectors in the installation. The first connects the fiber to electronics. This connector is usually on indoor fiber. This fiber connects the equipment room to the building entrance for the outdoor (buried or aerial) fiber. There is another connector on this end of the indoor cable and one on the outdoor cable. This is repeated at the other end of the network for a total of six connectors. Individual networks can vary, however, and the exact number must be determined. The connector manufacturer or the installer provide connector loss. Multiply the number of connectors by the loss for each connector to get total connector loss. Each of these lossescable attenuation, connector loss, and splice lossis then subtracted from the available power. If this number is negative, there is no need to continue, as there is not enough power to drive the network. If this number is positive, there are two more things to considerfirst, what happens if the fiber gets cut and has to be spliced back together? A proper design will count on this happening and account for it in the power budget; also, an estimation of the number of anticipated repairs over the life of the fiber needs to be made. These repairs will add splice loss, so we must multiply the number of anticipated splices by the loss of each splice (same number we used above), and subtract this from the remaining power. The number should still be positive. Finally, we must account for temperature extremes, as well as any other unforeseen factors. Determining a safety factor (margin) typically does this. This number will be different for every organization, depending on how much risk they want to assume in their network. Typically a value around 3 dB is used. To guarantee error-free operation, a value no less then 1.7 dB should be used. This safety factor is subtracted from the remaining power
188
Transmission Systems Design Handbook for Wireless Networks
from above. If the number is still positive after all of this, we can be assured that the fiber network will deliver the required performance over the life of the installation. Table 5.2 contains some typical numbers that can be used to approximate optical-link budget calculations. If at all possible, real numbers from the network in question should be used. The simple worksheet presented in Table 5.3a can help in these calculations. Table 5.3b is an example that plugs the numbers from the previous discussion into the worksheet presented in Table 5.3a. Table 5.2 Fiber-Optic Losses TIA standard for connector loss
0.75 dB
Typical cable attenuation at 850 nm (MM)
3.0 dB
Typical cable attenuation at 1,310 nm (MM)
1.0 dB
Typical cable attenuation at 1,310 nm (SM)
0.4 dB
Typical cable attenuation at 1,550 nm (SM)
0.2 dB
Typical distance between splices
6.0 km
Typical safety margin
3.0 dB
Typical splice attenuation (usually < 0.05 dB for SM and < 0.1 for MM)
0.1 dB
Table 5.3a Optical Budget Calculator Minimum Transmit Power
________
Minimum Receive Sensitivity
________
Available Power
________
________ km of cable
×
_______ dB/km
________
________ connectors
×
_______ dB/con
________
________ splices
×
_______ dB/splice
________
Link Margin _______ repair splices ×
________
_______ dB/splice
________
Safety Margin
________
Excess Power
________
Theory and Principles of Fiber-Optic Transmission
189
Table 5.3b Optical Budget CalculatorExample Minimum Transmit Power
10 dBm
Minimum Receive Sensitivity
33 dBm
Available Power
23 dB
×
____0.5 dB/km
_____10.0
_______6 connectors
×
____0.75 dB/con
_____x4.5
_______4 splices
×
____0.1 dB/splice
______0.4
_____ 20 km of cable
Link Margin ______ 5 repair splices ×
______8.1
_____ 0.1 dB/splice
_____ 0.5
Safety Margin
______3
Excess Power
______4.6
5.3 Synchronous Digital Hierarchy 5.3.1
Basics of Synchronous Systems
The fiber-optic media have proven to be the best high-speed transmission media where electrical and physical environmental factors are more considerable for enterprise network infrastructure design. It may primarily resolve some of the networking issues, but still durability and survivability of linear fiber-optic implementation raise more questions regarding cable cuts, disaster recovery, and so on. Since initial fiber installation followed a linear route between endpoints and the back-up fiber was installed with same right of way (meaning that the main and back-up fiber are probably placed very close to each other and possibly even in the same cable or same cable duct), definitely creates serious restoration problem in case of a natural or man made disaster. Usually, the prediction and prevention of such an undesired condition are unavoidable. A fiber-optic cut may involve the primary and the back-up cable in a linear route. In addition to the actual cut itself, glass fibers may experience stress fractures several hundred feet in either direction from the cut. Restoration under such circumstances may take a few days or weeks to complete. The above factors were not alone sufficient to drive SONET as a significant
190
Transmission Systems Design Handbook for Wireless Networks
transmission technology. The clients increasing demand for bandwidth, error-free delivery, and related requirements were too much for the older PDH solution, so the decision was made to use SONET for transmission means. Above all, PDH had the following limitations: • Lack of performance; • Incompatibility of different vendors equipment; • Optical interfaces that were not universally defined; • No self-checking; • No standard for high-bandwidth links; • Not synchronous above DS1.
To find a solution for PDH problems, we need an error-free, synchronous, and high-speed technology. In 1984 Bell Communications introduced SONET, a high-speed fiber-optic system that provides an interface and mechanism for optical transmission of digital information with nearly errorfree conditions. Later, SONET was quickly accepted by ANSI. In 1988 the Comite Consultatif Internationale de Telegraphique et Telephonique (CCITT) published a similar set of standards called SDH, currently used in Europe and many other parts of the world. SONET/SDH are true implementations of fiber-optic media, and because they are typically WAN standards, SONET/SDH use a point-to-point connection type. They use time-division multiplexing over mesh and ring physical topology. One of the principal purposes of SONET is to improve the durability and survivability of fiber-optic networks. SONET/SDH are physical transmission mechanisms for various high-end implementations, such as FDDI, ATM, and SMDS. SONET is not a network, but is used as an infrastructure transmission means for the foundation of a high-quality network. It is a set of standards that define the rates and formats for optical networks as specified in ANSI T1.105, ANSI T1.106, and ANSI T1.117. Both SONET and SDH are now the industry standards for voice, video, and data optical transmission. Three key requirements have driven the development of SONET. The first was to push the multiplexing standards beyond 44.736 Mbps that is DS-3 level. A number of vendors have introduced their own proprietary schemes for combining multiple DS-3s into an optical signal. The second requirement is to provide economic access to small amounts of traffic within the bulk payload
Theory and Principles of Fiber-Optic Transmission
191
of a SONET signal. The third is to provide bandwidth on demand and virtual private networking with a sophisticated optical error-free mechanism. Almost all new fiber-optic transmission systems, as well as highcapacity microwave systems, now being installed in public or private networks use SDH or SONET. Bit rates in long-haul systems are expected to rise to over 40 Gbps over the next few years and at the same time, systems of 155 Mbps and below penetrate more deeply into access transmission networks. New 3G wireless networks will definitely have a big impact on the urgency of adding more and more capacity into the existing transmission networks. 5.3.2
Benefits of SONET
Some of the benefits of using SONET systems in the transmission-network design include the following: • Performance. SONET provides excellent performance with almost •
• •
•
•
5.3.3
100% error-free transmission through the optical medium. Integration. The SONET services combine bandwidth and multiplexing capabilities to let users fully integrate voice, video, and data over a single means of transmission facility. Security. SONET maximizes security and eliminates contention by containing the traffic on the dedicated private lines. Disaster recovery. SONET is designed to provide survivability and reliability to self-heal within 50 ms in case of any physical medium failure. High-speed transmission. The SONET/SDH supports up to 9.9 Gbps transfer rate. In the DS-level hierarchy, SONETs basic bandwidth is a DS-3 (44.736 Mbps) channel plus overhead. Stability and robustness. Todays SONET architecture offers more stable and robust communication that also eliminates multiple mapping and demapping multiplexers in the long distribution circuit.
SONET Architecture
5.3.3.1 Transmission Layers
At the interface, SONET converts digital signals from electrical form to optical form and back to electrical form at the destination. It supports a transmission range from 51.84 Mbps to multiples of DS3s, such as OC-248 (9.9
192
Transmission Systems Design Handbook for Wireless Networks
Gbps) and the multiplexing of lower-capacity channels, down to the 64Kbps DS0 channel. The SONET standard defines four layers to deal with the task involved in getting transmission from one end point to another. It can adjust timing and framing during operation with additional support to drop and insert capabilities that make it easier to identify and remove channels going to different destinations. SONET can be used as carrier service for ATM, SMDS, FDDI, and other broadband ISDN networks. It uses an 810-byte frame as its basic transmission and transmits 80,000 frames/s. The four layers of SONET are photonic, section, line, and path (see Figure 5.2). Photonic is the physical layer of SONET that deals with cable, signal, and component specification, such as the optical fibers minimum power requirements and the dispersion characteristics of transmitting lasers. Signals are converted between electrical and optical form in this layer. In the section layer, basic frames are created and scrambled whenever appropriate. It converts electrical signals to photonic ones and monitors for errors. The line layer is responsible for getting frames from one end of a line to the other. The synchronization and multiplexing of data onto the SONET frames are done in this layer, so timing adjustments and adding and dropping are done at this level. This layer is also responsible for protection, maintenance functions, and switching. The path layer completes the transmission (e.g., getting from one end to another with overall path); this layer provides transport of data at the appropriate signaling speed. At the end-point terminal, a signal is converted to synchronous transmission signal (STS), which is the building block of the SONET optical interface, with a rate of 51.84 Mbps. An STS signal travels through various SONET networks in STS format until it terminates. The terminating Path layer Line layer Section layer Frame Light Photonic layer
Terminal
Envelope Section Photonic
Line Section Photonic
Path layer Line layer Section layer Photonic layer
Regenerator
STS mulitplexer
Terminal
STS-N block
Figure 5.2 SONET transport layers.
Theory and Principles of Fiber-Optic Transmission
193
equipment converts the STS to the user format. The STS consists of two parts: the payload and the overhead. The payload of STS carries the information portion of the signal, and the overhead carries the signaling as well as protocol information. These allow communication between intelligent nodes in the network, permitting operation, administration, maintenance, and provisioning (OAM&P) of a network from a central location. The end points are the source and destination for the DS-3 or smaller channel that makes up the SONET transmission. 5.3.3.2 Basic Frame Format
The functionality of SONET is achieved by defining the basic STS-1 signal and an associated byte-interleaved multiplex structure that creates a group of standard rates at N times the STS-1 rate (see Table 5.4). The frame structure must be defined describing how bits are assembled into standard units fit for transmission. N can be any integer value from 1 to 255. Today, the following values for N are available and defined as N = 1, 3, 9, 12, 18, 24, 36, 48, 96, 192, and 255. The transmission over fiber-optic facilities also describes an optical counterpart of the STS-1 signal called optical carrier level 1 (OC-1). Three STS-1 channels can be combined into one 155.52-Mbps (STS-3) channel by merging the frames, producing frames that are three times larger. Similarly, the STS-48 is 48 times the basic SONET rate of 2,488.32 Mbps. Table 5.4 High Data Rates STS Level
OC Level
STM Level Bit Rate (Mbps) DS3
001
001
51.84
001
003
003
01
155.52
003
006
006
311.04
006
009
009
466.56
009
012
012
04
622.08
012
018
018
933.12
018
024
024
1,244.16
024
036
036
1,866.24
036
048
048
16
2,488.32
048
096
096
4,976.64
096
192
192
9,953.28
192
255
255
13,219.20
255
194
Transmission Systems Design Handbook for Wireless Networks
The OC-1 signal forms the basic SONET transmission building block from which higher-level signals, such as OC-3, OC-48, and so on, are derived. For example, OC-48 systems carry 48 DS-3s (1,344 T1s). In the international arena, STS-3 equates with STM-1; thus, STS-48 equates with STM-16. It is possible to examine the SONET rate by applying the frame-format attributes to see how an STS-1 signal is formed. So, the STS-N will be multiples of STS-1 replaced by the value of N. A SONET frame is 810-bytes long and transmitted every 125 ms at 8,000 frames/s. Each frame is a coordination of 9 rows by 90 columns. In other words, the 810-byte frames are grouped into nine 90-byte portions that are then transmitted one after another. Therefore, the STS-1 line rate can be derived as follows: 9 rows × 90 columns × 8,000 frames/s × 8 bits/byte = 51.84 Mbps Three bytes, or the first three columns in each row, are the transmission overhead; the remaining 90 − 3 = 87 bytes are data or payload. The overhead in the first three rows is allocated for the monitor section; in the remaining six rows, it is for the line overhead. So, the STS-1 frame format can be summarized as follows: • STS-1 frame is the basic building block of SONET. • It consists of 810 bytes, 9 rows, and 90 columns. • There are 27 bytes of overhead formed from the first three octets of
each row, 9 used for section overhead, 18 for line overhead.
• Payload is 87 × 9 = 783 bytes. One column of the payload is the
path overhead, positioned by a pointer in the line overhead.
• Transmission is top to bottom, row by row, from left to right. • STS-1 frame is transmitted every 125 ms.
Column 4 to column 90 contain the SPE. One column in the SPE is used for path overhead. Both the section and line overhead include channels for communicating. These channels are used to send alarms and other administrative information. The line overhead includes several bytes for pointer. These allow the SPE to be moved. A good source of information on SONET and SDH vendors and analysis of key products available on the market can be found in [1].
Theory and Principles of Fiber-Optic Transmission 5.3.4
195
SONET Availability Requirements
In the SONET network design, the end-to-end availability objective for a DS1 channel transported along the 250-mile short-haul interoffice route is 99.98%. This is the equivalent of 0.02% unavailability or 105 minutes per year. This 0.02% unavailability along the route is due to unavailability of all facilities, including the terminal and repeater equipment, fiber-optic cable, connectors, splices, software failures, and procedural errors. This objective is linearly prorated with distance [2, 3]. The allocation to hardware reliability failures is 15 min/yr for a 250-km route and prorated for shorter distances. 5.3.5
SDH
ETSI defined the SDH for Europe. It is now used everywhere outside North America and Japan. SDH has provided transmission networks with a vendor-independent signal structure that has resulted in new network applications, the deployment of new equipment in new network topologies, and management by operations systems of much greater power than previously seen in transmission networks. The development of optical-fiber transmission systems made more complex standards possible. It was widely accepted that the new multiplexing method should be synchronous and based on byte interleaving and not on bit interleaving, as was the PDH. SONET is an ANSI standard; it can carry as payloads the North American PDH hierarchy of bit rates1.5/6/45 Mbps, plus 2 Mbps. SDH embraces most of SONET and is an international standard, but it is often regarded as a European standard because its suppliers, with few exceptions, carry only the ETSI-defined European PDH bit rates of 2/34/140 Mbps (8 Mbps is omitted from SDH). The flexibility of SDH can be used to best advantage by introducing new network topologies. Traditional networks make use of mesh and hub arrangements, but SDH, with the help of DACSs and hub multiplexers, allows these to be used in a much more comprehensive way. SDH enables these topologies to be combined with rings and chains of ADMs to improve flexibility and reliability across the core and access parts of the network. SDH defines traffic interfaces that are independent of vendors. At 155 Mbps they are defined for both optical and copper interfaces and at higher rates for optical ones only. These higher rates are defined as integer multiples of 155.52 Mbps in an N × 4 sequence, giving, for example, 622.08 Mbps (622 Mbps) and 2,488.32 Mbps (2.5 Gbps). To support network growth and the demand for broadband services, multiplexing to even higher rates is being
196
Transmission Systems Design Handbook for Wireless Networks
developed with upper limits set by technology rather than by lack of standards, as was the case with PDH. Each interface rate contains overheads to support a range of facilities and a payload capacity for traffic, and both of them can be fully or partially filled. Rates below 155 Mbps can be supported by using a 155-Mbps interface with only a partially filled payload area. An example of this is a microwave radio system whose spectrum allocation limits it to a capacity less than the full SDH payload, but whose terminal traffic ports are to be connected to 155-Mbps ports on a cross connect. Interfaces are sometimes available at a lower synchronous rate for access applications. North America has for some time used 51.84-Mbps SONET, and ETSI has defined a 34-Mbps SDH interface whose data rate is identical to that of a 34-Mbps PDH system. To support a range of operations, SDH also includes a management layer whose communications are transported within dedicated datacommunications-channel (DCC) timeslots inside the interface rate. They have a standard profile for the structure on network management messages, irrespective of vendor or operator. However, there has been very little agreement on the definition of the message sets to be carried, so there is very little interworking of management channels between equipment vendors at the SDH interface. More information on the management functions of the SDH can be found in [4].
5.4 DWDM 5.4.1
DWDM Overview
DWDM is a fiber-optic transmission technique that employs light wavelengths to transmit data parallel-by-bit or serial-by-character and can, at present, increase the bandwidth of the existing fiber-optic facilities up to 32 times. There is a huge importance of scalable DWDM systems in enabling service providers to accommodate consumer demand for ever increasing amounts of bandwidth. DWDM is becoming a crucial component of optical networks for the transmission of e-mail, video, multimedia, data, and voicecarried in IP, ATM, and SONET/SDH, respectively, over the optical layer. In order to squeeze more bandwidth out of their fiber networks, longhaul carriers are deploying dense-wavelength-division multiplexers to build backbones that might have dozens of channels riding on a single strand of fiber, with each channel operating at multigigabit speeds. Now the DWDM equipment vendors are moving their technology away from the WAN and positioning it in the metropolitan area network (MAN). The target customers
Theory and Principles of Fiber-Optic Transmission
197
for metro DWDM technology are the incumbent local exchange carriers (ILECs) and competitive local exchange carriers (CLECs), cable companies, wireless-network operators, and even enterprise customers with dark fiber. 5.4.2
DWDM Capacity (Bandwidth) Requirements
To understand the importance of DWDM and optical networking, these capabilities must be discussed in the context of the challenges faced by the telecommunications industry, and in particular, service providers. Most U.S. networks were built using estimates that calculated bandwidth use by employing concentration ratios derived from classical engineering formulas, such as Poisson and Reeling. Consequently, forecasts of the amount of bandwidth capacity needed for networks were calculated on the presumption that a given individual would only use network bandwidth for six minutes out of each hour. These formulas did not factor in the amount of traffic generated by Internet access (300% growth per year), faxes, multiple phone lines, modems, teleconferencing, Internet, and data and video transmission. Had these factors been included, a far different estimate would have emerged. In fact, today many people use the bandwidth equivalent of 180 minutes or more each hour. No one could have predicted the network growth necessary to meet the demand. For example, one study estimated that from 1994 to 1998 the demand on the U.S. IXCs network would increase sevenfold, and for the U.S. LEC network, the demand would increase fourfold. In actuality, one company indicated that its network growth was 32 times that of the previous year, while another companys rate of growth in 1997 alone equaled the same size of its entire network in 1991. Another operator has said that the size of its network doubled every six months in that 4-year period. In addition to this explosion in consumer demand for bandwidth, many service providers are dealing with fiber exhaust in their networks. Today, many carriers are nearing 100% capacity utilization across significant portions of their networks. Another problem for carriers is the challenge of deploying and integrating diverse technologies in one physical infrastructure. Customer demands and competitive pressures mandate that carriers offer diverse services economically and deploy them over the embedded network. DWDM provides service providers with an answer to that demand (Figure 5.3). Use of DWDM allows providers to offer such services as e-mail, video, and multimedia carried as IP data over ATM and voice carried over SONET/SDH. Despite the fact that these formats (IP, ATM, and
198
Transmission Systems Design Handbook for Wireless Networks
DWDM/point-to-point optical transport
Fixed DWDM/multipoint network
l1 l2 lN
l1 l2 lN
...
... l1 l1
Optical XC reconfigurable DWDM/multipoint network
Capacity expansion
Bandwidth management
lk lk
oxc
Figure 5.3 Fiber-optic capacity expansion and bandwidth management.
SONET/SDH) provide unique bandwidth-management capabilities, all three can be transported over the optical layer using DWDM. This unifying capability allows the service provider the flexibility to respond to customer demands over one network. DWDM can be deployed in unidirectional and bidirectional applications [5]. Unidirectional applications use two fibers to send and receive traffic separately in the opposite direction. Bidirectional applications, on the other hand, use a single fiber to carry traffic in both directions. Bidirectional transmission, which typically supports fewer wavelengths than unidirectional transmission, is a good solution only if there is no need for upgrading to large capacity of hundreds of gigabytes per second in the network. 5.4.3
Network Growth and Flexibility of DWDM
Faced with the challenges of increased service needs, fiber exhaust, and layered bandwidth management, service providers need options to provide an economical solution. One way to alleviate fiber exhaust is to lay more fiber, and for those networks where the cost of laying new fiber is minimal, this will prove the most economical solution. However, laying new fiber will not necessarily enable the service provider to provide new services or utilize the bandwidth-management capability of a unifying optical layer.
Theory and Principles of Fiber-Optic Transmission
199
A second choice is to increase the bit rate using TDM, where TDM increases the capacity of a fiber by slicing time into smaller intervals so that more bits (data) can be transmitted per second. Traditionally, this has been the industry method of choice (DS1, DS2, DS3, and so on). However, when service providers use this approach exclusively, they must make the leap to the higher bit rate in one step, having purchased more capacity than they initially need. Based on the SONET hierarchy, the next incremental step from 10-Gbps TDM is 40 Gbps, which is a huge leap that many believe will not be possible for TDM technology in the near future. This method has also been used with transmission networks that are based on either the SONET standard for North America or the SDH standard for international networks. The third choice for service providers is DWDM, which increases the capacity of embedded fiber by first assigning incoming optical signals to specific frequencies (wavelength λ) within a designated frequency band and then multiplexing the resulting signals out onto one fiber. Because incoming signals are never terminated in the optical layer, the interface can be bit-rateand format-independent, allowing the service provider to integrate the DWDM technology easily with existing equipment in the network while gaining access to the untapped capacity in the embedded fiber. DWDM combines multiple optical signals so that they can be amplified as a group and transported over a single fiber to increase capacity. Each signal carried can be at a different rate (OC-3/12/24, etc.) and in a different format (SONET, ATM, data, etc.). For example, a DWDM network, with a mix of SONET signals operating at OC-48 (2.5 Gbps) and OC-192 (10 Gbps) over a DWDM infrastructure, can achieve capacities of over 40 Gbps. A system with DWDM can achieve all this while maintaining the same degree of system performance, reliability, and robustness as current transmission systemsor even surpassing them. Future DWDM terminals will carry up to 80 wavelengths of OC-48, a total of 200 Gbps, or up to 40 wavelengths of OC-192, a total of 400 Gbps. By beginning with DWDM, service providers can establish a grow-asyou-go infrastructure, which allows them to add current and next-generation TDM systems for virtually endless capacity expansion. Carriers can address specific problem areas that are congested because of high-capacity demands. This is especially helpful where multiple rings intersect between two nodes, resulting in fiber exhaust. Service providers searching for new and creative ways to generate revenue while fully meeting the varying needs of their customers can benefit from a DWDM infrastructure as well. By partitioning
200
Transmission Systems Design Handbook for Wireless Networks
and maintaining different dedicated wavelengths for different customers, for example, service providers can lease individual wavelengths. Furthermore, DWDM allows service providers to increase capacity on a broad range of wavelengths in the 1.55-µm region. For example, with a DWDM system multiplexing up to 16 wavelengths on a single fiber, carriers can decrease the number of amplifiers by a factor of 16 at each regenerator site. Using fewer regenerators in long-distance networks results in fewer interruptions and improved efficiency. 5.4.4
Optical Layers
Aside from the enormous capacity gained through optical networking, the optical layer provides the only means for carriers to integrate the diverse technologies of their existing networks into one physical infrastructure. DWDM systems are bit-rate- and format-independent and can accept any combination of interface rates (e.g., synchronous, asynchronous, OC-3, -12, -48, or -192) on the same fiber at the same time. If a carrier operates both ATM and SONET networks, the ATM signal does not have to be multiplexed up to the SONET rate to be carried on the DWDM network. Since the optical layer carries signals as-is without any additional multiplexing, carriers can quickly introduce ATM or IP without deploying an overlay network. But DWDM is just the first step on the road to full optical networking and the realization of the optical layer. The concept of an all-optical network implies that the service provider will have optical access to traffic at various nodes in the network, much like the SONET layer for SONET traffic. Optical wavelength add-drop (OWAD) offers that capability, where wavelengths are added or dropped to or from a fiber, without requiring a SONET terminal. But ultimate bandwidth management flexibility will come with a crossconnect capability on the optical layer. Combined with OWAD and DWDM, the optical cross connect (OXC) will offer service providers the ability to create a flexible, high-capacity, efficient optical network with full optical bandwidth management. These technologies are todays reality: DWDM has been utilized in the long-distance network since 1995; OWAD products became available in 1998; and the first all-OXC was showcased at industry conventions in 1997. 5.4.5
Protection in DWDM Networks
The basic types of network failures generally considered are link and node failures. Link failure usually occurs because of cable cuts, while node failure is
Theory and Principles of Fiber-Optic Transmission
201
due to equipment failure at network nodes. Besides node and link failures, which are common failure situations in any communication network, channel failure is also possible in DWDM optical networks [6]. Offering protection on a per-channel basis has become an important factor as equipment providers roll out their metro DWDM systems. A number of equipment suppliers already offer wavelength protection on a per-channel basis as part of their metro platforms. Per-channel protection is becoming increasingly important because carriers want to cap their investments in legacy SONET/SDH equipment but still protect higher layers of traffic. Because SONET/SDH already offers protection, the ability to turn DWDM protection on and off depending on the traffic type is essential. The reason behind per-channel protection is dualif one channel suffers a failure and if you dont have protection on a per-channel basis, all channels suffer a 50-ms switch. If half of the channels already have SONET/SDH protection, it is not necessary to provide duplicate protection.
5.5 Optical Switching Optical networking provides the backbone to support existing and emerging technologies with almost limitless amounts of bandwidth capacity. Most current networks employ electronic processing and use the optical fiber only as a transmission medium [7]. All-optical networking (not just point-to-point transmission) enabled by optical cross connects, optical programmable adddrop multiplexers, wavelength routers, and optical switches provides a unified infrastructure capable of meeting the telecommunications demands of today and tomorrow. DWDM is a key technology in long-haul networks, because it gives carriers the ability to multiply the capacity of fibers by factors of 16, 32, and even higher. Service providers are moving from leasing dark fiber (fiber without terminal equipment), the physical medium, to leasing light paths, an optical end-to-end path through the network. Carriers are starting to offer individual wavelengths providing clear channel OC-48c capacity to their wholesale customers. The service provides a transparent interface with a customers network for completion of SONET rings, construction of new or diverse routes, and transmission of ATM or IP traffic. The digital cross connects are able to switch a large number of circuits, but they have to convert optical signals to electrical, switch them, and then convert them back into wavelengths of light and send them out. New optical cross connects do the switching on the optical level without converting optical signals into the electrical signals. Miniature mirrors with
202
Transmission Systems Design Handbook for Wireless Networks
surfaces no bigger than the diameter of hair are bouncing waves off the surface whose angles are changed by the small changes in temperature. Optical switches aim for the ability to provision bandwidth intelligently and switch and route wavelengths based on demand, the way standard Class 5 switches do today, but in the optical layeron a wavelength level rather than an individual voice channel. DWDM in combination with optical switching will provide a huge increase in the capacity of the new and existing transmission systems. It is probably the only available solution for OC-768/STM-256 (40 Gbps), OC-1920/STM-640 (100 Gbps), and OC-3840/STM-1234 (200 Gbps) systems over one fiber pair. Also, a wavelength-routing device can route signals arriving at different input fibers (ports) of the device to the different output fibers (ports) based on the wavelengths of the signals. A new optical internetworking initiative, the Optical Domain Service Interconnect (ODSI), is building interoperability between electrical and optical systems and among optical vendors, as well. ODSI is attempting to develop an interface that will allow electrical network elements, such as routers, ATM switches, and cross connects, to provision services, such as setting up and tearing down, on optical networks, on demand.
References [1]
Goralski, W., SONET, Second Edition, New York: McGraw-Hill, 2000.
[2]
Lewin, B., SONET Equipment Availability Requirements, IEEE, 1989.
[3]
Bellcore TR-TSY-000418, Generic Reliability Assurance Requirements for Fiber-Optic Transport Systems, Issue 2, September 1989.
[4]
ITU-T Recommendation G.784, Synchronous Digital Hierarchy (SDH) Management.
[5]
Fujitsu, Wave Division Multiplexing, white paper, April 1998.
[6]
Zhou, D., and S. Subramanim, Survivability in Optical Networks, IEEE Network, November/December 2000, pp 1623.
[7]
Sivalingam, K. M., and S. Subramaniam, eds., Optical WDM Networks: Principles and Practice, Norwell, MA: Kluwer Academic Publishers, 2000.
6 Microwave Point-to-Point System Design 6.1 Basic Microwave Transmission Theory RF signals can be transmitted over a wide range of frequencies, expressed in cycles per second, or hertz (Hz). Furthermore, 1 kilohertz (kHz) equals 1 thousand Hz, 1 megahertz (MHz) equals 1 million Hz, and 1 gigahertz (GHz) equals 1 billion Hz. Amplitude modulation (AM) radio signals are at the lower end of the RF spectrum, while other radio services, such as analog and digital TV (DTV), cellular and PCS telephony, and point-to-point microwave (MW) services are much higher in frequency. Free-space loss is defined as the loss between two isotropic antennas in free space in the absence of ground or atmospheric influences or obstructions; in other words, where fades due to refraction, reflection, and diffraction fade activity and atmospheric absorption do not occur. Radio energy is lost in free space because of the triangular spreading of energy in the wavefront as it travels through space in accordance with the inverse-square law (wavefront area quadruples for each doubling of distance for a 10 log 1/4 = 6 dB increase in path loss). The derivation of the free-space loss formula is presented in other references and will not be repeated here. The formula itself, however, is as follows: A = 96.6 + 20 log f + 20 log D 203
204
Transmission Systems Design Handbook for Wireless Networks
where A = free-space attenuation between antennas (dB); f = frequency (GHz); D = path length (miles); or A = 92.4 + 20 log f + 20 log D where A = free-space attenuation between antennas (dB); f = frequency (GHz); D = path length (km). The unfaded receive-signal level (RSL) is computed by introducing the net path loss (NPL) in decibels between a radio transmitter output and the far-end radio receivers input. The microwave radio transmit power is reduced by the NPL as determined by free space and atmospheric absorption losses, antenna gains, and waveguide feeder and network losses at both ends of the single radio link (or hop, as it is commonly referred to). Unfaded RSL at the distant receivers RF input is RSL = Pt − NPL = Pt + G t − L al + G r − L f − L mis where RSL = unfaded RSL (dBm); Pt = transmit power (dBm); Gt = transmit antenna gain (dB); Lfs = free-space loss (dB); Lal = absorption loss (dB); Gr = receive-antenna gain (dB); Lf = coax or waveguide feeder loss (dB); Lmis = network and miscellaneous losses (dB). The propagation of radio waves is generally affected by several factors, irrespective of the radio communication service or the specified purpose of telecommunication. These factors are described below.
Microwave Point-to-Point System Design
205
Frequency Effects
The relative importance of the factors influencing the propagation of radio waves mainly depends on the frequency band. In the low-frequency (LF) and medium-frequency (MF) bands the propagation is strongly influenced by the electrical characteristics of the ground and by the ionospheric conditions. In the VHF and UHF bands the effects from the ionospheric conditions decrease to such an extent that the terrain features and, to some degree, the meteorological characteristics of the troposphere (the lower parts of Earths atmosphere) begin to dominate. At microwave frequencies, the importance of the terrain features and the meteorological characteristics of the troposphere are still predominant. However, above about 6 GHz the effects of gas absorption and precipitation must also be taken into account. At frequencies close to 10 GHz the effects of precipitation begin to dominate. Gas absorption starts influencing at about 22 GHz, where the water vapor shows a characteristic peak. Radio waves theoretically travel in a straight line in a vacuum. However, due to nonhomogeneous pockets of warm and cold air, in actuality waves can and do bend much the same as a light wave propagates through a prism. Long waves will bounce from the ionosphere above the Earth and travel back to Earth, thereby allowing an electromagnetic signal to be transmitted over a very long distance to a remote receiver. Microwaves and other very short waves (for example, 15, 23, and 38 GHz) do not bounce off the ionosphere and suffer much higher attenuation as they travel through the air. In addition, millimeter waves are subjected to rain losses (above 10 GHz), which can impact availability of the radio link if there is no sufficient fade margin to handle the rain-fading effect. Terrain Effects
When radio waves propagate near the surface of the Earth, their characteristics are dominated by the electrical characteristics of the Earth and by the topography of the terrain, including the vegetation and man-made structures. Tropospheric Effects
The gaseous constituents of the atmosphere influence the propagation of radio waves both by absorbing energy and by variations in the refractive index. Variations in the refractive index of the atmosphere cause radio waves to reflect, to refract, and to scatter. The magnitude of these effects depends, of course, on the frequency.
206
Transmission Systems Design Handbook for Wireless Networks
Multipath Effects
The term multipath effects applies to those cases in which the effective received signal is made up of several components arriving at the receiving antenna over different paths. The components may have different phases and different amplitudes, and their mutual relationship may also vary continuously with time. Multipath effects result from reflections from buildings, from the surface of the Earth, or from horizontal interfaces between different layers in the atmosphere. Multipath effects caused by reflections are responsible for the fast fading observed on microwave radio links. They can seriously degrade the quality of a service. On frequencies below 10 GHz, multipath fading is the main concern of microwave engineers. Multipath fades result when there are two or more transmission paths for a communication link and there are many active transmission paths for the signal to propagate on, and all contribute to the signal at the receiver. Signals that arrive delayed have been phase shifted as a function of frequency with respect to those signals that travel a shorter path. At some frequency the delayed signal phase change will be 180° out of phase with the nondelayed signal (frequency-selective fading), and at this frequency the received frequency spectrum will contain a notch, since the two received signals will add destructively. The depth of the notch will depend on the strength of the delayed and nondelayed signals and will introduce dramatic distortions, which must be corrected in some way; the use of antenna space diversity or equalizers is a potential solution. To test in the lab (two radios in back-to-back configuration) the receivers ability to cope with dynamic effects of multipath fades, it is necessary to be able to accurately simulate these degradations in the lab environment, and by measuring M-curves (or signature) determine the equalizer performance. The M-curve must always include the BER threshold value and the delay. The M-curve is the most common multipath measurement, and the smaller the M-curve, the better the radio can cope with the multipath.
6.2 Theoretical Aspects of Microwave Link Design 6.2.1
Microwave-Radio-Path Calculation Overview
This is an overview of the basic calculations associated with the digital microwave radio path using widely recognized performance and availability objectives. More information can be found in [13]. Calculations of the
Microwave Point-to-Point System Design
207
short-term (due to multipath fading) outage time are based on Vigants widely used, field-verified model as defined in [4] for North American links and [5] for international (ITU-R) links. Newer ITU-R Rec. PN.530-7 multipath-outage-prediction methods (with Hosoyas space-diversityimprovement model) can produce results widely different from Vigants model, especially for flatland paths using space diversity in difficult geoclimatic regions. Long-term (due to rain fading) outage-time calculations are based upon Cranes model, which user-selects either Cranes or ITU-Rs point rain-rate tables for both North American and international (ITU-R) radiorelay links. Cranes point rain rates [6] are in close agreement with ITU-R point rain-rate data [7] in all regions worldwide. The actual rain regions and their letter designations in Crane and ITU-R are, however, different. While the Crane data may be used for international rain calculations (and vice versa), Crane regions are usually user-selected for North America and ITU-R regions are usually selected internationally. The ITU-R PN.530-7 rain model, which typically computes lower rain attenuation and, thus, less rain outage on shorter high-frequency links, will not be discussed here. 6.2.2
Design Fade Margins
The thermal (or flat) fade margin is the difference between the unfaded receive signal level and the receivers static or dynamic threshold, as measured with back-to-back radios, at a given BER. In digital microwave links, MW equipment specifications provide the ability to select either the 10−3 BER dynamic threshold for outage computations or the 10−6 BER static threshold for other performance and availability computations. It is important to understand that an internationally defined outage in a digital radio link corresponds to a 10−3 BER severely errored second (SES) event at the receivers dynamic threshold, very near the initiation of a T1 or E1 alarm (AIS) condition in the connected PABX trunk, channel bank, transmultiplexer, and so forth. The 10−6 digital-radio-static threshold (operating point) is used for factory and in-service manual measurements (with attenuators) and as a measure of circuit quality, not outage, although it is sometimes assigned as the outage threshold by some users. This method could lead to overdimensioning of the MW system, making it more expensive than necessary. A few digital microwave radio links with adverse geometry are susceptible to dispersive fading (spectrum distorting), as well as interference and flat fading, discussed above [8, 9]. Dispersive fading may cause the loss of the
208
Transmission Systems Design Handbook for Wireless Networks
digital receiver synchronization due to amplitude or group delay slope and notch effects even in the absence of significant flat fade activity. In addition to antenna heights, sizes, and vertical alignments optimized to the path geometry for minimum dispersive fading, effective countermeasures include space diversity (SD), adaptive time domain equalization (ATDE), and adaptive IF slope amplitude equalization (ASAE). As the RF bandwidth or number of signaling states increases, the effect of dispersive fading intensifies, requiring these more sophisticated countermeasures typical of todays digital microwave receivers. Fade margins in digital radios are far more complex than in analog microwave systems. In addition to the thermal (flat) fade margin, dispersive and interference fade margins may impact upon the performance of some digital links and, thus, have to be considered. The formula for the composite fade margin (CFM), accommodating all of these influences upon digital link performance, is CFM = TFM + DFM + IFM The dispersive fade margin (DFM) provided by the manufacturer is included, where appropriate, in the CFM computation. It is derived from the digital radios 6.3-ns multipath delay signature curve for preliminary link design, where the actual delay (and therefore the actual link DFM) from path geometry calculations has not yet been made. Final design (beyond the scope of this tutorial) computes the link DFM (with antenna discriminations and actual multipath delay) from path geometry computations. The term interference fade margin (IFM) (usually TFM + 6 dB, a TIA standard) defines the digital links vulnerability to cochannel and adjacent channel interference and is provided by the frequency search company based on the manufacturers threshold-to-interference (T/I) curves and the interference ambiance. IFM is often ignored in the preliminary link design. Based on the standard TIA-TSB-10F, a 1-dB or other threshold (fade margin) degradation due to interference may be entered as a miscellaneous loss, if desired. The combining of different decibel power levels requires more than simply arithmetic addition or subtraction: CFM = −10 log (10 − TFM /10 + 10 − IFM /10 − 10 − DFM /10 ) where TFM = thermal (flat) fade margin (dB);
Microwave Point-to-Point System Design
209
IFM = interference fade margin (dB); DFM = dispersive fade margin (dB). Once the composite or thermal (flat) fade margin has been determined, it is used to compute outage times and expected one-way path reliability or two-way availability. Path reliability (from the annual one-way short-term outage time) is influenced by the effects of the following: • Multipath fading, defined by geographical propagation characteris-
tics (climate and terrain) for given locales; • Average annual temperature for the same geographical region.
CFM is also input for computing the expected annual two-way path availability in high-frequency links that accommodates all long-term (10 CSES traffic disconnect) rain outage events. Annual rain outage and path availability are computed from the following: • Rain attenuation tables, which assign coefficients based upon fre-
quency and polarization; • Rain-rate tables based upon thunderstorm and similar high rain-rate activity in all worldwide regions.
Rain outage is considered a self-healing unavailability event, and unlike other such events, rain outage is statistically predictable over a long-term (more than one year) period. Equipment, antenna system, and infrastructure failures and manual intervention (e.g., switching), which could also cause long-term outage, are not a part of these two-way availability calculations. The fade margin of a digital or analog microwave radio path provides a safety margin to protect the microwave signal from the adverse effects (carrier-to-noise degradations) of multipath fading, interference, and rain attenuation. Digital-microwave-link fade margins are typically smaller than for analog radio links whose fade margins are often increased to provide baseband quieting (low thermal noise) even in short and other nonfading paths. The provision of an adequate fade margin and path clearance to protect against long-term CFM degradations and outage due to surface ducting and earth blocking, plus optimum diversity on longer or otherwise vulnerable
210
Transmission Systems Design Handbook for Wireless Networks
paths, assure per-hop propagation reliabilities in a Rayleigh-distributed multipath fade environment exceeding 99.999%. Path reliability and quality [residual bit error rate (RBER), %EFS, etc.], define digital-radio-link performance during traffic availability periods. Path availability is that percentage of time that a given microwave link is operational (traffic is not dropped or disconnected and performance measurements are possible) over a specified period, typically a year. With the exception of high-frequency (10 GHz) microwave links in rain areas, the availability objective of most microwave links is 100%. Predicted (measured) only during available (traffic-connected) periods, path reliability is a measure of annual short-term (CSES) one-way multipath fade outage occurring over a 24.5-month fade season. 6.2.3
Diversity Improvement
The probability of outage for a diversity system is U d = U nd /I d where Ud = one-way probability of outage for a diversity path; Id = diversity-improvement factor. The diversity-improvement factor differs for each type of diversityspace, frequency, hybrid, and so on. Space-diversity (SD) systems provide more efficient use of the spectrum, as well as extremely good diversity protection. In a typical diversity system, the radio at each terminal may contain an unprotected transmitter or Monitored Hot Stand By (MHSB) transmitters and redundant receivers (Figure 6.1). The transmitter is energized and operates on the same RF channel as
Data in
Tx A Tx B Rx A
Data out
Switch Rx B
Switch Antenna Data in −1 dB Power splitter −7 dB
Tx A Tx B Rx A
Data out
Figure 6.1 Hot stand by and space diversity.
Switch
Switch Rx B
−1 dB
Main antenna
330 m. 1090 ft. Vertical separation
Diversity antenna
Microwave Point-to-Point System Design
211
the diversity receivers at the far end. It uses the same antenna as one of the diversity receivers at the same site. Both receivers operate on the same RF channel (as the far-end transmitter); however, each receiver uses a separate antenna. When equipment failure or path fading affects the main path to the point where the signal is degraded, a decision is made (electronically) at the receiver end of the path as to which RF path is placed on-line. The SD arrangement has two receivers connected to vertically spaced main and diversity dishes. Hitless or errorless switches are used for switching the signal between the main and diversity branch. In a frequency-diversity (FD) system, the radio at each terminal contains redundant transmitter-receiver pairs. Each transmitter operates on a different RF channel and both transmitters are energized, and similarly, each receiver operates on a different RF channel, but is identical to the corresponding transmitter at the far end. When equipment failure or path fading affects an RF channel to the point where the signal is degraded, a decision is made at the receiver end of the link as to which RF channel is placed on-line. There are a number of other diversity methods for improving the particular microwave hop performance. Angle diversity (AD) has been used in LOS digital microwave links since the mid-1980s and in troposcatter links since the 1950s. The calculations use Vigants model, requiring vertically spaced dishes, and assume flat (analog) rather than dispersive (spectrumdistorting digital) fade activity on a path. This method used to be popular when digital radios were less robust, but today AD antennas are assigned mostly where installation constrictions (space, aesthetics, tower loading, and so on) prohibit SD and, thus, justify these less effective, more costly dishes. The AD antenna is a single dish with two feeds vertically offset by about 1° (the smaller the better). AD is most effective when path outages are dominated by dispersive fade activity (dispersive fade outage approaches or exceeds flat fade outage). As reported in the 1980s in AD studies [10], links with several of the above characteristics showed improvement factors ten or more times greater than space diversity (500 versus 50, for example). Lowercapacity radio links and those with new digital radios with more robust modulation and adaptive countermeasure schemes realized less improvement from angle diversity. Depending on path geometry and climatic conditions, AD improvements of perhaps 20 or even much more can be achieved. Optimum AD improvements are only obtained through an antennaalignment procedure that matches the antenna size and alignment to the path and its climatic characteristics. AD dishes require a more exacting, long-term alignment procedure than that for space and nondiversity antennas.
212
Transmission Systems Design Handbook for Wireless Networks
Hybrid diversity (HD) is an enhancement of space diversity, using FD, when permitted (SD + FD). Hybrid diversity is the most effective of all of the diversity arrangements and is preferred in difficult propagation areas and in space-limited mountaintop, urban areas, and other sites restricted to single antennas. It is possible to compute frequency- and hybrid-diversity improvements for links in regions where regulatory rules or waivers so permit. The hybrid-diversity improvement factor, Ihd, is derived either from the SD or FD improvement factor Ifd described above. Above 3 GHz, Isd is nearly always higher (better) than Ifd and is therefore selected unless the diversity spacing exceeds about 5% (300 MHz/0.3 GHz in the 6-GHz band, for example). Below 3 GHz, Ifd is usually larger than Isd and therefore selected. The higher T&R frequencies must always be assigned to the upper antenna at the SD (usually the lower elevation) end of each HD link for optimum performance. 6.2.4
North American and ITU Objectives
The recommended short-term one-way outage objective for a T1/E1 trunk or FDM circuit, regardless of system length, is 1,600 outage seconds per year (SES/yr) for a 99.995% end-to-end propagation reliability. This is as equally true of a 5-hop short-haul system as it is of a 150-hop long-haul system, except that in a very long-haul system (perhaps > 50 hops), only half are considered as fading hops. A 99.999% per-hop reliability (320 SES/yr outage) objective (floor) is often assigned in spur links and on short systems of less than about 5 tandem hops. Although computed over the fade season, this is considered an annual outage compared, perhaps, to Bell short-haul (< 400 km/250 mi), Bell longhaul (6,400 km/4,000 mi), or other end-user objective. The long-term rain outage in a high-frequency (10 GHz) microwave link is often engineered to this same 99.995% objectivenow an annual two-way rain availability, not a one-way multipath reliability, objective. This corresponds to the ITU-R per-hop availability objective, since the 60-hop 2,500-km ITU-R reference circuit end-to-end two-way availability objective of 99.7% [11] scales to a 99.995% availability objective for a typical 40-km link. The 1,600 SES/yr outage objective for a system (longest one-way T1 trunk or FDM message circuit) may be over links making up a part or the whole of a long-haul system, a short-haul system, or other distance. While the following computations are for link(s) making up a full short- or long-
Microwave Point-to-Point System Design
213
haul system, the 1,600 end-to-end SES/yr outage allocation may be proportioned only to the actual number of tandem hops in an actual system. That means the assignment of a 1,600/20 = 80 SES/yr (99.99975% path reliability) per-hop objective for a 20-hop digital microwave system. The ITU publishes recommendations for the telecommunication and radio communication areas. The recommendations for the telecommunication part are published by ITU-T, whereas the radio communication part is published by ITU-R. The ITU also provides maps and data, proven adequate and reliable, which allow identification of the applicable rain-rate statistics in any part of the world. For this purpose, the ITU has divided the world into 24 regions, each identified by similar rain-rate statistics. For example, in Europe, the maximum rain rate (i.e., rain rate not exceeded for more than 0.01% of the time) varies from about 20 mm/hr in southern Europe to about 40 mm/hr in most parts of Mediterranean Europe. It peaks at 60 mm/hr in small areas of southern Europe. In total, the ITU identified six regions in Europe and five regions in the United States, each with the same maximum rain rate. Regions in the United States range from 19 mm/hr on the west coast to 63 mm/hr in the southernmost states (based on 99.99% availability). The recommendations G.801, G.821, and G.826 define error performance and availability objectives. The objectives for digital links are divided into separate grades. These are high, medium, and local grade. The medium grade has four quality classifications. The following grades are usually used in wireless networks: • Medium-grade Class 3 for the access network; • High grade for the backbone network.
The ITU Recommendation G.703 defines the physical and electrical characteristics of digital interfaces, while Recommendation G.704 defines synchronous frame structures. These two recommendations are followed for all links in the transmission network. An RBS site becomes unavailable when the transmission link connecting it to the next site or BSC becomes unavailable. ITU Recommendation G.821 provides guidance on allowable unavailable time (AUT) for networks. For wireless networks, transmission link performance is typically based on meeting ITU medium-grade Class 2 or Class 3 objectives. They are based on a hypothetical reference connection, as depicted in Figure 6.2. Specific details concerning the application of performance and availability objectives to telecommunications connections can be found in ITU-T
214
Transmission Systems Design Handbook for Wireless Networks 1,250 km
LE Local grade
25,000 km
IG
1,250 km
LE
IG High grade
Medium grade
Medium grade
Local grade
IG = International gateway LE = Local exchange
Figure 6.2 ITU-T G.821 hypothetical reference connection.
Recommendation G.821, and the application of these objectives to digital radio-relay systems (microwave links) is detailed in ITU-R Recommendation 696-2. Table 6.1 shows the Rec 696-2 medium-grade performance and availability objectives that are typically applied to the design of microwave links operating at fundamental bit rates below 2 Mbps (this covers PDH links typically used in the access network). These ITU objectives are typically applied to microwave links in wireless networks. If we take the medium-grade Class 3 objective and consider the unavailable time (U) objective: U = 0.05% per year The intention of the ITU is that this objective should be applied to an entire connection; for example, if there is a chain of 10 links connecting RBSs to Table 6.1 Medium-Grade Performance and Availability Objectives Percentage of Any Month Performance Parameter −3
Percentage of Any Month
Class 2 (280 km) Class 3 (50 km)
SES (BER10 )
0.0075
0.002
ES
0.16
0.16
Unavailability
0.05
0.05
Microwave Point-to-Point System Design
215
the BSC, then the unavailability objective for each link should be 0.05%/10 for each link, or 0.005%. In a real network there may be fewer than 10 links connecting a series of RBSs to a BSC; however, the network operator may still choose to apply the figure of 0.005% to each link, to build in a performance margin. 6.2.5
Reliability and Availability Overview
In considering the assignment of a realistic short- or long-term outage objective, several things need to be kept in mind. A single overall design objective of not more than x hours, minutes, or seconds outage over some period such as a year, is an oversimplification. The character of the particular kind of outage and its effect on the system should be taken into account, and perhaps there should even be different objectives for different types of outages. For example, propagation outages due to multipath fading are usually short. A cumulative outage of an hour per year due to multipath might represent thousands of individual outages, each averaging one second or less (one SES) on a properly engineered path. On the other hand, propagation outages totaling an hour per hop due to rain attenuation might consist of only four or five individual outages averaging 10 to 15 minutes each. The effects of long-term and short-term system outages on trunks are very different. The many short-term unreliability outage events do not disconnect circuits or reduce (in most circuits) data throughput. The few longterm unavailability events cause both traffic disconnect and loss-of-data throughput. A distinction should be made between those circuits for which an outage of a few seconds or a few minutes is just an inconveniencefor example, spur cell site in the wireless networkand those circuits for which such an outage might result in a danger to life, great economic loss (important or hub sites in a wireless network), or other catastrophic consequences. The suitability or unsuitability of a higher frequency band such as 18- or 38GHz, in a high-rain-rate region could differ widely for these two situations. Even if the maximum possible reliability and availability objectives are established and a path or a system is engineered to the full limit of the state of the art, the probability of outage can never be eliminated, but only reduced to a very low value. Thus, it is imperative to make any very important services as fail-safe as possible against a loss of the communications channel. The system should be engineered with appropriate protection schemes and diversity arrangements so that short- or long-term outages are tolerated, or at least kept within acceptable bounds. Ring (route diversity) protection is often used to eliminate long-term rain outage, for example.
216
Transmission Systems Design Handbook for Wireless Networks
It seems that in most cases, a more relaxed attitude might be taken toward rain-induced outages than toward multipath outages or even equipment outages. In several respects, rain outage is somewhat benign in nature. If the fade margins are kept high and the paths are not stretched out too much, even in less advantageous areas, the number of outages per year should not be very large. The length of individual rain outages on a hop should only rarely exceed 5 to, perhaps, 10 minutes. Short (fewer than 2 seconds) microwave outages, common on a typical longer diversity or shorter nondiversity digital microwave link with adequate fade margin, will not drop telephone or data lines. Such outages quickly clear with all circuits remaining connected and little note taken of these transient events. Longer outages associated with low fade margins, rain, and so on, disconnect all subscribers and may block access to a digital link for at least 10 seconds after each long-term outage event. Such traffic disconnects are unacceptable to most users; thus, these more vulnerable links clearly require appropriate diversity or ring protection. For high-reliability links (usually in long-haul systems with many hops in tandem), the per-hop objective may approach or exceed 99.9999%, allowing only 20 to 30 seconds of per-hop outage per year. Short-haul systems up to about 10 hops are often assigned a per-hop design objective of about 99.9995% for 160 SES/yr outage. In a wireless network, this would apply to the BSC-MSC and MSC-PSTN connections. Spur legs or short systems with 1 to 5 hops may be assigned a relaxed 99.999% perhop path-reliability objective equating to 320 SES (5.3 minutes) outage per year. In wireless networks, this would apply to hub or other important sites. For other services, even dramatically lower path reliabilities may be acceptable, perhaps approaching 99.99% or about 1 hour of outage per year. This would apply to the end cell sites in the wireless network. It is important to note that all the path reliability formulas and models in this text represent short-term one-way outage. To calculate two-way multipath outage time (rarely used to characterize link performance), double the calculated multipath outage. Unavailability outage times due to rain fading are not doubled, because they occur simultaneously in both directions of transmission and are always two way. It is important also to understand that per-hop performance and availability objectives are just that, objectives, not per-hop requirements. Seasonal, terrain, and geoclimatic changes average out on longer systems, thus permitting end-to-end guaranteed compliance to a users performance requirement. But on a per-hop basis, many links will perform much better, a few somewhat worse, than performance and availability objectives computed here.
Microwave Point-to-Point System Design 6.2.6
217
Effects of Rain on Microwave Propagation
The principal gaseous absorption is by oxygen and water vapor. The attenuation due to oxygen is relatively constant in the 214-GHz frequency range. Water-vapor absorption, on the other hand, is highly dependent on the frequency, as well as the density of the water vapor (absolute humidity, gm/m3) [10]. Heavy rainfall, usually in cells accompanying thunderstorm activity and weather fronts, has a great impact on microwave-path availability above 10 GHz. This long-term (515 minutes) outage time usually causes traffic disconnects, and such long-term outage is never added to short-term multipath outage (previously discussed). Rain outage increases dramatically with frequency, and then with path length. Duration fades of 10 to 15 minutes to over 50 dB have been recorded on an 18-GHz, 5 km/3 mile path in Houston, for example, and increased outage at 23 GHz can require a 2-to-1 reduction in path length compared to 18 GHz for a given availability. The predicted annual outage may not occur for years, and then accumulate over a single rainy season for a long-term average [12]. Early studies, both theoretical and experimental, resulted from the recognition of the importance of rain in designing microwave paths with an availability objective in excess of 99.9%. In recent years the emphasis has been on establishing predictive techniques for the statistical estimation of the attenuation-probability distribution for a particular path. R. K. Crane has developed a model for determining the attenuation due to rain based on several factors, including path length, frequency, and point rain rates. Coefficients given by Crane [13] and ITU-R Rec. 838 are identical. The point rain-rate tables in Crane and ITU-R [2] are, however, dissimilar only as worldwide rain regions are assigned different letters, although the rain rates are very nearly the same. We may select either Crane or ITU-R point rain letters worldwide for all regions. However, Crane rain regions are usually selected from maps for North America, and ITU rain regions are usually selected from maps for international links. The more optimistic ITU-R PN.530xx rain model typically computes lower rain attenuation and, thus, less rain outage on shorter high-frequency links. Rain attenuation at the higher microwave frequencies (>10 GHz) has been under study for more than 40 years. Much is known about the qualitative aspects, but the problems faced by the microwave transmission engineerwho makes quantitative estimates of the probability distribution of the rainfall attenuation for a given frequency band, polarization, path length, and geographic (rain distribution rate) areastill remain difficult. In order to estimate this probability distribution, instantaneous rainfall data is needed. Unfortunately, the
218
Transmission Systems Design Handbook for Wireless Networks
available rainfall data is usually in the form of a statistical description of the amount of rain that falls at a given measurement point over various time periodsgenerally at least an hour in length. The rain-induced attenuation along a given path at a given instant in time is a function of the integrated effect of the rainfall existing at all points along the path. It is affected not only by the total amount of water in the path at that instant, but also by its distribution along the path in volume and drop size (intensity). For heavy rain rates, the instantaneous distribution of volume and drop size along the path is highly variable and is difficult to predict with the accuracy expected of the rainfall data generally available. Increasing the fade margins, shortening path lengths, and increasing antenna sizes are the most readily available tools for reducing the per-hop annual rain outage in a given area. Route diversity (ring protection) or a lower, less vulnerable frequency band (perhaps 10 GHz, in digital systems) is often considered to reduce or essentially eliminate the impact of rain outage on system availability. The total annual rainfall in an area has little relation to the rain attenuation for the area. Within the United States, for example, the northwestern states have the greatest annual rainfall (in excess of 100 in/2,500 mm per year) produced, however, by long periods of steady rain of relatively low intensity at any given time. Other areas of the country with lower annual rates experience thunderstorms and frontal squalls, which produce short-duration rain rates of extreme intensity. It is the incidence of rainstorms of this type that determines the rain rates for an area and, thus, the high-frequency microwave links long-term path outage time and unavailability. Even the rain statistics for a day or an hour have little relationship to rain attenuation. A day with only a fraction of an inch or centimeter of total rainfall may have a path outage due to a short period of concentrated, extremely high-intensity rain. Another day with several inches or centimeters of total rainfall may experience little or no path attenuation because the rain is spread over a long time period or area. The most common reason for a preference for a lower frequency is the susceptibility of bands above 10 GHz to rainfall attenuation. Although the effect is present to some degree at lower frequencies, it increases rapidly with frequency. For example, a rain-cell intensity causing only a few decibels of attenuation at lower frequencies could be sufficient to cause a path outage at 18 GHz. Although fades caused by rain cells are occasionally observed at lower frequencies (1020-dB fades at 6 GHz have been recorded even in North America), this type of fade generally causes outages only on paths above
Microwave Point-to-Point System Design
219
10 GHz. The outages are usually caused by blockage of the path by the passage of rain cells (e.g., thunderstorms), perhaps 4 to 8 km, or 2.5 to 5 mi, in diameter and 5 to 15 minutes in duration on the path. Such fading exhibits fairly slow, erratic level changes, with rapid path failure as the rain cell intercepts the path. The fades are nonselective in that all main and diversity paths in both directions are affected simultaneously. Vertical polarization is far less susceptible to rainfall attenuation than horizontal polarized frequencies. Increased fade margin is of some help in rainfall attenuation fading; margins as high as 45 to 60 dB, some with automatic transmitter power control (ATPC), have been used in some highly vulnerable links for increased availability. When permitted, seldom used crossband diversity is totally effectivethe lower-frequency path is stable (affected only by multipath fading) during periods when the upper-frequency path is obstructed by rain cells. Route diversity (ring-protected paths separated by more than about 8 km, or 5 mi) is also used successfully. In summary, these are the things to bear in mind in connection with rain attenuation fades: • Multipath fading is at its minimum during periods of heavy rainfall
with well-aligned dishes, so the entire path fade margin is available to combat the rain attenuation (wet-radome loss effects are minimized with shrouded antennas). • Neither SD nor in-band FD, provides improvement against rain attenuation fade outage. • Vertically polarized high-frequency-link rain outage is 40% to 60%
less than those links horizontally polarized. • Use of mircowave systems below 10 GHz will eliminate problems caused by rain.
6.3 Practical Aspects of Microwave-Link Design 6.3.1
Design Overview
The first step in the microwave system design for the wireless network involves determining the actual system requirements and site locations, which should typically be defined by the RF group in coordination with the transmission planning group, and with the help of the real estate group as well as path surveyors. Shown here is a general list of the areas of information that must be gone over with the customer during the proposal and initial
220
Transmission Systems Design Handbook for Wireless Networks
planning stage; more detailed questions will have to be answered during the design and deployment stage. • Information on the existing PSTN facilities in the areaswitching and
transmission, ILECs, CLECs, and so on. If possible, a map should be created showing all of the below that exist in the area. • Microwave backhaul questionsfrequency bands, regulatory issues, preferred suppliers, existing systems in the area, and so on. • Fiber-optic backhaul questionsexisting fiber-optic facilities, leasing of dark fiber, types of cables and equipment, and so on. • Operator preferences regarding owned versus leased facilities, equipment supplier, engineering and installation services, points of interconnect (POI) questions, and so on. • Miscellaneous (optional) questionsorganizational and logistical information. Based on the information provided in answering these questions, transmission engineers will work with microwave engineers to determine the best approach and strategy in formulating an overall transmission plan. The transmission plan could be completely microwave or it could be a combination of microwave systems and leased lines. Routing design and preliminary path analysis should be done prior to any field trips. Computer programs are available to plot an initial system map, which shows the sites in geographical relationship to each other. Approximate site coordinates (avoid GPS unless the measurement is done in differential mode) and the feasibility of various paths are determined from topographical maps, digital terrain databases (minimum 1:50,000), or old survey information. This map is used together with traffic requirements to define the paths between sites and the type of radio for each hop. For very short microwave hops (sites very close to each other, less than 3 km), a detailed path survey will not be required. Construction managers could be trained to perform this quick analysis and, using binoculars, see the remote site of the MW hop. This is generally sufficient to confirm LOS. Any written observations, comments, and photographs will later be useful to the microwave designer. Field verification of the LOS (path survey) between proposed wireless sites should be done before any further negotiations with the potential landlord take place. Interference analysis and frequency coordination play an important part in proposed route design. Governments usually require users of the radio spectrum to frequency-coordinate their planned and existing microwave radio
Microwave Point-to-Point System Design
221
systems with other users of the radio frequency spectrum. Such coordination is a prerequisite in any microwave radio license application by a microwave radio system operator. The radio license applicant must determine if the planned radio system will experience any interference from the existing environment and vice versa. Potential interference can be calculated for three different cases: interference between microwave point-to-point radio stations, between microwave radio relay stations and earth stations [14], and between radio relay stations and geostationary satellite orbit. The results of these calculations will indicate whether or not there is potential interference and whether redesign or relocation of the planned MW system is required. The microwave radio site survey report (checklist) is a document that will include site details, address, directions to the site, access, restrictions, general information on the site, tower or rooftop leasehold, available space and equipment at the site, dc and ac power availability, and zoning restrictions [13]. The document is usually issued and signed by the construction manager or project manager. At this point, if all the results of the previous analysis are positive and the sites have been accepted by all groups, a detailed microwave path design (based on required availability and reliability of the system) and equipment selection can be completed. The typical output of the MW system engineering design is as follows: • MW path calculations; • System layout; • T1/E1 plan; • System block and level diagrams; • Rack profiles; • Tower loading specifications and antenna placement. 6.3.2
Protected and Nonprotected Microwave Systems
The terms protection and diversity are often used interchangeably when applied to microwave links. This is not correct, since protection commonly improves long-term traffic interruptions [10 consecutive severely errored seconds (CSES) or more], while diversity arrangements greatly reduce the number and duration of short-term outages (less than 10 CSES). The radio path has two transmitters and two receivers always on-line (hot). There is a switch that keeps one radio transmitting or receiving until a failure occurs,
222
Transmission Systems Design Handbook for Wireless Networks
and at that moment, the signal is switched to the standby radio. The MSHB configuration (1+1) protects against equipment failures only, not path propagation problems. The hot-standby-protection system uses a common frequency channel, with only one of the two transmitters at any end of the microwave link active at any particular time. The transmitter in standby mode will be fully operational except that the transmitter will be muted. The hot-standby (protected) system configuration provides hitless receiver changeover on each side of the radio relay link in case of receiver equipment failure or sudden propagation path fading on one of the four microwave paths. If a transmitter fails, there will be a short break in transmission until the standby transmitter is activated. In contrast to receiver changeover, transmitter changeover will therefore not be hitless. 6.3.3
Microwave Repeaters
Sometimes it is impossible to connect to Point B from Point A in one microwave hop, although from a propagation or quality perspective it is feasible. There could be an obstacle (a tree, building) between points A and B obstructing LOS. If there is no other solution and it is imperative to connect these two points (cell sites, cell site to the switch office), a microwave repeater must be used [15]. An active MW repeater site contains two complete microwave radio terminals, together with the antennas, waveguides, or coax cables, and the like, connected back to back. It is a much more costly solution than the passive repeater. It requires enclosure for the equipment, a power plant, an antennamounting structure of some kind, and so on. In other words, it requires the complete microwave site without the benefits of a revenue-generating cell site providing the RF coverage to that area. This is a very expensive solution to be avoided if possible. The best way to avoid use of microwave repeater sites is to plan and execute the RF coverage carefully and to strategically place cell sites in such a way that they all have an LOS with at least one other cell site. This requires careful coordination between RF and transmission or microwave planners from the very beginning of the project. Where a direct microwave path cannot be established (no LOS) between two points, it is sometimes possible to establish a path by using a passive MW repeater. The function of such a repeater is to redirect the microwave beam in order to pass the beam around or over the obstacle (buildings, hills, etc.). The main requirement is that there is a LOS between the passive repeater and both sides of the microwave link. There are two types of passive repeaters in use: One requires two parabolic antennas connected back to
Microwave Point-to-Point System Design
223
back through a short piece of transmission line; the other, more commonly used, is a flat billboard-type metal reflector, which acts as a microwave mirror. The first type is rarely used due to its inefficiency. The effectiveness of a passive repeater is an inverse function of the product of the lengths of the two paths, rather than the sum of their lengths, as one might suppose [15]. Thus, it is highly desirable to keep one of the paths very short (a few hundred meters if possible). Billboard passives are used for frequencies of 6 GHz and above and fall into two basic configurations, depending on the geometric relationship. If the site of the passive repeater is off to one side, or behind the terminal, so that the angle between two paths at the reflector is less than 130° (the smaller the angle the better), a single billboard can be used. This is the most common application. However, if the only available location happens to be more or less in line with the path, a double billboard may be required, consisting of two reflectors usually fairly close together and geometrically arranged to reflect the beam at the proper angles. For single billboards, things are simple if the billboard is in the far field of both antennas. In that case, the antenna gains and the billboard gains are independent and do not interact with each other. Then we simply calculate a total path loss, which is the sum of the two antenna gains and the two-way gain of the reflector, in order to arrive at the end-to-end path loss through the reflector. Net path loss = A 1 + A 2 − G1 − G 2 − G3 where A1 = Short-leg attenuation (dB) A2 = Long-leg attenuation (dB) G1, G2 = Antenna gains (dB) G3 = Free-space, two-way gain of a single passive billboard (dB) Passive-repeater-gain calculations could be complicated in cases where the billboard is in the near field of one of the parabolic antennas. A rule-ofthumb formula to determine the near field boundary is D = 2 × f × B2 where D = Near-field zone distance (ft)
224
Transmission Systems Design Handbook for Wireless Networks
F = Frequency (GHz) B = Antenna diameter (ft) In that case, Net path loss = A 1 + A 2 − G1 − G 2 − (G3 − K ) where K = Correction factor (dB) Correction factor K can be calculated using empirical formulas and graphs, and its value will depend on the parabolic dish diameter, frequency, and distance between the parabolic antenna and the passive repeater. Its value is usually between 0.2 and 1.6 dB. 6.3.4
Microwave Path Calculations
Microwave path calculations are performed as part of a detailed microwave system design, and all the detailed hardware requirements are defined based on this information. There are a number of activities involved in preparation for path engineering: • Perform site and path surveys and become familiar with the market
and the terrain;
• Consider linear or ring architecture and MW repeater facilities; • Define availability, outage, and quality standards for the individual
links and the overall system;
• Specify capacity requirements and available frequency-band choices
(higher-frequency bands for shorter distances and lower-frequency bands for longer distances);
• Determine (equipment or route) protection requirements; • Determine propagation diversity arrangements; • Use a microwave design tool (e.g., Ericssons LinkPlanner or
vendor-independant PathLoss 4.0) for detailed path engineering and interference analysis;
• Generate a bill of material (BOM) for the microwave system.
Microwave Point-to-Point System Design
225
Transmission-network requirements for reliability of the wireless networks in North America are shown here (equipment not included). This is a basic rule of thumb and a fairly simplified practical approach, but can serve as a guideline. • For high-reliability links (usually in long-haul systems with many
hops in tandem or backbone), the per-hop objective may approach or exceed 99.9999%, allowing only 20 to 30 seconds of per-hop outage per year. • Short-haul systems of up to about 10 hops are often assigned a perhop design objective of about 99.9995% for 160 SES per year outage; this objective should be used to design BSC-PSTN MW links. • Spur legs or short systems with 1 to 5 hops in tandem assigned a relaxed 99.999% per-hop path-reliability objective equating to 320 SES (5.3 minutes) outage per year; the same criteria should be used for the links connecting important hub sites. • For single MW hops (or maybe up to 2 tandem hops), the required objective should be 99.995% per hop; this applies to RBS-BSC and RBS-RBS microwave hops. Sometimes these numbers are different and could be proposed by the customer; it is important to keep in mind that by definition, higher objective numbers lead to more expensive MW networks. Further information about the microwave system theory and practical microwave design can be found in [1618]. 6.3.5
Microwave Interference Analysis and Frequency Coordination
Radio frequency coordination is the term given to procedures followed by users of a common band of radio frequencies to minimize and control potential interference between systems. The key aspect of the procedure involves cooperative (or at least informed) radio-frequency planning. Radio systems should be designed in such a manner that they do not cause or suffer objectionable interference with other existing or planned systems using the same frequency band. This coordination is facilitated by sharing coordination data among users, so that accurate and up-to-date information is available with which estimates of potential interference can be made during the system design stage. Radio-frequency-interference studies and frequency coordination are necessary not only when designing a new system, but also when determining
226
Transmission Systems Design Handbook for Wireless Networks
the potential interference effects of other users radio construction proposals on existing and planned systems. Thus, coordination is involved when one party initiates construction plans as well as when reacting to other parties plans. The frequency spectrum is a valuable resource and is generally subject to appropriate planning and management to prevent misuse and interference between the many and varied applications. National administrations will allocate some or all of these bands for fixed microwave radio use in line with local requirements. Before network planning commences, an operator must determine available frequency bands and channel plans specific to that country. Often, and preferably, an operator, in this case a wireless operator, is able to obtain a number of frequency allocations as a block, enabling him to perform his own network planning in advance without risk of interference from other users. Most regulatory authorities also operate a local-link-length policy, where the length of a particular path will determine what frequency bands are available for the operator to choose from. Typically, the shorter the path, the higher the frequency required. The first step is to perform intrasystem frequency coordination (within the given network); then, if results are satisfactory, perform intersystem frequency coordination. As stated earlier, the radio license applicant must determine if the planned radio system will experience any interference from the existing environment and vice versa. Potential interference can be calculated for three different cases: interference between microwave stations, between microwave stations and earth stations, and between microwave stations and geostationary satellite orbit. The results of these calculations will indicate whether or not there is potential interference and whether redesign or relocation of the planned MW system is required. In many cases, the most reliable information on the potential interference cannot be obtained by calculations, since there is little or no information on the existing terrestrial or satellite systems in the area. The best way to determine potential interference is to sweep the entire spectrum using test equipment at the future microwave system antenna location. 6.3.6
Microwave System Design Guidelines
The transmission network must be designed to meet service demands, but always with the most economical routing in mind. Two scenarios are most common in wireless-network deploymentleased facilities or the microwave network. For larger networks, it is usually some combination of both, and even the leased-lines (facilities) network requires careful transmission-
Microwave Point-to-Point System Design
227
network planning. Survivability and reliability of the network are achieved by means of transmission loopsthe ring configuration or a combination of star and ring configurations. For a larger wireless network (and transmission network), the ring configuration is recommended as a high-capacity backbone carrying traffic to the switch location (Figure 6.3). Not all the cell sites are shown in this network diagram. As a backbone microwave system, 7-GHz, SDH microwave radios with the STM-1 (63 E1s) capacity are used. Cross-polarization can be used to further increase (double) the capacity of the ring. Higher-capacity microwave radios are used to bring traffic from the cell sites and hub sites into the ring and to the BSC. The following criteria can be used for determining whether to consider using a ring configuration: • More than 20 to 30 cell sites in the network; • More than 40 to 50 T1/E1s required; • More than 4 to 5 MW links required at the switch location;
Cell site
MW ring hub site/cell site
7 GHz/63 E1 BSC/MSC
13 GHz/16 E1
13 GHz/8 E1 PDH microwave link (13 GHz) SDH microwave link (7 GHz) SDH microwave link (XPIC)
Figure 6.3 Microwave system design.
228
Transmission Systems Design Handbook for Wireless Networks
• High network reliability required; • Customer requirement.
The ring configuration provides a reliable communication facility, since it offers automatic protection from site hardware (batteries, towers, antenna systems) failures, radio or MUX equipment failures, propagation failures in the microwave network, and cable cuts in the fiber-optic network. Since additional sites for microwave use only (MW repeater sites) are very expensive, they should be avoided if possible, so it is usual to assume one microwave hop per cell site. These are some of the basic steps for the microwave design as part of the overall wireless-network design: • Use an RF network design tool or existing terrain database to check • • • •
• • • • •
for the basic LOS; If possible, perform preliminary site and path surveys and become familiar with the market and the terrain; Consider linear, star, and ring architecture and MW repeater facilities; Define availability, outage, and quality standards; Specify capacity requirements and available frequency band choices (higher-frequency bands for shorter distances and lower-frequency bands for longer distances); Determine equipment (infrastructure, route) protection requirements; Determine propagation diversity arrangements; Choose the MW equipment vendor; Choose the MW installation services contractor (radio and antenna installation could be, and often is, provided by different contractors); Prepare the ATP.
In microwave systems, additional protection (e.g., space or frequency diversity) at lower frequencies may be required against short-term multipath outages. Diversity systems as a means of increasing reliability of the MW system should be used only for very high-capacity systems or systems of great strategic importance. All the sites that belong to the ring are considered hub sites and must be completed and built first to provide protection.
Microwave Point-to-Point System Design 6.3.7
229
Microwave Lookup Table
To determine roughly the number of MW hops required for the wireless network in a specific market, Table 6.2 can be used. Calculations are done using the PathLoss 4.0 microwave software tool [18]. Rain regions of the world are described using a range of letters, with A, B, and C referring to the areas with very little or no rain, and M, N, and P referring to the rain forest, which experiences heavy downpours. The statistical variations of rainfall intensity, specific attenuation, and attenuation along the path depend in a very complex way on the number, type, and intensity of rainstorms that traverse the
Table 6.2 Microwave Lookup Table Availability
Frequency
99.995%
99.999%
ITU Rain Region
A
B
C
D
E
F
G
H
J
K
L
M
N
P
2
43
43
43
43
43
43
43
43
43
43
43
43
43
43
6
28
28
28
28
28
28
28
28
28
28
28
28
28
28
7
27
27
27
27
27
27
27
27
27
27
27
27
27
27
8
26
26
26
26
26
26
26
26
26
26
26
26
26
26
10
23
23
23
23
23
23
23
23
23
23
22
21
20
15
11
23
23
23
23
23
23
23
23
23
23
22
21
20
15
13
22
22
22
22
22
22
22
22
21
21
19
18
17
17
15
22
22
22
22
22
21
21
21
20
19
17
16
14
13
18
14
14
14
14
14
13
13
13
12
11
9
9
7
6
23
12
12
12
12
11
10
10
9
9
8
6
6
4
3
38
11
9
7
7
6
5
5
4
4
3
3
3
2
1.5
8
29
29
29
29
29
29
29
29
29
29
29
29
29
29
6
19
19
19
19
19
19
19
19
19
19
19
19
19
19
7
18
18
18
18
18
18
18
18
18
18
18
18
18
18
8
17
17
17
17
17
17
17
17
17
17
17
17
17
17
10
16
16
16
16
16
16
16
16
16
16
15
14
13
9
11
16
16
16
16
16
16
16
16
16
16
15
14
13
9
13
15
15
15
15
15
15
15
15
14
14
12
11
9
8
15
15
15
15
15
15
14
14
13
13
12
9
7
6
5
18
10
10
10
10
10
9
9
8
8
7
5
5
4
3
23
9
9
9
8
8
6
6
6
5
5
3
3
2
2
38
9
6
5
4
3
3
3
3
2.5
2.5
2.2
2.2
1.5
1
Average temperature 10°C 211 GHz: 6-ft. dish 1315 GHz: 4-ft. dish 18 GHz and above: 2-ft. dish
Vigants/Crane model used (more pessimistic) Vertical polarization assumed Standard Tx power Average terrain
230
Transmission Systems Design Handbook for Wireless Networks
path each year. No theoretical basis exists for the calculation of the desired rainfall statistics, and they must be obtained empirically. Cranes rain-fading model [6] recommends using a rain-rate distribution tailored to each climatic area. The rain heavily affects frequencies over 10 GHz, while frequencies below 10 GHz are only affected by multipath. Frequencies in the 10-GHz neighborhood experience the effects of both. For example, if the distance between cell sites in the South American market (rain region M or N) is 5 km and we are using a 23-GHz microwave system, there is a chance that we might need MW repeaters between the sites. Since that would be an expensive solution, another option is to use a 15-GHz or even lower frequency, but we may be limited by local regulations or the customers license for microwave spectrum. It is obvious from the table that frequencies around 10 GHz are affected only by heavy rain, while even the smallest amount of rain could affect 38-GHz microwave systems. Cranes model is used mainly in North America, while the ITU method is used in the rest of the world. These two methods for calculating rain fading result in slightly different values. Cell-site backhaul is usually designed to satisfy 99.995% reliability (path propagation only; this does not include equipment), with important sites and higher-capacity hub sites designed for 99.999% reliability. There are a number of excellent microwave design tools on the market. Some microwave equipment manufacturers insist on using their own software tools while some operators and consultants prefer tools available on the open market. One of those vendor-independent tools is PathLoss 4.0. This is probably one of the best tools for complex microwave design including North American and ITU standards, different diversity schemes, diffraction and reflection (multipath) analysis, rain effects, and so on. PathLoss 4.0 is widely accepted by microwave system design engineers around the world. Figure 6.4 shows an example of a PathLoss 4.0 microwave path design worksheet.
6.4 Spread-Spectrum Microwave Systems Spread spectrum is a digital-coding technique originally developed for military purposes; coding protects the data by increasing the transmitted bandwidth and reducing the power density [16]. This technique resists jamming, interference, selective fading, and interception, as well as minimizing interference to and from other users. Spread-spectrum systems can coexist with other radio
Microwave Point-to-Point System Design
231
Microwave Worksheet027c049d.pl4 Elevation (m) Latitude Longitude True azimuth (°) Antenna model Antenna height (m) Antenna gain (dBi) Vertical angle (°) Connector loss (dB) Frequency (MHz) Polarization Path length (km) Free space loss (dB) Atmospheric absorption loss (dB) Net path loss (dB) Radio model TX power (watts) TX power (dBm) EIRP (dBm) Emission designator RX threshold criteria RX threshold level (dBm) Maximum receive signal (dBm) RX signal (dBm) Thermal fade margin (dB) Geoclimatic factor Path inclination (mr) Average annual temperature Worst month multipath outage (%) (sec) Annual multipath outage (%) (sec) (%-sec) Rain region Rain rate (mm/hr) Rain attenuation (dB) Annual rain outage (%-sec) Annual multipath + rain (%-sec)
027C 1536.67 20 40 52.00 N 103 18 07.60 W 84.17 VHP2-142 24.00 36.50 −0.32 0.00 14,825.00 Vertical 5.67 130.95 0.15 58.11 9415-UX (2E-1) 0.16 22.00 58.50 3.5MOD7W BER = 10-6 −89.00 −20.00 −36.11 52.89 4.45E-05 5.26 21.00 100.00000 2.61E-05 100.00000 1.09E-04 100.00000-0.00 D2-96 Temp. Continent 234.09 52.89 99.99996-12.55 99.99996-12.55
049D 1513.87 20 41 10.70 N 103 14 52.80 W 264.19 VHP2-142 17.00 36.50 0.28 0.00
58.11 9415-UX (2E-1) 0.16 22.00 58.50 3.5MOD7W BER = 10-6 -89.00 -20.00 -36.11 52.89
100.00000 2.61E-05 100.00000 1.09E-04
Figure 6.4 PathLoss 4.0 microwave worksheet.
systems, without being disturbed by their presence and without disturbing their activity. The immediate effect of this elegant behavior is that spread-
232
Transmission Systems Design Handbook for Wireless Networks
spectrum systems may be operated without a license, making spread-spectrum modulation the chosen technology for rapid-deployment systems. There are two basic types of spread-spectrum systemsdirect-sequence spread spectrum (DSSS) and frequency-hopping spread spectrum (FHSS). Fundamentally, DSSS is accomplished by chopping up the signal into data packets using a predefined sequence called the spreading code. Only the coded data, not the signal, is transmitted and only receivers with the matching encoding scheme can reconstruct the signal. A system with a particular encoding scheme will see any foreign encoding pattern or frequency as background noise. With FHSS, hopping the signal randomly to a different frequency several times per second spreads the signal. Both DSSS and FHSS use a pseudorandom sequence. While for DSSS the signal is mixed with the pseudorandom sequence, for FHSS the signal changes frequencies based on the pseudorandom sequence. The bandwidth of the transmitted signal is much greater than the bandwidth of the original message, and the bandwidth of the transmitted signal is determined by the message to be transmitted and by an additional signal known as the spreading code. Low power density relates to the fact that the transmitted energy is spread over a wide band, and therefore the amount of energy per specific frequency is very low. The effect of the low power density of the transmitted signal is that such a signal will not disturb (interfere with) the activity of other systems receivers in the same area. Redundancy relates to the fact that the message is (or may be) present on different frequencies from where it may be recovered in case of errors. The effect of redundancy is that spread-spectrum systems present a high resistance to noises and interference, being able to recover their messages even if noises are present on the medium. Spread-spectrum modulation techniques are composed of two consecutive modulation processes: one executed by the message to be transmitted and the other executed by the spreading code (spreading process). It is this spreading process that generates the wide bandwidth of the transmitted signal. In FHSS systems, the two modulation processes are as follows: • Process 1. The original message modulates the carrier, thus generat-
ing a narrowband signal. • Process 2. The frequency of the carrier is periodically modified (hopped) following a specific spreading code. In FHSS systems, the
Microwave Point-to-Point System Design
233
spreading code is a list of frequencies to be used for the carrier signal. The amount of time spent on each hop is known as dwell time. In DSSS systems, the two modulation processes are as follows: • Process 1. The original message is modulated by the spreading code.
In DSSS systems, the spreading code is a sequence of bits (known as chips), and the first modulation step is an exclusive-OR (XOR) operation executed between the message and the spreading code (in a process known as chipping). The result of the first modulation step is that a 0 bit of message is converted into a chip sequence representing the 0 bit, and the 1 bit of message is converted into another chip sequence, representing the 1 bit. Instead of transmitting the original message bit, a chip sequence representing the bit will be transmitted. • Process 2. The sequences representing the message bits modulate the carrier signal. Dwell time in FHSS is represented as a 3x data bit duration. Spreading sequence in DSSS is represented as being five chips long. In DSSS systems, colocation could be based on the use of different spreading codes (sequences) for each active system. For FHSS systems, IEEE 802.11 defines 79 different hops for the carrier frequency. Using these 79 frequencies, IEEE 802.11 defines 78 hopping sequences (each with 79 hops) grouped in three sets of 26 sequences each. Sequences from the same set encounter minimum collisions and they may be allocated to colocated systems. Theoretically, 26 FHSS systems may be colocated. However, as synchronization among independent systems is forbidden (synchronization would eliminate collisions), the actual number of systems that can be colocated is around 15. Figure 6.5 shows an example of the spread-spectrum microwave system application in the wireless network where the hub site is connected to the BSC/MSC site, as well as to other cell sites, by means of this kind of microwave point-to-point system. The greatest interest in spread-spectrum products has resulted from the creation of rules permitting spread-spectrum transmitters to operate on an uncoordinated basis without requiring individual user licenses. The 2.4 and 5.8 GHz bands are unlicensed bands (in North America) that can be used without any interference analysis or frequency coordination with the existing
234
Transmission Systems Design Handbook for Wireless Networks
Road coverage
Base station Urban/suburban coverage 2.4 or 5.8 GHz Hub site
Base station
MSC/end office switch
Figure 6.5 Spread-spectrum MW system.
microwave systems. Spread-spectrum MW systems should be used only as a temporary solution when rapid deployment is required or in rural areas only. The reason is that the 2.4 GHz band is also used with a number of other applications within the unlicensed ISM band, such as garage-door openers, microwave ovens, and Bluetooth systems. Special rules to take advantage of land-based spread-spectrum technology have been issued by the United States, Canada, Argentina, and the United Kingdom. The United States permits unlicensed operation with high power (1W) in three bands under Part 15.247 of the FCC Rules: 902928 MHz, 2,4002,483.5 MHz, and 5,7255,850 MHz. Argentina has followed with rules identical to current U.S. rules. Those in Canada are similar, and based on Industry Canada, Spectrum Management, RSS-210, Low Power Licence-Exempt Radiocommunication Devices. (More information about
Microwave Point-to-Point System Design
235
Industry Canada can be found on its Web page: http://strategis.ic.gc.ca/ sc_mrksv/spectrum/engdoc/spect1.html.) Most of the regulations permitting unlicensed spread spectrum penalize the use of directional antennas. Manufacturers using directional antennas are forced to reduce the transmitter output power by the gain of the antenna to comply with an effective radiated power (ERP) limit. Manufacturers using omnidirectional antennas can use the full power, usually 1W. The use of digital radio communication has increased tremendously in recent years, and the FCC and other agencies have made the ISM band available for wireless digital communication in order to satisfy the increasing demand. It is well known that microwave ovens are the most significant man-made noise source within the ISM band because of their widespread popularity. They have become a major source of interference for digital communication systems. Microwave ovens are designed for heating and cooking food through the use of electromagnetic energy at frequencies in the ISM bands, ranging from 890 to 6,000 MHz. In particular, the frequency of 2,450.50 MHz is applied to drive most ovens used for heating and cooking food in the household or commercial establishment. A primary source for oven emission comes from the leakage of microwave radiation from the door. Certainly, designs and research studies should be initiated to mitigate the effect of this man-made noise and to enhance the performance of wireless communication systems under this man-made-noise environment. Nonionizing radiation is the energy associated with electromagnetic radiation and may be classified in terms of the quantum energy required to eject or promote electrons from biological materials exposed to electromagnetic radiation. Microwave radiation has low quantum energy; under ordinary circumstances, it is too low to affect ionization, excitation, ejection, or promotion of electrons. Consequently, microwave radiation is referred to as nonionizing radiation. A new and more realistic performance standard serves to better control the emission of nonionizing radiation from microwave ovens. It will minimize human exposure and will add to the protection of the publics health and safety. Concurrently, it will provide relief to the spectral congestion in a frequency band allocated to digital radio communication. It also will help to improve the performance of many digital radio communication systems using the ISM band. From the regulatory perspective, these unlicensed bands come with two major constraints, a transmit-power limitation of essentially 1W and a minimum processing gain of 10 dB for either a frequency-hopping or directsequence system. This implies that the desired data capacity per bandwidth (in other words, bandwidth efficiency) may have to be sacrificed to achieve
236
Transmission Systems Design Handbook for Wireless Networks
the processing gain, and the total transmit power is not high enough to support multilevel quadrature-amplitude-modulation (QAM) techniques to increase the data rate. The challenge is to overcome these difficulties and still achieve sufficiently high data capacity.
6.5 Microwave Compatibility and Safety The Radiation Control for Health and Safety Act (PL90-602) was adopted by the U.S. Congress in October 1968, to protect the public from unnecessary exposure to potentially harmful radiation, which includes microwaves emitted by electronic products. The act prescribes different and individual performance standards, to the extent appropriate and feasible, for different electronic products so as to recognize their different operating characteristics and uses. The ANSI guidelines have been updated and refined several times since 2 the first frequency-independent limit of 10 mW/cm was issued. In 1986, a report from the National Council for Radiation Protection and Measurements (NCRP) provided a comprehensive evaluation of the scientific literature on the biological effects of radio-frequency electromagnetic fields (RFEMs). On the basis of this evaluation, the report made recommendations for exposure criteria (Biological Effects and Exposure Criteria for RadioFrequency Electromagnetic Fields, NCRP Report No. 86, National Council on Radiation Protection and Measurement, Bethesda, Maryland, 1986). Several key aspects of the NCRP recommendations were incorporated subsequently into the ANSI/IEEE C95.1-1992 guidelines. These guidelines and recommendations formed the basis for mandatory standards adopted by several federal agencies, including the FCC. They are also used extensively by standards-development organizations throughout the world. The FCCs guidelines establish separate maximum-permissibleexposure (MPE) limits for general population/uncontrolled exposure and for occupational/controlled exposure. The general population/uncontrolled limits set the maximum exposure to which most people may be subjected. People in this group include members of the general public not associated with the installation and maintenance of the transmitting equipment. Higher exposure limits are permitted under the occupational/controlled exposure category, but only for persons who are exposed as a consequence of their employment (e.g., wireless radio engineers or technicians). To qualify for the occupational/controlled exposure category, exposed persons must be made fully aware of the potential for exposure (e.g., through training), and they
Microwave Point-to-Point System Design
237
must be able to exercise control over their exposure. In addition, people passing through a location, who are made aware of the potential for exposure, may be exposed under the occupational/controlled criteria. The MPE limits adopted by the FCC for occupational/controlled and general population/uncontrolled exposure incorporate a substantial margin of safety and have been established to be well below levels generally accepted as having the potential to cause adverse health effects. Determining whether a potential health hazard could exist with respect to a given transmitting antenna is not always a simple matter. Important questions must be considered in making that determination the following: • What is the frequency of the RF signal being transmitted? • What is the operating power of the transmitting station and what is
the actual power radiated from the antenna?
• How long will someone be exposed to the RF signal at a given dis-
tance from the antenna?
• What other antennas are located in the area, and what is the expo-
sure from those?
For all frequency ranges at which FCC licensees operate, the FCCs rules establish MPE limits. The MPE limits vary by frequency because of the different absorptive properties of the human body at different frequencies when exposed to whole-body RF fields. The FCC establishes MPE limits in terms of electric field strength, magnetic field strength, and far-field equivalent power density. For most frequencies used by wireless services, the most relevant measurement is power density. The MPE limits for power density 2 are given in terms of milliwatts per square centimeter, or mW/cm . In terms of power density, for a given frequency the FCC MPE limits can be interpreted as specifying the maximum rate that energy can be transferred (i.e., the power) to a square centimeter of a persons body over a period of time. In practice, however, since it is unrealistic to measure separately the exposure of each square centimeter of the body, actual compliance with the FCC limits on RF emissions should be determined by spatially averaging a persons exposure over the projected area of an adult human body. Electric field strength and magnetic field strength are used to measure near field exposure. At frequencies below 300 MHz, these are typically the more relevant measures of exposure, and power density values are given primarily for reference purposes. However, evaluation of far-field equivalent
238
Transmission Systems Design Handbook for Wireless Networks
power density exposure may still be appropriate for evaluating exposure in some cases. For frequencies above 300 MHz, only one field component need be evaluated, and exposure is usually more easily characterized in terms of power density. Transmitters and antennas that operate at 300 MHz or lower include radio broadcast stations, some television broadcast stations, and certain personal wireless service facilities (e.g., some paging stations). Most personal wireless services, including all cellular and PCS, as well as some television broadcast stations, operate at frequencies above 300 MHz. As noted above, the MPE limits are specified as time-averaged exposure limits. This means that exposure can be averaged over the identified time interval (30 minutes for general population or uncontrolled exposures, or 6 minutes for occupational or controlled exposures). However, for the case of exposure of the general public, time averaging is usually not applied because of uncertainties over exact exposure conditions and difficulty in controlling time of exposure. Therefore, the typical conservative approach is to assume that any RF exposure to the general public will be continuous. The FCCs limits for exposure at different frequencies are shown in Table 6.3. Finally, it is important to understand that the FCCs limits apply cumulatively to all sources of RF emissions affecting a given area. A common example is where two or more wireless operators have agreed to share the cost of building and maintaining a tower and to place their antennas on that joint structure. In such a case, the total exposure from the two facilities taken together must be within the FCC guidelines. It has been determined through calculations and technical analysis that due to their low power or height above ground level, many facilities by their very nature are highly unlikely to cause human exposures in excess of the guideline limits. Operators of those facilities are exempt from routinely having to determine compliance, and facilities with these characteristics are considered categorically excluded from the requirement for routine environmental processing for RF exposure. If a facility is categorically excluded, an applicant or licensee may ordinarily assume compliance with the guideline limits for exposure. However, an applicant or licensee must evaluate and determine compliance for a facility that is otherwise categorically excluded if specifically requested to do so by the FCC. No radio or television broadcast facilities are categorically excluded. Thus, broadcast applicants and licensees must affirmatively determine their facilitys compliance with the guidelines before construction, and upon every facility modification or license renewal application. With respect to personal wireless services, a cellular facility is categorically excluded if the total ERP of all channels operated by the licensee at a
Microwave Point-to-Point System Design
239
Table 6.3 Limits for the Maxium Permissible Exposure (MPE) (A) Limits for Occupational/Controlled Exposure
Frequency Range (MHz)
Electric Field Strength (E) (V/m)
Magnetic Field Strength (H) (A/m)
Power Density (S) (mW/cm2)
Averaging Time |E|2, |H|2 or S (minutes)
0.33.0
614
1.63
(100)*
6
3.030
1,842/f
4.89/4
(900/f2)*
6
30300
61.4
0.163
1.0
6
3001,500
f/300
6
1,500100,000
5
6
(B) Limits for General Population/Uncontrolled Exposure
Frequency Range (MHz)
Electric Field Strength (E) (V/m)
Magnetic Field Strength (H) (A/m)
Power Density (S) (mW/cm2)
Averaging Time |E|2, |H|2 or S (minutes)
0.31.34
614
1.63
(100)*
30
2
1.3430
824/f
2.19/f
(180/f )*
30
30300
27.5
0.073
0.2
30
3001,500
f/1500
30
1,500100,000
1.0
30
f = frequency in MHz. * = Plane-wave equivalent power density. Note 1: Occupational/control limits apply in situations in which persons are exposed as a consequence of their employment provided those persons are fully aware of the potential for exposure and can exercise control over their exposure. Limits for occupational/controlled exposure also apply in situations when an individual is transient through a location where occupational/controlled limits apply provided he or she is made aware of the potential for exposure. Note 2: General population/uncontrolled exposures apply in situations in which the general public may be exposed, or in which persons that are exposed as a consequence of their employment may not be fully aware of the potential for exposure or cannot exercise control over their exposure.
240
Transmission Systems Design Handbook for Wireless Networks
site is 1,000W or less. If the facility uses sectorized antennas, only the total effective radiated power in each direction is considered. In addition, a cellular facility is categorically excluded, regardless of its power, if it is not mounted on a building and the lowest point of the antenna is at least 10m (about 33 ft) above ground level. A broadband PCS antenna array is categorically excluded if the total ERP of all channels operated by the licensee at a site (or all channels in any one direction, in the case of sectorized antennas) is 2,000W or less. Like cellular, another way for a broadband PCS facility to be categorically excluded is if it is not mounted on a building and the lowest point of the antenna is at least 10m (about 33 ft) above ground level. The power threshold for categorical exclusion is higher for broadband PCS than for cellular, because broadband PCS operates at a higher frequency where exposure limits are less restrictive. At present, the relevant safety guideline for human exposure in the 2 United States has been reduced to f /1,500 mW/cm , where f is expressed in MHz. For example, in case of 2,450 MHz the safety guideline translates to 1.67 mW/cm2. In point-to-point microwave radio systems, the only danger for the technicians is during the installation and testing of live systems, and only in cases where they spend extended time directly in front of the microwave parabolic antenna. Due to the fact that the microwave antennas are installed high above ground level and are highly directive, there is a very low (practically insignificant) level of the microwave radiation on the ground affecting any living creatures. In other words, point-topoint microwave radios use low-power transmitters and narrow, directed antenna beams (parabolic dish antennas) to reduce radiation levels of many orders of magnitude to well within safety limits. Therefore, MW systems are not dangerous to the public and do not present a safety or a health hazard.
6.6 Coordinate Systems, Datums, and GPS 6.6.1
About Datums
Geodetic datum (singular of the Latin word data, meaning given things) defines the reference systems that describe the size and shape of the Earth. Hundreds of different datums have been used to frame position descriptions since Aristotle made the first estimates of Earths size, and datums have evolved from those describing a spherical body to ellipsoidal models derived from years of satellite measurements. Geodetic datums and the coordinate reference systems based on them were developed to describe geographic
Microwave Point-to-Point System Design
241
positions for surveying, mapping, and navigation. Through a long history, the figure of the Earth was refined from flat models to spherical models of sufficient accuracy to allow for global exploration, navigation, and mapping. True geodetic datums were employed only after the late 1700s, when measurements showed that the Earth was ellipsoidal in shape. Modern geodetic datums range from flat Earth models used for plane surveying to complex models used for international applications, which completely describe the size, shape, orientation, gravity field, and angular velocity of the earth. Cartography, surveying, navigation, and astronomy all make use of geodetic datums. Referencing geodetic coordinates to the wrong datum can result in position errors of hundreds of meters. Different nations and agencies use different datums as the basis for coordinate systems used to identify positions in geographic information systems, precise positioning systems, and navigation systems. The diversity of datums in use today and the technological advancements that have made possible global positioning measurements with submeter accuracy requires careful datum selection and careful conversion between coordinates in different datums. Datum types include horizontal, vertical, and complete datums. The Global Positioning System (GPS) is based on the World Geodetic System 1984 (WGS-84). Coordinate values resulting from interpreting latitude, longitude, and height values based on one datum as though they were based in another datum can cause position errors in three dimensions of more than 1 km. Datum conversions are accomplished by various methods. Complete datum conversion is based on seven parameter transformations that include three translation parameters, three rotation parameters, and a scale parameter. Simple three-parameter conversion between latitude, longitude, and height in different datums can be accomplished by conversion through Earth-centered, Earth-fixed, XYZ Cartesian coordinates in one reference datum and three origin offsets that approximate differences in rotation, translation, and scale. Parameters for simple XYZ conversion between many datums and WGS-84 are published by the Defense Mapping Agency.
6.6.2
Geometric Earth Models
Early ideas of the figure of the Earth resulted in descriptions of the Earth as an oyster (the Babylonians, before 3000 B.C.), a rectangular box, a circular disk, a cylindrical column, a spherical ball, and a very round pear (Columbus, in the last years of his life).
242
Transmission Systems Design Handbook for Wireless Networks
Flat Earth models are still used for plane surveying, over distances short enough so that the Earth, curvature is insignificant (less than 10 km). Spherical Earth models represent the shape of the Earth with a sphere of a specified radius. Spherical Earth models are often used for short-range navigation (VOR-DME) and for global distance approximations. Spherical models fail to model the actual shape of the Earth. The slight flattening of the Earth at the poles results in about a 20-km difference at the poles between an average spherical radius and the measured polar radius of the Earth. Ellipsoidal Earth models are required for accurate range and bearing calculations over long distances. Loran-C and GPS navigation receivers use ellipsoidal Earth models to compute positions. Ellipsoidal models define an ellipsoid with an equatorial radius and a polar radius. The best of these models can represent the shape of the Earth over the smoothed, averaged seasurface to within about 100m. 6.6.3
Reference Ellipsoids and Coordinate Systems
Reference ellipsoids are usually defined by semimajor (equatorial radius) and flattening (the relationship between equatorial and polar radii). Other reference ellipsoid parameters, such as semiminor axis (polar radius) and eccentricity, can be computed from these terms. The Earth has a highly irregular and constantly changing surface. Models of the surface of the Earth are used in navigation, surveying, and mapping. Topographic and sea-level models attempt to model the physical variations of the surface, while gravity models and geoids are used to represent local variations in gravity that change the local definition of a level surface. The topographical surface of the Earth is the actual surface of the land and sea at some moment in time. Aircraft navigators have a special interest in maintaining a positive height vector above this surface. Sea level is the average (methods and temporal spans vary) surface of the oceans. Tidal forces and gravity differences from location to location cause even this smoothed surface to vary over the globe by hundreds of meters. Gravity models attempt to describe in detail the variations in the gravity field. The importance of this effort is related to the idea of leveling. Plane and geodetic surveying uses the idea of a plane perpendicular to the gravity surface of the Earth, the direction perpendicular to a plumb bob pointing toward the center of mass of the Earth. Local variations in gravity, caused by variations in the Earths core and surface materials, cause this gravity surface to be irregular.
Microwave Point-to-Point System Design
243
Geoid models attempt to represent the surface of the entire Earth over both land and ocean as though the surface resulted from gravity alone. Bomford described this surface as the surface that would exist if the sea was admitted under the land portion of the Earth by small frictionless channels. The WGS-84 geoid defines geoid heights for the entire Earth, and the U.S. National Imagery and Mapping Agency (formerly the Defense Mapping Agency) publishes a 10° × 10° grid of geoid heights for the WGS-84 geoid. By using a four-point linear interpolation algorithm at the four closest grid points, the geoid height for any location can be determined. The same grid can be used to produce a contour map of geoid heights for the globe. The National Imagery and Mapping Agency publishes a 0.25° model of the WGS-84 geoid. Coordinate systems to specify locations on the surface of the Earth have been used for centuries. In western geodesy, the equator, the tropics of Cancer and Capricorn, and then lines of latitude and longitude were used to locate positions. Eastern cartographers used other rectangular grid systems as early as A.D. 270, and various units of length and angular distance have been used throughout history. The meter is related to both linear and angular distance, having been defined in the late eighteenth century as one ten-millionth of the distance from the pole to the equator. The most commonly used coordinate system today is the latitude, longitude, and height system. The prime meridian and the equator are the reference planes used to define latitude and longitude. The geodetic latitude (there are many other defined latitudes) of a point is the angle from the equatorial plane to the vertical direction of a line normal to the reference ellipsoid. The geodetic longitude of a point is the angle between a reference plane and a plane passing through the point, both planes being perpendicular to the equatorial plane. The geodetic height at a point is the distance from the reference ellipsoid to the point in a direction normal to the ellipsoid. Earth-centered, Earth-fixed X, Y, and Z Cartesian coordinates are also used to define three-dimensional positions with respect to the center of mass of the reference ellipsoid. The Z-axis points toward the North Pole. The X-axis is defined by the intersection of the plane defined by the prime meridian and the equatorial plane. The Y-axis completes a right-handed orthogonal system by a plane 90° east of the X-axis and its intersection with the equator. Hundreds of geodetic datums are in use around the world.
244
6.6.4
Transmission Systems Design Handbook for Wireless Networks
GPS
6.6.4.1 About GPS
GPS is based on the information users receive from GPS satellites. The purpose of GPS is to provide its users with the ability to compute their location in three-dimensional space [19, 20]; in order to do that, the receiver must be able to lock on to signals from at least four different satellites. Moreover, the receiver must maintain a lock on each satellites signal long enough to receive the information encoded in the transmission. Achieving and maintaining a lock on four or more satellite signals can be impeded, because the signal is transmitted at 1.575 GHz, a frequency too high to bend around or pass through solid objects in the signals path. This is the reason that GPS receivers cannot be used indoors. Outdoors, tall buildings, dense foliage, or terrain that stands between a GPS receiver and a GPS satellite will block that satellites signal. It has to be noted that the inherent accuracy of the GPS is better than 100m. Also, the signals from the GPS satellites have been degraded intentionally by the U.S. Department of Defense (DOD) for the purpose of national security. This performance degradation is known as selective availability (SA), and only DOD-approved users have access to satellite signals without SA. However, a policy statement recently issued by the White House indicates that SA will be turned off before 2006, and the accuracy of GPS position fixes will improve significantly. Authorized users with cryptographic equipment and keys and specially equipped receivers use the Precise Positioning System (PPS). U.S. and Allied military, certain U.S. government agencies, and selected civil users specifically approved by the U.S. government can use the PPS. PPS Predictable Accuracy • Horizontal accuracy of 22m; • Vertical accuracy of 27.7m; • Time accuracy of 100 ns.
Civil users worldwide use the Standard Positioning Service (SPS) without charge or restrictions. Most receivers are capable of receiving and using the SPS signal. Today, the DOD still intentionally degrades the SPS accuracy by the use of selective availability.
Microwave Point-to-Point System Design
245
SPS Predictable Accuracy • Horizontal accuracy of 100m; • Vertical accuracy of 156m; • Time accuracy of 340 ns.
These GPS accuracy figures are from the 1994 Federal RadioNavigation Plan. The figures are 95% accurate; for horizontal accuracy figures, 95% is the equivalent of 2 rms (two-distance root-mean-squared), or twice the radial-error standard deviation. For vertical and time errors, 95% is the value of two standard deviations of vertical error or time error. SA was officialy turned off as of May 2001. Typical commerical GPS positioning accuracy will be 10 to 20m depending on the GPS receiver and the configuration of the satellite constellation. 6.6.4.2 Differential GPS Techniques
The idea behind differential positioning is to correct bias errors at one location with measured bias errors at a known position. A reference receiver, or base station, computes corrections for each satellite signal. Because individual pseudoranges must be corrected prior to the formation of a navigation solution, differential GPS (DGPS) implementations require software in the reference receiver that can track all SVs in view and form individual pseudorange corrections for each SV. These corrections are passed to the remote, or rover, receiver that must be capable of applying these individual pseudorange corrections to each SV used in the navigation solution. Applying a simple position correction from the reference receiver to the remote receiver has limited effect at useful ranges. This is due to the fact that both receivers would have to be using the same set of SVs in their navigation solutions and have identical GDOP terms (not possible at different locations) to be identically affected by bias errors. Differential corrections may be used in real time or later, with postprocessing techniques. Real-time corrections can be transmitted by radio link. The U.S. Coast Guard maintains a network of differential monitors and transmits Differential Code GPS (DGPS) corrections over radio beacons covering much of the U.S. coastline for the purpose of navigation. DGPS corrections are often transmitted in a standard format specified by the Radio Technical Commission Marine (RTCM). Corrections can be recorded for postprocessing. Many public and private agencies record DGPS corrections for distribution by electronic means. Private DGPS services use leased FM
246
Transmission Systems Design Handbook for Wireless Networks
subcarrier broadcasts, satellite links, or private radio beacons for real-time applications. To remove SA (and other bias errors), differential corrections should be computed at the reference station and applied at the remote receiver at an update rate that is less than the correlation time of SA. Suggested DGPS update rates are usually less than 20 seconds. DGPS removes common-mode errors, those errors common to both the reference and remote receivers (not multipath or receiver noise). Errors are more often common when receivers are close together (less than 100 km). Differential position accuracies of 1 to 10m are possible with DGPS. Differential carrier GPS is used for surveys. All carrier-phase tracking is differential, requiring both a reference and remote receiver tracking carrier phases at the same time. In order to estimate the number of carrier wavelengths at the reference and remote receivers correctly, they must be close enough to ensure that the ionospheric delay difference is less than a carrier wavelength. This usually means that carrier-phase GPS measurements must be taken with a remote and reference station within about 30 km of each other. Special software is required to process carrier-phase differential measurements. Newer techniques, such as real-time-kinematic (RTK) processing, allow for centimeter relative positioning with a moving remote receiver. 6.6.4.3 GPS Techniques and Project Costs
Receiver costs vary depending on capabilities. Small civil SPS receivers can be purchased for under $200; some can accept differential corrections. Receivers that can store files for postprocessing with base station files cost more ($2,000$5,000). Receivers that can act as DGPS reference receivers (computing and providing correction data) and carrier-phase-tracking receivers (and two are often required) can cost many thousands of dollars ($5,000$40,000). Military PPS receivers may cost more and can be difficult to obtain. Other costs include the multiple receivers (when needed), postprocessing software, and specially trained personnel. Project tasks can often be categorized by required accuracy, which will determine equipment cost: • Low-cost, single-receiver SPS projects (100m accuracy); • Medium-cost, differential SPS code positioning (110m accuracy); • High-cost, single-receiver PPS projects (20m accuracy); • High-cost, differential-carrier-phase surveys (1 mm to 1 cm accuracy).
Microwave Point-to-Point System Design 6.6.5
247
Useful Facts to Remember • Coordinates can be taken from topographical maps, surveys, or GPS. • In North America, coordinates are based on NAD27 or NAD83
(North American datum).
• NAD27 is based on Clarke ellipsoid 1866 and uses a stone in Meades
Ranch in Kansas as a reference point.
• The NAD83 Geodetic reference system is based on satellites. • All older North American topographical maps are based on NAD27. • NGVD29 is a National Geodetic Vertical Datum. • NAVD88 is a North American Vertical Datum. • No universal formulas exist for the conversion between NAD27 and
NAD83.
• Ordinary GPS is usually not precise enough for the MW design, fre-
quency coordination, and licensing process. • Three satellites can be used for a two-dimensional position fix, but four are required for height above terrain (HAT).
6.7 Managing the MW Radio Network 6.7.1
Introduction
Microwave radios are sometimes used in large communications networks where network management must deal with particular network configurations: • Several hundred sites to be monitored; • Several hundred subnetworks not interconnected; • Fast notification of alarms when problems occur on any site of any
subnetwork; • The ability to manage radios from other NMS providers.
The domain manager application, hereafter referred to as the domain manager, is an on-site software tool for management and support of the digital radio relay link. It is also often called craft interface. Its software runs on a PC-based workstation connected to the indoor MW unit via a serial communications interface. The domain manager provides extensive element-
248
Transmission Systems Design Handbook for Wireless Networks
management functions on site and, via the microwave radio link, can be used to access any other link in a chain of stations. The domain manager can also be connected to the on-site Ethernet LAN, with transparent communications to and from any particular link whose indoor unit (IU) is connected to the LAN (as in the case of a splitconfiguration microwave system). The usual application software modules are as follows: • Element-management application; • Alarm-management application; • Configuration-management application; • Performance-monitoring application.
The element-management application allows the user to access, monitor, and control the radio stations in the network domain. In a multihop link system, any of the associated stations in the domain can be accessed. In a simple, single-hop link, only the near- and far-end stations can be accessed. All elements of information maintained within the management information base (MIB) of the selected radio station are accessible to the operator. The operator is alerted to alarms occurring within the domain by means of color changes affecting the relevant radio stations shown on the display list. The radio station signaling an alarm can then be accessed at will to establish the cause of the alarm condition and to determine appropriate steps to correct the condition. The alarm-management application monitors alarms from all radio stations within the network domain to which it has access. It also logs these alarms in a local or network database. Proper alarm-management procedures are enforced by this application in the sense that operators will have to formally acknowledge each alarm and report on the treatment thereof before being allowed to clear it. In this manner a full alarm audit trail can be established for all radio stations within the particular network domain. The alarm-management application contains a subsystem specifically designated to be adapted to control external alarm annunciation interfaces that may be specific to a particular service providers TMN or any other environment. The configuration-management application allows an operator to back up or restore the configuration of any particular radio station within the network domain to which it has access. The configuration information maintained by this application includes operational configuration parameters as
Microwave Point-to-Point System Design
249
well as hardware and software build state parameters. The configurationmanagement application, besides allowing operators to restore the operational configuration state of radio stations in case it should become corrupted, is indispensable to service providers in whose organizations proper asset management is a prerequisite. The performance-monitoring application monitors measured, as well as statistical, performance parameters of all radio stations within the network domain to which it has access. It then logs this information in a local/ network database in a format that will facilitate the graphical display of this information using standard spreadsheet application programs. The variation in the values of parameters such as power supply voltages, outdoor MW unit temperature, received signal strength, and so on, over time can thus be subjected to graphical trend analysis. The ability to perform trend analysis is especially useful for preventative maintenance. The microwave radio craft-interface system can be integrated into another open network management platform that oversees the overall telecommunications transmission network and has full remote management capabilities. The optional network-management features include a simple network management protocol (SNMP) network agent in the microwave equipment. This allows a remote NMS to monitor and control the microwave system and to access its MIB. This includes G.826 performance monitoring information. In a similar manner, a proxy server can be configured to support other networking standards, such as Q3. Network links to the MW system can be any of the following: dial-up modem, leased line, remote via a radio link, via an on-site LAN, or by a WAN. 6.7.2
Managing a Microwave Network with SNMP
Typical microwave networks are composed of one or a combination of the network topologies shown in Figure 6.6, including meshed networks, where each node can be connected to any number of other nodes. 4 1
2
3
3 1
3 5
Figure 6.6 Linear, ring, and star topologies.
2 4
250
Transmission Systems Design Handbook for Wireless Networks
Although three nodes are theoretically sufficient to form the ring, due to the high-low violation, there must be even numbers of MW hops in the ring. In order to pass traffic and telemetry information, each radio has up to three connection points: the radio frequency (RF) link and the two wired connections (repeater in, repeater out). When three or more radios require interconnection at the same node, this may be implemented by daisy chaining the IDUs to form a ring. In typical wireless deployments, cells are established through an urban city. Each of these cells carries the call traffic of only its cell. In order to link the cells together and to the PSTN, microwave radios are used. The radio equipment is installed in the RBS at the center of the cell (hub site). Network management is used to monitor and control the radios of the network (Figure 6.7). In order to access and communicate with radios for control and monitoring functions, each radio within the network can be configured to have a unique identification number. This number can be in the form of the network element (NE) number or the IP address.
RBS 1
RBS 6 RBS 4 RBS RBS
3
RBS
2
5 6 1
Network management
Figure 6.7 Microwave network NMS.
1
2 2
3
3
4 8
12
9
5
4 7
6
5
10 11
Microwave Point-to-Point System Design
251
SNMP Overview
The SNMP protocol was developed to provide a basic, easy-to-implement network-management tool that could be common to the network equipment and, yet, less restrictive of the management system used. SNMP allows the monitoring and controlling of radios by any network manager that supports the protocol (Figure 6.8). In addition, SNMP provides a quick and dynamic response to alarm conditions, which accelerates alarm notification and reduces the need for polling, but requires embedded radio software. The second method of implementing SNMP in a network of radios using a proprietary protocol is by inserting a proxy agent between the network manager and the radio network. This proxy agent is usually in the form of a computer and performs the following three functions: • Translates the proprietary protocol messages into SNMP (and vice
versa);
• Sequentially polls each radio of the network to obtain status
information;
• Reports alarm conditions to the network manager.
This type of arrangement is usually slower and still requires polling of every radio in the network, but the advantage is that it can convert legacy networks into an SNMP network. The SNMP network is made up of the following: • SNMP manager;
SNMP network
SNMP network manager
Embedded SNMP agent in all radios
Figure 6.8 SNMP network.
252
Transmission Systems Design Handbook for Wireless Networks
• SNMP agent; • MIB.
The SNMP manager uses SNMP to manage devices equipped with an SNMP agent such as microwave radios as well as other types of products. The management function is in the form of controlling and monitoring the network devices in a centralized fashion. It also serves as the man-machine interface to all network devices. The SNMP agent is the implementation of a network management protocol that exchanges network-management information with the network management station. The SNMP agent resides in each element (radio) of the network. The function of the agent is to respond to requests, such as the following, from the manager: • GET: Enables the management station to retrieve the value of
objects at the agent, such as status or configuration information;
• SET: Enables the management station to set the value of objects at
the agent, such as configuration parameters;
• TRAP: Enables an agent to notify the management station to signify
events; these messages are used to report unsolicited events, such as alarms.
SNMP traps, which are unsolicited alerts (alarms), are generated under the conditions listed below. It should be noted that traps are not generated for all of the radio alarms. A minor alarm does not generate a TRAP message. This is to avoid congestion of network messages that do not require immediate intervention. MIB is a structured collection of variables that can be monitored and managed by the network manager. This database is made up of all of the radios parameters, each parameter identified by a unique variable name that is used by the network manager. During the configuration phase, the MIB file is first executed on the network manager station. This action informs the network manager of the parameters that can be accessed. Direct Connection
When the network manager is colocated with a radio terminal (Figure 6.9) and only one radio subnet is to be monitored, then a direct connection (as illustrated) between the manager and the radio is possible. A different type of NMS direct connection, utilizing IP addressing, is shown in Figure 6.10.
Microwave Point-to-Point System Design
253
MW radio network SNMP network manager
Direct connection or dedicated line RS-232
Figure 6.9 Direct NMS connection.
198.105.19.32
198.105.16.218 198.105.16.219 198.105.16.217 198.105.16.220 (router) 3 1
2 5
NMS
198.105.16.222
198.105.16.32 PPP 198.105.16.9
4
5
6
198.105.16.222
1
2
6
IP
IP
IP
IntNet
IntNet
IntNet
198.105.16.217
198.105.16.218
198.105.16.222 MW radio subnetwork
Figure 6.10 Direct connection and IP addressing.
Other NMS Connection Methods
When the network manager is not located near the radio terminal equipment, then a connection via modems (PSTN connection) is possible. A terminal server with built-in modems permits the access to several subnetworks.
254
Transmission Systems Design Handbook for Wireless Networks
The network manager can also be linked to radio terminal equipment via a LAN or WAN. In the case of the LAN connection, a terminal server is required at the radio end to convert the Ethernet to RS-232.
References [1] ITU-Recommendations of the CCIR, Fixed Service Using Radio-Relay Systems, Vol. 9, Part 1, 1990. [2] ITU, Handbook of Radiometeorology, Geneva, Switzerland, 1996. [3] ITU HandbookDigital Radio-Relay Systems, Geneva, Switzerland, 1996. [4] Vigants, A., Space Diversity Engineering, Bell System Technical Journal, Vol. 54, No. 1, January 1975. [5] ITU-R Rep. 338-6, Propagation Data and Prediction Methods Required for Terrestrial Line-of-Sight Radio-Relay Systems, Geneva, Switzerland, 1990. [6] Crane, R.K., Prediction of Attenuation by Rain, Proc. IEEE Trans. on Communications, Vol. Com-28, No. 9, September 1980. [7] ITU-R Rec. PN.837-1, Characteristics of Precipitation for Propagation Modeling, Geneva, Switzerland, 1994. [8] Mojoli, L.F., and U. Mengali, Propagation In Line of Sight Radio Links (Part II Multipath Fading, Telletra Review, 1988. [9] Mojoli, L.F., and U. Mengali, Propagation In Line of Sight Radio Links (Part I Visibility, Reflections, Blackout), Telletra Review, 1988. [10] ITU-R Rec. 836, Surface Water Vapor Density, Geneva, Switzerland, 1992. [11] ITU-R Rec. F.557-3, Availability Objective for Radio-Relay Systems, Geneva, Switzerland, 1994. [12] Rummler, W. D., Advances in Microwave Radio Route Engineering for Rain, ICC 87, Seattle, WA, June 1987. [13] Andrew Corporation, Catalog 38. [14] Rec. ITU-R SF.1006, Determination of the Interference Potential Between Earth Stations of the Fixed-Satellite Service and Stations in the Fixed Service. [15] GTE Lencurt, Inc., Engineering Considerations for Microwave Communications Systems, 1970. [16] Ivanek, F., Terrestrial Digital Microwave Communications, Norwood, MA: Artech House, 1989. [17] Henne, I., and P. Thorvaldsen, Planning of Line-of-Sight Radio Relay Systems, NERA Telecommunications, Second Edition, 1999.
Microwave Point-to-Point System Design
255
[18] Contract Telecommunications Engineering, PathLoss 4.0, User Manual. [19] Abbott, E., and D. Powell, Land-Vehicle Navigation Using GPS, Proceedings of the IEEE, Vol. 87, No. 1, January 1999. [20] Herring, T. A., Geodetic Applications of GPS, Proceedings of the IEEE, Vol. 87, No. 1, January 1999.
7 Transmission-Network Planning and Design 7.1 Overview The U.S. telecommunications industry was changed forever on February 1, 1996, with the passage of the Telecommunications Act. Telecommunications providers are now offered much more opportunity with fewer governmental restrictions, and yet this opportunity comes with the threat of increased competition and the need for greater focus on improved customer acquisition and support. Todays deregulated, competitive environment calls for solutions that provide a sustainable advantage in acquiring new customers and retaining existing customers. This is especially important to wireless carriers who bid large amounts of money for licenses, compete with entrenched and new competitors, and provide innovative services that are supported by increasingly sophisticated networks. Today, many wireless telecommunications providers are thinking of how to do the following: • Expand their footprint by finding and securing viable new markets
through licenses or strategic partnerships; • Plan, build, and maintain a cellular, PCS, enhanced specialized mobile radio (ESMR), also known as intergrated dispatch enhanced network (iDEN), or other broadband network; 257
258
Transmission Systems Design Handbook for Wireless Networks
• Provide improved customer service while reducing churn and pre-
venting fraud;
• Market and sell their network and services to new and existing
customers.
In order to do these things, telecommunications providers, regardless of the market or country, must take the following steps: • Quickly and accurately estimate market potential to generate greater
return on investments made in costly licensing and infrastructure;
• Focus deployment on those areas within a market that generate
higher returns sooner;
• Increase quality and number of services offered by designing and
building networks faster;
• Reduce costs and increase speed of finding, acquiring, and building
cell sites;
• Visualize competitive threats or opportunities and identify potential
alliances in existing market areas;
• Decrease the amount of time between reported trouble and penalties
(from immediate credit to improvements in the network);
• Target marketing campaigns based on geographic profiles of cus-
tomers in a wireless network to attract new customers and increase sale of products;
• Share data among marketing, customer service, and engineering
departments for faster, more integrated decision making that benefits the entire enterprise.
Wireless networks are expensive in terms of the amount of labor needed to plan and build them, the amount of information that has to be generated, and the cost of equipment. Telecommunications companies will have to try to save time and money and ensure that the network is built cost-effectively and efficiently. Outsourcing engineering design and implementation services is becoming more popular every day. Wireless networks are increasingly complex (Figure 7.1), and different types of expertise are involved in their planning, design, and implementation (deployment).
Transmission-Network Planning and Design
PSTN SS7 ISUP links (STP, tandem)
PSTN (LEC, IXC)
SS7 gateway
LE, AT
MSC-to-PSTN voice trunks
Enhanced service platforms (VMS, PPS) ESP
MSC-to-BSC A-Interface
259
VLR
MSC
SS7 network
STP Signaling links HLR
BSC BSC
ANSI-41 links to other MSC/HLR
WIN
Wireless Digital TCP/IP communication Intelligent network (data) Network (SCP, SN/IP)
Transmission network
Service regions
Microwave radio links Fiber rings Copper cables
RBS
Figure 7.1 Wireless network architecture.
7.2 General Wireless-Network Planning and Design Principles 7.2.1
Identifying the Opportunity and Strategic Planning
The two most important elements of strategic decision making are determining the opportunities available to the enterprise and implementing a plan to take advantage of those opportunities. This is most critical to wireless carriers that may face as many as eight competitors in their geographic market offering wireless voice and data. The first order of business is finding unfulfilled needs or services, and identifying products inadequately supplied in the marketplace. In the wireless industry, it may be markets showing rapid growth in population, where
260
Transmission Systems Design Handbook for Wireless Networks
local wire line telephone companies are unable to provide basic telephone service (WLL). It may also be populations where the shift is away from static jobs and workplace environments to more mobile, professional types of employment (mobile wireless systems). The actual installation of a telecommunications networkthe wireless access links to users, the fiber-optic backbone, the microwave systems, the switchesis only one part of delivering a profitable and reliable service. Equally important are all the support systems that actually turn that network into a viable business. These include systems to bill users, to create and manage services across the network, to monitor failures of the network and expedite repairs, and to plan traffic capacities to ensure the optimum use of equipment. Previously, each support system operated independently and could only communicate with others with great difficulty, causing major inefficiencies or a reliance on expensive human labor. New systems using object-oriented software techniques mean that changes in one part of the system or network cause automatic updates across all related systems, creating a nearly seamless environment for service creation, delivery, billing, and customer relationship management. In most countries, telecommunications was traditionally seen as of such strategic importance to the nation that services were provided by state-owned utilities. As the political philosophies of much of the world shifted away from state control and ownership and toward private enterprise in the 1980s and 1990s, these existing telecommunications monopolies were challenged, and regulatory regimes were liberalized. In addition, many developing countries recognized the importance of telecommunications in underpinning economic and social growth and encouraged inward investment by new operators. Competition has usually been introduced to once-closed markets in a series of standard phases beginning with the opening up of the commercial data-communications services market. The next step would be the entry of mobile operators, the development of competition in long-distance and international services, and the freedom for cable TV (CATV) entertainment providers to offer telephony services over cable networks. In many parts of the world, the final stage has been reached, with operators of all types able to compete in providing domestic voice telephony services to the mass market. Complementing this, parts of the radio spectrum have been opened up from military and other closed-group use and are being auctioned off by governments for fixed and mobile voice, data, and multimedia communications services. The political shift toward open markets has also driven the increasing globalization of world trade and the development of multinational corp-
Transmission-Network Planning and Design
261
orations. For these to operate effectively across different time zones and regions, telecommunications is crucial. This factor has driven technological, regulatory, and infrastructure development, with business location often dictated by the quality of local communications. In turn, increasing competition has led to a greater emphasis on speed of response, time to market, and the integration of the activities of distributed workgroups, workers, and outsourced suppliers. Whether it is the highly mobile international executive or the home-based teleworker, it is telecommunications that has made these new ways of doing business possible. Incumbent telecommunications operators, while often facing part privatization, have also confronted intense pressures as a result of the emergence of leaner and fitter competitors, usually using new technologies, while they themselves have the inheritance of aging legacy networks and systems. Major changes in their strategies and organizational structures have had to be engineered at short notice, and the traditional public service utility culture of most existing operators has undergone a shift toward a business-oriented, customer-focused environment. As open-market concepts have spread, a growing array of alliances, partnerships, and investments has sprung up to provide both national and international telecommunications services to business and domestic users. Many operators are now seeking to add value to their services to customers and differentiate themselves in increasingly crowded markets. Once opportunities are identified, it is important to identify and assess the competition. For example, PCS licenses were auctioned based on definitions of major trading areas (MTAs) and BTAs, while cellular licenses were awarded via lottery based on metropolitan statistical areas (MSAs) and rural statistical areas (RSAs). No single wireless company provides ubiquitous national coverage. A wireless operator must establish roaming agreements with other carriers to provide subscribers coverage while roaming. To complicate matters, there are multiple technology standards including GSM, WCDMA, TDMA, CDMA2000, and others. By layering ownership information with choice of technology, a wireless operator can learn what possible relationships are viable and the relevance of those relationships. This can lead to better decisions on the choice of the operators own technology and create opportunity rather than the alternative: isolation and failure. Once market goals for the future wireless network have been established, a thorough analysis of potential technologies is performed. Technology assessment compares each technologyvendor performance scenario against the established market goals. Once completed, the technology assessment is a strategic plan involving a complex
262
Transmission Systems Design Handbook for Wireless Networks
trade-off between the advantages, limitations, and risk inherent in each scenario. 7.2.2
Customer Requirements Analysis
Crucial to successful market analysis is the identification of customers most likely to purchase service packages. Naturally there are many types of customers, but the most important is the one offering the greatest revenue potential in the shortest time frame. As an example, a given PCS operator may elect to target households for second line PCS service. The companys research may show that residents with $40,000 or more per year in income and having one or more children would be more likely to subscribe to that service. With this information, the operator can analyze the demographic variables on a geographic basis and visualize the results and quickly identify those license areas with people most likely to buy the service. A wide range of corporate and consumer applications are enabled not only by voice services, but by nonvoice wireless services, such as the following: • Internet chat groups have proven a very popular application of the
Internet, enabling groups of like-minded peopleso-called communities of interestto use nonvoice mobile services as a means to communicate.
• Textural and visual information with a wide range of content, from
share prices to sports scores, to weather and flight information, to news headlines, to traffic information and location-sensitive services, can be delivered to mobile phone users. This information need not necessarily be textualit can consist of maps, graphs, or other types of visual information.
• Still images such as photographs, pictures, postcards, greeting cards
and presentations, and static Web pages can be sent and received over the wireless network as they are across fixed telephone networks.
• Moving images are very important, since over time, the nature and
form of mobile communication is getting less textual and more visual. Videoconferencing applications, in which teams of distributed sales people can have a regular sales meeting without having to go to a particular physical location, is another application for moving images.
• Web browsing has never been an enduring application for mobile
users. Because of the slow speed of circuit-switched data, it takes a
Transmission-Network Planning and Design
263
long time for data to arrive from the Internet server to the browser. Alternatively, users switch off the images and just access the text on the Web, and end up with difficult-to-read text layouts on small screens. • Document sharing and collaborative working let different people in
different places work on the same document at the same time. Multimedia applications combining voice, text, pictures, and images can even be envisaged. These kinds of applications could be useful in any problem-solving exercise, such as firefighting, combat (to plan the route of attack), medical treatment, advertising-copy setting, architecture, journalism, and so on. Anywhere somebody can benefit from being able to comment on a visual depiction of a situation or matter, such collaborative working can be useful.
• Corporate e-mail is another very important nonvoice application.
With up to half of employees typically away from their desks at any one time, it is important for them to keep in touch with the office by extending the use of corporate e-mail systems beyond an employees office PC.
• Internet e-mail services come in the form of a gateway service where
the messages are not stored, or mailbox services in which messages are stored. In the case of gateway services, the wireless e-mail platform simply translates the message from SMTP, the Internet e-mail protocol, into SMS and sends it to the SMS center. In the case of mailbox e-mail services, the e-mails are actually stored and the user gets a notification on his or her mobile phone and can then retrieve the full e-mail by dialing in to collect it, forward it, and so on. Upon receiving a new e-mail, most Internet e-mail users do not currently get notified on their mobile phone. When they are out of the office, they have to dial in speculatively and periodically to check their mailbox contents. By linking Internet e-mail with an alert mechanism, however, users can be notified when a new e-mail is received.
• Vehicle positioning applications integrate satellite positioning systems
that tell people where they are with nonvoice mobile services that let people tell others where they are. The GPS is a free-to-use global network of 24 satellites run by the U.S. Department of Defense. Anyone with a GPS receiver can receive his or her satellite position and thereby find out where he or she is. Vehicle positioning applications can be used to deliver several services including remote vehicle diagnostics, ad hoc stolen-vehicle tracking, and new rental car fleet
264
Transmission Systems Design Handbook for Wireless Networks
tariffs. SMS is ideal for sending GPS position information such as longitude, latitude, bearing, and altitude. GPS coordinates are typically about 60 characters in length. • Wireless LAN access comes into play when mobile workers are away
from their desks and they clearly need to connect to the LAN in their office. Wireless LAN applications enable access to any applications that an employee would use when sitting at his or her desk, such as access to the intranet and corporate e-mail services. The mobile terminal, such as the handheld or laptop computer, has the same software programs on it as the desktop, cut-down client versions of the applications accessible through the corporate LAN. This application area is therefore likely to be a conglomeration of remote access to several different information typese-mail, intranet, and databases accessible through Web browsing tools or that require proprietary software applications on the mobile device.
• File transfer applications encompass any form of downloading size-
able data across the mobile network. This data could be a presentation document for a traveling salesperson, an appliance manual for a service engineer, or a software application such as Adobe Acrobat Reader to read documents. The source of this information could be one of the Internet communication methods, such as File Transfer Protocol (FTP), telnet, http, or Java, or a proprietary database or legacy platform. Irrespective of source and type of file being transferred, this kind of application tends to be bandwidth intensive. It therefore requires a high-speed mobile data service to run satisfactorily across a mobile network.
• Home automation applications combine remote security with remote
control. Basically, users can monitor their home from wherever they areon the road, on holiday, or at the office. If the burglar alarm goes off, the user is not only alerted, but can go live and see who the perpetrators are, and perhaps even lock them in. Not only can users see things at home, they can do things as well. It is possible to program the video recorder, switch the oven on so that the preheating is complete by the time the user arrives home (traffic jams permitting), and so on. As the IP will soon be everywherenot just in mobile phones but in all manner of household appliances and in every machineremote devices can be addressed and instructed. A key enabler for home automation applications will be Bluetooth, which allows disparate devices to interwork.
Transmission-Network Planning and Design 7.2.3
265
Interconnection
In current usage, most wireless calls are terminated to wire-line networks, typically the local LEC, on the PSTN. However, with recent telecommunications legislation, wireless carriers can opt for competitive access providers (CAPs) that also provide local transmission in North America. In establishing connectivity to an LEC or CAP, a wireless operator must understand the ability of that wire-line carrier to facilitate interconnection. For example, most CAPs concentrate their networks in only the largest cities. By mapping incumbent LEC facilities and the facilities of available CAPs, planners can choose who handles calls terminating to wired telephone networks. Ultimately, this can lead to cost savings as LECs and CAPs begin to compete for business and interconnection rates drop. Here, it is important to identify markets with high concentrations of customers most likely to purchase a specific service package. Service options can be customized and marketing dollars precisely targeted in those areas. The Erlang B distribution is used for dimensioning trunk routes. It is also used in wireless networks, and it is based on the following assumptions: • There are an infinite number of sources. • Calls arrive at random. • Calls are served in order of arrival. • Blocked calls are lost. • Holding times are exponentially distributed.
To calculate the capacity (number of T1/E1 lines) required for the interconnect (PSTN lines), it is important to calculate the BSC-MSC traffic and add a few percent for additional services like prepaid services, voice mail, and so on. The BSC-MSC traffic depends entirely on the total number of Erlangs for the market and can be calculated in two ways. One of the methods is to use Erlang B tables, which give a slightly more precise result. In case we want to use Excel spreadsheet, the best way is to use the following formula:
(
N = E +
)
E /M
where N = Number of T1 or E1 lines; E = Total number of Erlangs (Erlangs/sub × number of subs);
266
Transmission Systems Design Handbook for Wireless Networks
M = number of DS0s (24 for T1, 30 for E1). It is obvious that this formula does not take GoS into account. The results of the calculations for a very low number of Erlangs (a few hundred) follow closely 1% GoS, while higher numbers of Erlangs (a few thousand) follow closely 0.1% GoS. In any case, the maximum error that is made by using this formula instead of Erlang B tables and the given GoS is less than 10%, and in most cases close to 5%. For example, the number of required E1 circuits for the offered traffic of 500 Erlangs based on the tables and 1% GoS is 18 E1s, and based on 0.1% GoS it is 19 E1s. At the same time, the formula gives 18 E1s as a result. For the offered traffic of 5,000 Erlangs and 1% GoS, the required number of E1s is 165 and 172 E1s for the 0.1 GoS. The formula gives 170 E1s as a result. These results are somewhat overdimensioned, since not all the traffic will terminate on PSTN. In mobile networks, about 10% or less of traffic is between subscribers within the same network, and those calls do not get terminated on the PSTN. On the other hand, the call model is used for PSTN trunk design, BSC, and MSC capacity calculations. The call model is in compliance with the traffic formula: Erlang/Sub = ( AHT (s ) × BHCA /Sub)/3600 where AHT = Average call-holding time (sec). The call volume includes complete and incomplete calls. BHCA = Busy-hour call attempt. Again, this includes all call types. The customer usually defines AHT, but typically it is between 90 and 180 seconds. In the interconnection to the public network (PSTN), we design the network to connect to two or more tandem offices for local, national, and international traffic. In the design, the minimum trunk and signaling link requirement is two for reliability and redundancy reasons. 7.2.4
Spectrum Auctions
While the FCC (http://www.fcc.gov) has completed several sets of spectrum auctions, more spectrum auctions are scheduled for the months and years ahead. Congress has realized that auctioning radio spectrum is lucrative, and will continue pushing the FCC to make more spectrum available to
Transmission-Network Planning and Design
267
companies. Participation in any spectrum auction should be based on ration and the ability to generate substantial financial returns within a set period of time. Given forecasted penetration rates and estimated per-subscriber revenues, planners can identify the cost per POP and can use this information as a gauge in further bidding. In some countries (Mexico for example) there have been a number of auctions for the microwave spectrum as well (18, 23, 38 GHz). New wireless operators can lease spectrum directly from the government or in this case from the new spectrum owners and use it for the microwave systems connecting RBSs with the BSC (access transmission network or backhaul). 7.2.5
Clearing Spectrum and Microwave Relocation
One of the greatest obstacles faced by new PCS licensees is an incumbent microwave operator in the auctioned spectrum (2 GHz). These users are typically municipalities, pipeline and utility companies, or private firms. In order to carry wireless voice and data in the same spectrum, PCS licensees must provide compensation or comparable facilities to incumbent microwave operators. The FCC has undertaken a series of rulemakings that reallocated 220 MHz of fixed frequency bands near 2 GHz from fixed microwave to various PCS applications. The existing 2-GHz microwave links are permitted to move into several migration bands higher in frequency (4 GHz, 6 GHz, 10.5 GHz, and 11 GHz). The process will also provide narrower channelization in the destination bands to more efficiently handle the low-capacity links commonly in use at 2 GHz. The initial step in this process is identifying who the incumbent microwave users are and where they are located. This is accomplished by displaying the routes that microwave signals traverse in the PCS carriers license area and who the transmitter is. Overlaying this with the BTA or MTA boundary and cell site location can reveal impacts and help in decision making. If the microwave radio signals are in remote areas, the PCS carrier may be able to begin operation while negotiations continue with the incumbent microwave user. Otherwise, adjustments to network build-out will be needed, and negotiations with the incumbent will take a higher priority. Different countries have different approaches in resolving this issue. Of course, getting a license areas spectrum ready for use in PCS means more than just relocating microwave usersfor example, how much noise exists and will remain to cause problems for the future PCS network and how low this floor can be brought to ensure high quality of the wireless network.
268
7.2.6
Transmission Systems Design Handbook for Wireless Networks
RF Design
RF design is a complex process that includes analysis of physical topography and morphology, propagation analysis, and different equipment scenarios. Before and during network build-out, it is important to model how RF propagation will occur to and from subscribers using their handsets. In doing this, many geographic characteristics have to be considered, including the following: • Building heights and their locations; • Topography; • Population densities; • Traffic volumes; • Zoning laws; • Green areas.
The end result is that cell sites are selected based on propagation models. These results are provided to site acquisition personnel who can proceed to negotiate for use of, and access to, each site. Once cell sites have been acquired and constructed, base station equipment can be tested to confirm results and to improve RF propagation (drive testing and optimization measurements). This information can be analyzed and compared to original plans for further network refining. An important aspect of this activity is that it can also be conducted on a competitors network. This can reveal coverage lapses and weaknesses in signal strength that could play to a competing wireless operators advantage in securing subscribers. RF design for a new wireless network starts with the definition and understanding of the customer requirements. The initial design has to satisfy the signal coverage of the specified square mileage; the design to satisfy capacity requirements will follow in the next phase of the network build-out. Erlangs represent the average activity during a period of time. An Erlang requirement on a system node represents, then, the ability of that node to allocate enough resources to serve as many customers as the requirement and still meet its performance goals. Performance goals are expressed in the form of blocking probability or delay. What is referred to by resource depends on the network type and call type. Required GoS is also one of the factors determining how many base stations, sectors, FDMA frequency channels, TDMA timeslots, or CDMA code channels will be deployed in a given situation. It is common for the wireless systems to require GoS, usually
Transmission-Network Planning and Design
269
specified as a blocking probability using Erlang B formula, of 2% in North America and between 1% and 2% in Europe. A circuit-switched network establishes an end-to-end circuit involving various network facilities that are held for the duration of the call. From a network point of view, it is the holding of these resources that is important when specifying traffic capacity, not the flow of information within individual circuits. On the other hand, packet-switched networks are directly concerned with the actual flow of information, since in these systems, traffic on a transmission link is directly related to the activity of the sources. The trafficcarrying capacity of a packet-switched network is a function of the packetgeneration characteristics of the sources and transmission link capacity. In 3G networks things are more complicated since the traffic is a mixture of voice and different data services. Network resources used by a packetswitched call are not necessarily held for the duration of the call. Wireless technology is realizing a smooth evolution from circuit-switched technology to packet-based switching. This evolution can proceed over a period of time with ATM cell-based switching. ATM is an established and well-proven technology that is used by network operators to typically carry circuitswitched voice. The QoS mechanisms in ATM allow it to roll out multiservice (voice and data) networks with extremely tight delay requirements in combination with low-bandwidth radio links. In Universal Mobile Telecommunication System (UMTS), ATM is the prescribed transport mechanism in the radio access network. Here base stations are connected with many E1/T1s or higher-bit-rate connections. The small and delay-sensitive voice packets have to coexist with bursts of large IP data packets. Today, classic IP grabs all of the bandwidth, but ATM provides the required QoS for the delay-sensitive voice packets, even during handoffs from one network to another network. In the future, 3G standardization will most likely move toward the 3G real-time all-IP networks, where all services, transport and access alike, use the IP protocol stack. Initially IP will be used for simple data communication services, such as Web browsing, e-mail, and file transfer, later for streaming data, and finally for conversational services with real-time requirements. Examples of real-time services are voice, video, and multimedia IPall will be IP end-to-end. All-IP requires developments in the transport network and in managing the end-to-end QoS. Regardless of whether it is a 2G or 3G wireless network, the result of the RF plan is a number of cell sites, their location, and the type of the cellsite infrastructure required for the project. Calculations are based mainly on the given number of Erlangs per subscriber (or total number of Erlangs for the given market) and GoS.
270
7.2.7
Transmission Systems Design Handbook for Wireless Networks
Transmission Media and Topology Planning
Transmission planning is a subset of the total wireless-network planning, both when creating new networks and extending existing ones. Modern networks allow a very great flexibility in routing and intelligent switching features. It is important, however, that transmission aspects are not forgotten in the planning process. For complicated networks one should keep in mind the abilities of the particular signaling that is implemented. Advanced signaling systems could, in addition to performing their normal functions, convey information about certain transmission parameters in connections. Examples of transmission parameters of interest are accumulated delay, existence of echo cancellers in the path, terminals not needing network echo control, accumulated impairments, choice of particular routes for calls with special requirements for high-quality connections, and so on. The goal is to build the network that will provide a reliable transmission network capable of delivering enough capacity for present needs as well as ensuring seamless expansion in the future. The cost of transmission for base station access and inter-MSC backbone networks is not insignificant, typically 20% to 25% of infrastructure costs. Its impact on operator profitability is also influenced by the way the network rollout speed and transmission-related network downtime affect the customers overall perception of service quality. Usually, transmission (transport) facilities are either leased or owned (copper, microwave, fiber optic), or most likely a combination of both. In many cases, project managers and those involved in the time-to-market assessment conclude that a faster way to build the network is to lease T1/E1 circuits from the local telephone companies rather than build their own microwave system. That may not be the case in every situation and may prove, for a number of reasons, to be a much more expensive and lengthy process than originally anticipated. Transmission-system requirements vary throughout the network life cycle and the most successful new and expanding operators offering mobile services take future transmission requirements into account in the initial service launch phase. A successful transmission platform includes necessary network functionality for the entire network life cycle, supporting network growth (scalability) and reconfiguration (flexibility) as well as fast multiservice deployment. The transmission network can be presented either with a logical or physical view. The logical view is useful for initial network dimensioning. It shows where a connection is needed in order to transport traffic from source to destination. It is, however, important also to look at the physical implementation in
Transmission-Network Planning and Design
271
order to build an optimized network. Transmission architecture in 3G wireless networks will be slightly different than in 2G wireless networks. The main reason for this is required flexibility for expansion and increased capacity that will be mandated by new and bandwidth-consuming applications and also a number of new nodes that will be introduced in 3G networks. Even before the RF network design has been completed, it is possible to estimate roughly if the available microwave band will allow a simple microwave network or whether microwave repeaters or other microwave bands will be required. For example, in the 800-MHz cellular band, CDMA cell sites have a typical radius of 2 km for dense urban morphology, 3.5 km for urban, 7.5 km for suburban, 20 km for rural, and 25 km for highway cell sites. That will mean that in most parts of the world (except in areas with an extreme amount of rain) we could use 15- or 18-GHz bands to connect dense urban, urban, and suburban cell sites. For rural and highway sites, lower bands will be required, and most likely in the 6- or 7-GHz band or 5.8-GHz spreadspectrum (ISM band), microwave radios will be used. 7.2.8
Mobile Positioning
Location-related products are the next major class of value-added services that mobile network operators can offer their customers. Not only will operators be able to offer entirely new services to customers, but they will also be able to offer improvements on current services such as location-based prepaid or information services. The FCC adopted a ruling in June 1996 that requires all mobile network operators to provide location information on all calls to 911, the emergency services. The FCC mandated that by October 1, 2001, all wireless 911 calls must be pinpointed within 125m, 67% of the time. On December 24, 1998, the FCC amended its ruling to allow terminal-based solutions as well as network-based. There are a number of regulations that location-based services must comply with, not least of all to protect the privacy of the user. In 2000, Ericsson, Nokia, and Motorola joined forces to create a common standard for positioning for wireless networks and creating the Location Interoperability Forum (LIF). Members of this forum will be a mix of operators, suppliers, and application developers. By providing open interfaces as standardized by LIF, applications will be able to operate identically in systems from different suppliers. Different location-based applications require different levels of accuracy and clearly it is important for ships to know exactly how far they are from shore and the water depth, whereas for people location accuracy of a
272
Transmission Systems Design Handbook for Wireless Networks
hundred meters or so would often be acceptable. Driving directions, roadside assistance and tracking of fleet, packages, and people will each become more prevalent with the availability of improved location accuracy. While some early adopters in vertical markets will be satisfied to know in what part of town their trucks or packages are located, most customers in the horizontal market will want to know exactly where their assets, or even their children, are. Similar dynamics surround roadside assistance. Although roadside assistance providers can gain efficiency from knowing in what part of town motorists are stranded, most subscribers believe they need to be located within 125m to feel safe. The provision of turn-by-turn driving directions, the most complicated location service, requires even higher accuracy and frequent updates. The key to incorporating accuracy into a network-based location service is the ability to combine improved location finding technology with less accurate methods, such as cell of origin data, so that network operators can achieve scale quickly. In addition, they can market the services broadly to existing customers instead of targeting only subscribers with enhanced phones. Customer care, billing systems, and switches can be integrated with the location applications without waiting for networkwide coverage or complete handset penetration. Finally, customers can immediately use their new location services in 100% of their coverage area and receive service 100% of the time, regardless of the inherent inconsistencies in location accuracy and availability. More information about different methods of mobile positioning is provided in other chapters of this book. 7.2.9
Toll QoS
QoS is defined as the collective effect of service performances that determine the satisfaction of a user of a service. These quality of service performance factors are applicable to all services, such as service support performance, service operability performance, service accessibility performance, service retainability performance, service integrity performance, and other factors specific to a given service. Network performance is defined as the ability of a network or network portion to provide the functions related to communications between users; it contributes to service accessibility, retainability, and integrity. Network performance parameter values are usually derived from quality of service parameter values [1]. Connections including the wireless system should, under error-free conditions, achieve subjective ratings comparable to those of connections in
Transmission-Network Planning and Design
273
the ordinary PSTN. This is a necessary, but in no way sufficient, condition for being considered PSTN quality or toll quality. To realize the widespread acceptance, from a quality perspective, of the PSTN, many other requirements should be satisfied by wireless systems. For example, the received speech should sound natural, and users should be able to recognize the voices of people with whom they are familiar. In addition, wireless systems should be robust to transcodings (such as when used in tandem with a far-end wireless system) and robust to a reasonable level of bit and frame error. It also has to be robust to the wide variety of ambient noise conditions (e.g., offices, outdoors, highways) under which such systems will be used. Also, highly interactive conversations should be possible with minor effort, meaning that excessive delays cannot be introduced and no annoying effects should be imposed on call progress tones, network announcements, or music-on-hold. Severe channel impairments, such as signal dropouts shall not be frequent or regular, and speech processed through such systems shall be recognizable by network-based speech recognition systems (that already work well with PSTN-originated speech). Failing to support these capabilities, or providing performance levels that prove unsatisfactory to the user, may cause wireless systems not to have the widespread user acceptance that is the clear goal for these systems. 7.2.10 Network Performance
In todays telecommunications networks, wireline, wireless, and Internet communications converge. To be able to provision, test, and monitor the network (and of course bill the customer), it is necessary to have remote access to every piece of equipment in that network. The key to providing service quality lies in the successful integration of network technologies that provide both voice and data. Networks are becoming more intelligent, geographically dispersed, and larger every day. We continue to deploy thousands of intelligent software NEs throughout wireless networks, from the backbone to the periphery of the WAN and the last mobile unit in the network. These NEs will intelligently administer tasks across the network, such as accounting functions, user authentication, inventory control, and configuration and monitoring of remote devices, while also generating new revenues and reducing network operation and management costs. The integrated NMS should be vendor-independent and able to do the following: • Provide an impartial view of the service quality provided by each
vendors equipment;
274
Transmission Systems Design Handbook for Wireless Networks
• Extract service quality information from all of the protocols and
technologies in use in the network; • Provide a view of end-to-end delivered-service quality.
In todays complex multiservice environment, the customer care and billing system must be flexible to handle the multitude of services that may be billed on different usage schemes. Whereas traditional telephony calls are billed on a duration basis, this charging model is, in many cases, not appropriate to the new band of services, which may require different models such as usage or volume-based. Additionally, services may need to be bundled to remain competitive and so the billing system will need to support the complex rulesbased rating models needed to bill for these. The time-to-market for new services is essential, and therefore the customer care and billing system must be flexible to support this. It is important to remember that superior customer service is the key factor in retaining subscribers. While this has become a cliché in many industries, including telecommunications, it is of critical importance now more than ever to wireless carriers. Fast resolution of customer complaints and the ability to quickly correlate multiple troubles is part of this customer service initiative. When a customer calls in, the customer service representative (CSR) must be able to record the trouble, the time when it occurred, and the location of the trouble. The CSR can also assess the problem based on specific categories: 1. Service quality Coverage Dropped calls Blocked calls 2. Voice quality 3. Special features SMS Voice mail Call waiting Call forwarding 4. Equipment problems No connection Handset problems 5. Fraud
Transmission-Network Planning and Design
275
Subsequently, these problems can be mapped and correlated to problems other customers are encountering. This information provides engineering with the data it needs to spot recurring problems and to resolve them. Voice quality is usually based on subjective tests performed in the lab environment as well as the opinion of actual users of the system. The average of the scores obtained from the number of listeners is called the mean opinion score (MOS). The toll-quality voice usually has MOS > 4 on a scale of 1 to 5, with 5 being the best (speech is perfectly understandable) [2]. It is important to note that all of the problems listed above could be caused by over-the-air interface (RF), subscriber units, network elements (switching), or transmission facilities (E1/T1 circuits). When engineers start troubleshooting the network problems, all the components of the network architecture are the suspects. The network performance improvement and optimization process examines dropped calls, interference, handoff performance, coverage verification, customer churn, and other competitive factors. It focuses on data collected from the network in live tests. Drive tests using wireless measurement equipment, actual phone quality messages, switch data collection, and network performance-reporting software are important elements of the optimization process. Using this data, it is possible to troubleshoot problems and correct them accordingly. Network management involves assuring quality service for subscribers wireless voice and data communications. However, no network is completely immune to problems, natural or man-made. While cell sites can frequently impact call quality, there may also be problems in handling traffic from cell sites to the switching center during peak periods of the day. A group of cell sites may need to be upgraded from a T1 line to a T3 line to handle the volume of calls. To support network operation and maintenance and fast problem resolution, many wireless operators are creating site information databases. These databases detail information on each cell site including location, topography, RF propagation, equipment inventory, and physical structure. In turn, this information can be made accessible to field personnel who have responsibility for maintaining and inspecting each cell site. It is a tool that helps in identifying and deploying network components to begin operations and to sustain them with quality call completion. An increasing problem in providing wireless service is fraud. It accounts for close to $1 billion in cellular costs per year, and this is expected to grow. While providing quality customer service, a wireless operator must also pay attention to those end-users who are trying to exploit a wireless operators network illegally. The first step in combating fraud is identifying
276
Transmission Systems Design Handbook for Wireless Networks
those parts of an operators coverage area where much of it happens. Measures can then be taken to reduce the problem; preventative efforts such as requiring personal identification numbers (PINs) or security codes can be used to intercept calls and to validate subscribers. Sales to these areas can be restricted and certain types of calls (e.g., international) can be blocked. On a more active basis, investigators can be assigned to identify and prosecute perpetrators. Also, users are most vulnerable when they leave their home areas. Users may roam using dual mode analog-digital handsets, and may travel to areas where the provider has not yet implemented authentication. Authentication is a method by which analog and digital networks exchange algorithms between the switch and a handset to insure secure transmission before sending information. For network operators, fraud represents a very real problem. An effective fraud strategy can save millions of dollars in lost revenue and protect your relationships with your key customers. Faced with increasingly sophisticated fraudsters and hackers, operators must take additional steps to ensure the security, reliability and profitability of their services across all their networks: fixed, mobile, and intelligent. 7.2.11 Sales and Marketing
The key ingredient to near-term success for any wireless operator is the level of sales and marketing expertise that it concentrates on attracting subscribers. As a carrier builds out its network, it must match its network and capabilities to end-user needs and the likelihood of end users to purchase. Customer profiles must be built and analyzed, competitive offerings must be understood, and sales materials must be generated. As an example, those subscribers having high long-distance usage may tend to be located in certain license areas. These end users could be offered the operators branded long-distance for wire-line and wireless calls that, when bundled, provides cheaper overall monthly bills. Identifying this would assist in filing the appropriate tariffs and forms and arranging for connectivity with a wholesale long-distance carrier. Lastly, information on calling patterns can be used to create tailored rate plans. For instance, if a firm currently resells cellular service prior to operating its own PCS network, it can draw upon billing records to identify cell sites where its cellular customers are making calls. The correlation of where cellular calls are placed and where those end users live and work can result in PCS rate plans that are zoned and more attractive than cellular plans. Also, by obtaining information on competitors license areas and the status of their network,
Transmission-Network Planning and Design
277
a wireless operator can match those competitors or decide to focus on underdeveloped or neglected areas. In some cases, multiple players may fight aggressively in the same big city for large revenues while ignoring smaller cities with higher profit margins. An operator can determine this through competitive analysis and refocus its energies on the smaller cities. Over the next few years, many of the existing operators will migrate from the 2G toward 3G wireless networks. The following issues concerning that process stand out: • How to increase average revenue per user; • How to minimize costs of technology deployment and to adopt as
simple a deployment process as possible;
• How to deploy viable commercial services in a timely manner; • How to maintain a satisfactory end-user experience throughout this
process.
The latter two items on this list are especially important. If operators cannot deploy viable commercial services in a timely manner, they run the risk of losing competitive position. If the network or handsets deliver a negative end-user experience, they will not only fail to generate revenue, but will encourage churn as well. Not to be forgotten is the importance of voice. Using compression techniques and more sophisticated bandwidth management, 3G will eventually enable nonvoice applications such as full-motion video and multimedia, all in real time. This is the beginning of the mobile Internet era with possibilities to increase revenues from existing subscriber bases as well as from new market segments, segments that have thus far not been economically or technically possible to address. An indication of the business potential is presented by predictions that there will be approximately 600 million mobile Internet users by the end of 2004, with 3G leading us into a phase of mobile Internet hypergrowth. It is predicted that within a few years, more end users will access the Internet via handheld mobile terminals than wireline connections. Subscribers today are more sophisticated when choosing their wireless service provider. Increased subscriber dependency on the Internet and information services is creating new demands on wireless systems. In the near future, IP telephony has the potential to become the dominant technology for the provisioning of real-time wireless telephony services. The migration of all elements in the present circuit-switched domain (basically voice and video
278
Transmission Systems Design Handbook for Wireless Networks
applications) toward the IP-based packet-switched domain is expected in the medium to long term. The circuit-switched domain will, however, continue to grow for some years mainly due to the fact that this technology is currently the only one that will ensure the service performance and reliability required by the operators. Although wireless IP telephony is more flexible in its ability to provide a variety of innovative services, the strong consensus today indicates it may be several years beyond 2002 before it can compete with CS services in terms of radio spectrum efficiency, error robustness, and voice quality. Smooth migration toward an all-IP network means preparing for this evolution by basing the launch system on a future proof backbone and at the same time maintaining the support for circuit-switched domain traffic. The focus of a subscribers evaluation criteria when choosing a wireless service provider is based on value-added services and features. A sophisticated blend of information, voice, data, education, entertainment, multimedia, anywhere, anytime, will provide wireless operators with opportunities in offering differentiated services for targeted customer segments.
7.2.12 Regulatory Issues
The FCC is an independent federal regulatory agency that is responsible directly to Congress. Established by the Communications Act of 1934, it is charged with regulating interstate and international communications by radio, television, wire, satellite, and cable, and its jurisdiction covers the 50 states, the District of Columbia, and U.S. possessions. The general objectives of federal telecommunications regulations are to provide efficient use of the electromagnetic spectrum, which is considered a public resource and to develop a domestic telecommunications infrastructure able to provide service on the national level and compete on a global level. Especially in recent years, its task is also to provide a highly competitive economic market that spurs technological advances in the telecommunications industry. The FCCs Wireless Telecommunications Bureau (WTB) handles all FCC domestic wireless telecommunications programs and policies, except those involving satellite communications. Wireless communications services include cellular telephone, paging, personal communications services, public safety, and other commercial and private radio services. The WTB regulates wireless telecommunications providers and licenses. The bureau also serves as the commissions principal policy and administrative resource with regard to federal auctions for the private use of the public airwaves.
Transmission-Network Planning and Design
279
The WTB regulates the three primary wireless communications services: cellular, ESMR (also known as iDEN), and PCS. Portions of the frequency spectrum are allocated to specific uses (such as TV broadcast band, cellular), and specific frequencies within that part of the spectrum are assigned to licensed operators. These procedures are intended to prevent interference or conflicts among various operators or services at a given location attempting to use the same portion of the frequency spectrum. One of the functions of the FCC is to issue licenses to wireless communications carriers. The FCC issues licenses for certain frequency bands of the electromagnetic spectrum and effectively limits the number of wireless communications providers in a specific geographic service area. In addition to regulating licenses, the FCC establishes performance standards for cellular and PCS providers. The FCC requires cellular and PCS licensees to provide, within a specified period of time, a coverage ratio of a minimum quality for either a composite geographic service area or a percentage of an areas population. The FCC currently requires cellular, ESMR, and PCS providers to comply with the ANSI/IEEE standards for radio frequency electromagnetic fields as a condition of licensure. Under authority granted by the Federal Aviation Act, the FAA has jurisdiction over the following communication facilities: (1) towers that exceed 200 ft in height; (2) towers that are located within 20,000 ft of a major commercial or military airport; and (3) towers that are located within 10,000 ft of a general aviation airport. The FAA reviews the location and height of such towers and may require them to be painted or illuminated to prevent possible interference with nearby airport operations. The FAA also reviews possible interference issues with aircraft-to-ground communications that may be caused by transmission facilities located in or near airport flight paths. Under the requirements of the FAA, wireless communications providers are responsible for filing a notice with the FAA if their facilities are subject to FAA review. Most local governmental agencies regulate wireless communications facilities via land-use regulations contained in respective zoning ordinances and general plans and are responsible for reviewing and processing applications for discretionary and ministerial permits for these facilities. Local governments also have the broad authority to ensure the public health, safety, and welfare of their citizens. Local jurisdictions regulate wireless communications facilities through the permitting process. Most agencies require a discretionary permit, such as a conditional use permit, in order to construct a facility. Whether a permit is processed administratively or requires a public hearing varies among local
280
Transmission Systems Design Handbook for Wireless Networks
agencies. In general, administrative processing entails lower permit fees and shorter processing times, while the public hearing process involves higher permit costs and a longer permit turnaround time. Frequency spectrum is a valuable resource and is generally subject to appropriate planning and management to prevent misuse and interference between the many and varied applications. National administrations will allocate some or all of these bands for fixed microwave radio use in line with local requirements. Before microwave network planning commences, the wireless operator must determine available frequency bands and channel plans specific to that country. Often, and preferably, an operator may be able to obtain a number of frequency allocations as a block, enabling him to perform his own network planning in advance without risk of interference from other users. Most regulatory authorities also operate a local link length policy, where the length of a particular path will determine what frequency bands are available for the operator to choose from. Typically, the shorter the path the higher the frequency required. Frequency planning is the coordination of link frequencies to minimize any interference between links within the network and those operated by other users. In some instances, the local regulatory authority undertakes frequency planning. If, however, a block allocation has been obtained, then planning will be the responsibility of the wireless operator. The local requirement for equipment type approval will also vary from country to country, ranging from a simple paperwork exercise to a full product test program to local standards. Type approval is generally the responsibility of the radio supplier, but an operator should ensure that all requirements are satisfied before microwave links are deployed. Other limitations imposed by authorities can also have an impact on microwave radio deploymentfor example, tower height restrictions or limitations on antenna size. These factors can restrict effective radio lengths at the planning stage and should be ascertained in advance of the detailed link design stage. In every wireless and wire-line network there must be a provision for the lawful intercept (LI). The LI is used to intercept unobtrusively telephone calls to or from designated subscribers and send the information, composed of the activity on the connected line as well as call-related information (CRI), to Monitoring Center (MC). The LI is equipped to handle calls that require full monitoring, where the entire conversation is recorded and can be monitored by the appropriate law enforcement agency, and calls that only require statistical monitoring, where only CRI is kept in a database. CRI consists of the following information:
Transmission-Network Planning and Design
281
• Called and calling parties phone numbers; • Priority of the call (high or low priority); • Level of monitoring (full or statistical); • Call start time; • Call end time; • Call duration.
The LI receives a call to be intercepted from the switch on one trunk, makes an outgoing call to the switch on another trunk, and connects both legs of the call to carry the conversation. Additionally, in the case of fully monitored calls, the LI establishes another call to the MC where the call can be recorded. In the case of statistically monitored calls, the LI sends to the MC information such as time call started, time call ended, called and calling phone numbers involved, and other pertinent information to be stored in the data files. 7.2.13 Life Cycle of Wireless Networks
The life cycle of successful wireless networks involves development through steps: initial service launch, high growth phase, network changes and optimization, and new revenue opportunities involving the launch of new services. This stepwise development implies changes in focus for network optimization. Essential to competitive development through transmission technologies, these changes in focus drive changes in network planning and configuration throughout the network life cycle. Increasingly, new operators strategically create their own unique portfolio of value-added services designed to attract a targeted subscriber potential, specializing either in cost minimization or revenue maximization. The operators choice of strategy naturally varies from one country to another, depending on local or regional regulatory terms and charges. Successful mobile network operators launch their services by providing adequate geographical coverage in key areas of the network, including major cities main roads and selected special tourist or business locations. This coverage tends to be implemented through a network consisting of a few macrocellular base stations and an MSC. Fast service rollout for a new network is of vital importance to enable cash inflow and investment payback. The transmission network in such a launch phase is not usually optimized for link capacities or cost, but the transmission system and its management system
282
Transmission Systems Design Handbook for Wireless Networks
are advisable to be selected for their ability to meet potential requirements of future network growth. During later phases, many new base stations will share the main transmission links of the initial launch network. As the subscriber base increases, successful operators add new base stations, BSCs, and switches to serve all cities and improve roadside and indoor coverage. The number of macrocellular base stations can grow swiftly to hundreds, and ultimately to several thousands, and the number of switches could grow to between 15 and 30, resulting in a rapid increase in the number of transmission links. Managed network growth is essential in this phase, as installation and commissioning successfully require the minimum number and duration of site visits for a smooth and cost-effective transition. Cost optimization plays an increasing role during this phase, as grooming to carry traffic from several base stations over fewer T1/E1 links optimizes transmission links. Transmission hubs located in selected base station sites perform the grooming, and new operators who lease the transmission capacity also benefit from savings enabled by grooming. As the number of base stations grows larger, successfully expanding operators rearrange transmission links more frequently by adding capacity or rerouting the traffic for one or more reasons: • Macrocellular base stations are added; • Transmitter/receiver units (TRX) are added to existing base stations
to upgrade capacity;
• Microcellular cell systems are added; • New BSCs are installed, and base stations reconfigured to work
under new BSC serving areas;
• New switches are installed, and base stations reconfigured to work
under new switch serving areas;
• Automatic rerouting and repair capabilities are introduced; • New services, mobile and fixed, are introduced with minimum
rollout, with the same transmission infrastructure used for multiple services.
As the competition for a regions cellular market grows, service differentiation and introduction of new value-added services play an increasing role. Operators who use their own transmission networks tend to begin offering fixed services such LAN-interconnect, voice, IP and other services in any
Transmission-Network Planning and Design
283
phase. As a cost-effective strategy, the most successful operators may use the same transmission network for both mobile and fixed services.
7.3 Transmission System Design 7.3.1
Transmission System Design Process and Requirements
Process is, by definition, a linked set of activities, which defines an object and is directed toward achieving a specified and measurable result. Processes should be defined and described to improve performance within the organization and its projects. A process description includes descriptions of activities in the process as well as entry and exit criteria for these activities. Entry criteria are predefined criteria that should be fulfilled before a certain task can be started. Exit criteria are also predefined criteria that should be fulfilled for a certain task to be completed. The purpose of this chapter is to provide an overview of the transmission-network design methods, processes, and procedures used in the wireless networks and in transmission networks in general (Figure 7.2). It will also point out the need for the RF, switch, and transmission engineers to work closely together in defining, planning, and deploying the most reliable and cost-effective network possible. During the first stage of the network design process, a few initial questions need to be asked with regard to such areas as economics, the area topology, the existing network, and the services that the customer wishes to offer. The following are a few examples: • Is this an upgrade of the existing wireless network or a completely • • • • • • • •
new network (green field project)? What economic resources does the operator have? Are we planning 10 years ahead or just dealing with todays demand? Is the operator going to cover small areas (islands) or the entire country? Are we primarily aiming for the business or private subscriber? Is the main purpose of the new network for data traffic or speech traffic? Is CS traffic to be maintained on the existing network? Is there spare capacity on this existing network? What kinds of services is the operator going to offer?
284
Transmission Systems Design Handbook for Wireless Networks Site visit results
Customer/marketing input Other inputs
Draft RF network plan output
RF engineering
System engineering Transmission Switching
Customer input Draft Tx network plan output
Tx engineering
Switching/Network Architecture input
All Yes Final RF/Tx requirements network plan satisfied? No Deployment
Figure 7.2 Transmission-network design process.
Before the transmission-network planning can start, some basic activities have to take place in order to define the wireless networks requirements and expectations: 1. Review the proposed coverage area (or RF network plan, if completed) and identify the need for the transmission (backhaul)leased lines, fiber, or microwave facilities. 2. Determine the responsibilities for the backhaul (leased lines, fiber, MW) design and deployment. 3. Meet with customer(s), contractor(s), vendor(s), and partner(s). 4. Sign off the nondisclosure agreements with all parties (customer, vendors, partners) involved in the project. 5. Identify potential microwave sites and MW capacity requirements. 6. Identify need for the MW repeaters or SDH high-capacity MW backbone. 7. Identify MW frequency bands or channels to be used. 8. Identify availability of the unlicensed microwave spectrum. 9. Identify existing MW systems in the area and the source of information (microwave frequency coordination).
Transmission-Network Planning and Design
285
10. Provide information (drawings, maps) of the existing transmission facilities in the area (MW, fiber-optic, copper), as well as PSTN offices and POPs of the local telecommunication companies. 11. Determine availability numbers, reliability, and mean time to repair (MTTR) of the existing transmission facilities. 12. Identify backhaul requirements for the switch location. 13. Identify and visit potential switch site location(s). 14. Complete the site survey on some or all of the potential cell-site candidates and determine feasibility of the sites to be used as MW sites or hubs (determine LOS with the adjacent sites). 15. Determine existing tower and other antenna mounting structure capabilities, sufficient space for the MW radio equipment and antenna installation (provide site layouts and tower profiles) and access to those sites. 16. Determine power and battery back-up requirements. 17. Find out all the customer-specific requirements. 18. Organize and schedule detailed MW path or site survey if required. 19. Complete a scope and task delineation list (showing who is doing what). 20. Provide preliminary high-level transmission-network design (TND) and identify equipment required. 21. Identify equipment and service resources (for international projects, try to find local companies). 22. Develop preliminary transmission-network build-out schedule. The role of the transmission-network planner is no longer restricted to simply network and technology requirements. It is important for the network planner to be able to optimize existing and new network resources to create a high-quality and cost-efficient network. It is vitally important that the transmission-network planner understands the business objectives of the company and is able to respond to business requirements effectively. To plan a network effectively, a transmission-network planner needs to have a complete understanding of the whole network. At all stages, the emphasis should be on designing simple network architecture. This will be beneficial in terms of deployment and network management and will provide flexibility to allow for easy network expansion.
286
7.3.2
Transmission Systems Design Handbook for Wireless Networks
SDH and PDH Transmission Systems
In PDH, older T-carrier and E-carrier networks are plesiochronous; backhaul microwave systems for small to medium wireless networks are always PDH. One of the problems with the PDH is that it doesnt have standardized optical interfaces. SDH is a set of international standards for broadband communications over single-mode fiber-optic transmission systems. The standards were originally developed in the United States, where they are known as SONET. Basic signal levels of E1 (2 Mbps) and E3 (34 Mbps) can be multiplexed into an STM-1 or STM-4 signal. The SDH signal, being synchronous, allows direct access to all of the component signals without the need to demultiplex the complete signal; thus, only equipment sufficient to add or remove the required signal is needed at any location. SDH networks are usually configured to be highly redundant, with dual fibers providing backup. They can be laid out in a linear fashion; however the optimum network topology is a ring. In a ring configuration, the first fiber transmits in one direction, with the second fiber transmitting in the opposite direction. Through this layout, it is highly unlikely that any device on the network can be isolated through a catastrophic failure; this feature is known as self-healing. Digital microwave high-capacity networks today are designed with these same principles in mind. 7.3.3
SDH Transmission-Network Protection
A description of the generic protection types in an SDH network is provided in ITU-T Recommendation G.805. This ITU-T recommendation indicates how these generic types are applied in the case of the SDH. A detailed description of the implementation of some of these schemes is provided in ITU-T Recommendations G.783 and G.841. SDH multiplex section protection is a type of trail protection as described in ITU-T Recommendation G.805. Failure events are detected by the multiplex-section-termination (MST) function, and the reconfiguration uses the protection switching functions that are in the multiplex-sectionprotection sublayer. The resultant reconfiguration may involve protection switching in multiple SDH network elements. The coordination of such switching in multiple SDH network elements is by means of an automatic protection switching (APS) protocol [3]. In a 1+1 SDH multiplex-section-protection system, two multiplex sections are provided: One carries the traffic, and the other acts as a standby. A description of multiplex section 1+1 protection is given in ITU-T Recommendation G.783.
Transmission-Network Planning and Design
287
A 1:N SDH multiplex-section-protection system consists of N trafficcarrying multiplex sections that are to be protected, together with an additional multiplex section to provide protection. When not required for protection, this additional multiplex section capacity can be used to support lower-priority extra traffic. This extra traffic is not itself protected. A description of multiplex section 1:N protection together with the APS protocol is given in ITU-T Recommendation G.783. Multiplex-section shared-protection rings are characterized by dividing the total payload per multiplex section equally into working and protection capacity. That means that for a two-fiber STM-N ring, there are N/2 administrative unit groups (AUGs) available for working and N/2 AUGs for protection, while in a four-fiber STM-N ring, there are N AUGs available for working and N AUGs available for protection. Any multiplex section of a multinode ring under a section or node failure condition can access the ring protection capacity. Thus, the protection capacity is shared between multiple multiplex sections. This sharing of protection capacity may allow a multiplex-section shared-protection ring to carry more traffic under normal conditions than other ring types. Under nonfailure conditions, the protection capacity can be used to support lower-priority extra traffic. This extra traffic is not itself protected. A description of multiplex-section sharedprotection rings including the definition of the APS protocol, is provided in ITU-T Recommendation G.841. A multiplex-section dedicated-protection ring is a 1:N protection scheme where N = 1. A system consists of two counterrotating rings (each transmitting in opposite directions relative to the other). Under failure conditions, the entire working channel is looped to the protection channel. The APS protocol required for this scheme is not provided in ITU-T Recommendation G.841, since the maximum capacity of this type of ring is the sum of the capacity on each span, and therefore the applications for this type of protection scheme are limited. Subnetwork connection protection is described in ITU-T Recommendation G.805. It may be applied to either a SDH higher-order path or lower-order path. To support subnetwork protection, two dedicated subnetwork connections are provided: One carries the traffic, and the other acts as a standby. This protection mechanism can be used on any physical transport structure (e.g., meshed, rings, or mixed). It can be used to protect a complete end-to-end network connection or a portion of a network connection. Further details of the application of this scheme in the SDH are provided in ITU-T Recommendation G.841.
288
7.3.4
Transmission Systems Design Handbook for Wireless Networks
Ring Protection in the Wireless Network
For a larger transmission network, it is recommended that ring configuration be used as a high-capacity backbone (SONET or SDH) carrying traffic from the cell sites to the switch location. In this configuration BSC is always one node on this high-capacity ring. Ring configuration provides for a reliable communication facility, since it offers automatic protection from the following: • Site hardware (batteries, towers, antenna system) failures; • Radio and MUX equipment failures; • Propagation failures in the microwave network; • Cable cuts in the fiber-optic network.
It also provides basic user features such as simple operation, fault location, and maintenance, enables alternate routing of E1/T1 traffic automatically, and protects against loss of E1/T1 traffic due to single failure. The main rule is that each E1/T1 circuit must be dedicated completely around the ring. Reuse of the same E1/T1 in the opposite direction is not possible, and for ultimate reliability, both directions can be 1:1 hardware protected. All the sites that belong to the ring are considered hub sites and have to be completed first in order to provide protection. Figure 7.3 shows the principle of operation of the self-healing bidirectional ring configuration. In microwave systems, additional protection (e.g., space or frequency diversity) at lower frequencies may be required against short-term multipath outages. Added service reliability can be achieved even without the ring configuration, using automatic rerouting. Many successful mobile operators protect transmission by using automatic traffic rerouting on their most important links, assuring additional reliability in normal situations, such as when access microwave radio links suffer cutoff due to poor weather conditions or possible fiber-optic cable cuts or any other human error. With a flexible rerouting transmission system such as T1/E1 trunk rerouting, backup capacity can pass via physically separate routes, as the problem is not likely to interrupt both routes simultaneously. The base station trunk is the entire physical transmission link between two base stations or sites or between a base station and its base station controller, typically a T1/E1 or nxT1/E1 link. In case of traffic failure, trunk rerouting switches all traffic in the main trunk simultaneously to the backup trunk (Figure 7.4). Large base stations comprising a number of circuits are
Transmission-Network Planning and Design
E1/T1
E1/T1
RBS3
E1/T1 RBS2
Repeater
3
T1/E1 Protection-only direction
4
DS0 21 RBS4
1
RBS1
Repeater
Red alarm
RBS4
1
E1/T1
Rain outage
3 T1/E1 Wrap
RBS2 2
4
Red alarm
RBS3
E1/T1
2 T1/E1 Normal direction
289
RBS1 BSC
E1/T1
BSC
E1/T1
E1/T1
a.
b.
Figure 7.3 Ring protection in MW network: (a) normal (CW rotation) operation, and (b) failure between sites 3 and 4.
switched simultaneously for minimum service downtime. Rerouting can be arranged for all sites or only critical sites, such as base stations that are labeled as higher priority (e.g., hub sites). 7.3.5
Description of TND Deliverables
A deliverable is any output from the process of planning, design, testing, and so on. It could be a drawing, written document (report, white paper, spreadsheet), or just about anything two sides agree upon. This section describes briefly each deliverable during the transmission-network planning, design, and deployment stage and its contents. The preliminary network report is a description of the adopted transmission strategy. The topologies chosen for the different parts of the network are presented along with background for the choice. The report will also present the proposed media for the various parts of the network and the reasoning
290
Transmission Systems Design Handbook for Wireless Networks
RBS3
Traffic interruption RBS2
RBS1 BSC
RBS4
RBS5
RBS3
RBS2
RBS1 BSC
RBS4 Traffic rerouting RBS5
Figure 7.4 T1/E1 link (trunk) protection.
supporting the proposal. Furthermore, the estimated capacities in the network are presented. The report is written on an overview level. It does not account for each individual link. The preliminary network layout is a network drawing showing an overview of the access and backbone transmission networks. Links routing as well as estimated capacities and proposed media are indicated. If the nominal network planning is based on a channel-loading plan only, the layout will not show each individual link as the site positions are not known. In this case the layout will detail the main link routing such as the backbone transmission network, and the higher capacity links in the access network connecting a group of base stations and hub sites.
Transmission-Network Planning and Design
291
The preliminary equipment list is presented on a network basis, so the list is not broken down to show equipment per subnetwork or lower levels. The equipment list will detail all transmission equipment included in the preliminary network plan. The level of detail is sufficient to enable forecasting and ordering of equipment. Prices are not part of the equipment list. The list of approved sites presents which of the proposed sites have been approved from a transmission point of view. The list will also present rejected sites together with a short description of why they were rejected. Every site has to have either access to leased T1/E1 facilities or LOS for the microwave system. The MW LOS survey report contains all relevant site data for surveyed sites from a transmission point of view, such as site name, geographical position, street address, site height, and proposed tower or pole height. It also includes line-of-sight information on all surveyed paths detailing path name, bearing (azimuth), distance, a short path description, and LOS confirmation. Photographs for each surveyed path are included in the report. The transmission-network plan describes the final transmission-network design based on actual site positions. It includes trunk routing diagram, base station to switch logical connection diagram, timeslot routing diagrams and the network management system plan. The report is written on a detailed level and it accounts for each individual link. The diagrams are normally supplied as attachments to the report. The timeslot routing diagrams, are only produced for links connected to DACS equipment, if any. The layout is a network drawing showing the access and backbone (core) transport networks. Link routing as well as capacities and media are indicated and the layout will show each individual link. The final equipment list is presented according to customer preferences and on a network basis, subnetwork basis, per transmission site, or any combination thereof. The equipment list will detail all transmission equipment included in the final network design. The level of detail should be as close to the final requirements and final transmission-network plan as possible. The MW path and frequency-planning report (for microwave systems) includes a specification of recommended channels with exact frequency, output power, polarization, and antenna size and height. The report will also present the estimated threshold degradation for each link. The degradation value is a theoretically calculated value. Furthermore, the report includes performance prediction for each link. The calculated quality and availability figures are presented along with corresponding objectives. The calculations are based on North American and International (ITU) theoretical models.
292
Transmission Systems Design Handbook for Wireless Networks
7.4 Leased Lines in Wireless Networks Leased lines (known also as leased circuits or leased facilities) in the transmission network (usually T1/T3, E1/E3) are the optimal solution in the short term for the fast transmission-network deployment. Minimal resources are required from an engineering, implementation, project management, and operations perspective when compared to a microwave or fiber-optic build, but represent high recurring costs due to monthly fees. For example, orders are placed with a carrier when a site location is confirmed. Careful scheduling with carrier is needed to ensure minimal disruption or noise when installing facilities in a building and to meet the overall objective of delivering DS1s on time for turnup. Once in operation, a maintenance problem is simply addressed by placing a phone call to the carrier to report the problem. Unfortunately, there are a number of caveats when using leased facilities. First and foremost is the high recurring cost. It will last ad infinitum unless the intention is to replace it with a privately owned system. In addition to the recurring cost, there is a typical service chargeusually several thousand dollars per facilityand a construction chargefrom several thousands to hundreds of thousands of dollarsif the carrier is required to build facilities to a site. Another problem with leased facilities is the limited capacity and also the fact that leased transmission facility cost is linear with respect to bandwidth. For example, if twice the capacity is desired then twice the facility must be leased, thereby doubling the cost. Leased facilities are also inflexible in network design and it is a lengthy and costly process to reconfigure leased lines to address ongoing changes in the network. Finally, for any maintenance problems on the leased line, the time to repair is at the mercy of the carrier. Therefore, it is important to establish a good working relationship with the facility provider (telecommunication provider) in order to understand the interworkings of the company and hopefully to influence key departments when help is needed. Owned facilities in the wireless network usually require planning, building, and maintaining the microwave network. This process starts almost immediately after the launch of the wireless network, and by year three, most of the leased facilities should be replaced by the completely owned and operated transmission network. Details of the microwave network planning design and build-out are discussed in Chapters 6 and 9 of this book. Today, customers are also able to lease different services based on ATM from the local telecommunications company. Recommendation ITU-T G.176 provides transmission performance planning guidance to network and service planners who are responsible for the integration of ATM
Transmission-Network Planning and Design
293
technology (e.g., digital processing equipment, multiplexers, and switches) into the PSTN [4]. This recommendation recognizes and addresses the interconnection of other networks (e.g., private networks and digital cellular networks) with the PSTN and the continued need to support existing voiceband services. 7.4.1
Leased Access and Core Transmission Networks
Base station access networks in most parts of the world (except in North America) are most often both owned and operated by the mobile network operator as a strategic asset. The main reasons for this are profitability, together with the control it gives over rollout and services in terms of quality and the timely availability of new connections. Microwave access dominates in base station access network implementations, as it is often the fastest means for network rollout and capacity expansion. Using microwave transmission, an operator saves on operational expenses compared to laying his own cables or leasing connections. At least two-thirds of all base station connections are based on microwave. Access is copper-based transmission, when copper lines are widely available at an attractive price. This is particularly attractive if the operator owns copper lines. The access technology for the dark copper case has been high-bit-rate digital subscriber line (HDSL). However, copper-based connections may not always provide the same flexibility and controllability in rollout as wireless (microwave) alternatives. Optical fiber is constantly gaining a greater foothold and has a clear role in future (3G) transmission-network implementations, providing transmission capacity to regional hub sites, from where the capacity is further distributed by using wireless or copper media up to individual base stations. Optical fiber network may be used in the future to provide access even up to the base station sites. Hub sites are needed in base station access transmission network for grooming traffic and managing protection, especially when distances between the BSC and base station increase. The connections between switches (core network transmission), extending also to BSC connections, are of higher capacity than the base station access connections, and the distances to cover are longer. Not every mobile network operator has so far had the resources to build its own transmission for these long circuits, so these may be leased from a country-wide transmissionservices provider as protected E1 (2 Mbps) or T1 (1.5 Mbps) connections. BSC-to-MSC (sometimes called fronthaul) and interswitch BSC-BSC traffic (sometimes called sidehaul) is already carried mostly over optical fiber, even if this may be invisible to a customer subscribing to a protected N × 2
294
Transmission Systems Design Handbook for Wireless Networks
Mbps (1.5 Mbps) leased-line service. The existence of fiber-optic networks, and SONET/SDH technology in particular, compared to N × 2 Mbps (or 1.5 Mbps) leased-line-service prices, has led many operators to reconsider their core transmission sourcing policies. In Europe, many of the existing N × 2 Mbps leased lines have been converted into SDH leased transmission, providing 63 × 2Mbps capacity. The main reasons for this change are the savings achieved in operational expenses and the fast upgrades in capacity. The only changes involved for mobile network operators have been the purchase of an SDH multiplexer, located at their central transmission sites, to provide 2-Mbps electrical interfaces with the leased SDH stream. If the leasing alternative is chosen for backbone services, for greatest cost-efficiency, leasing should take place at the STM-1 (VC-4) level, not the N × 2Mbps level. With current leasing prices, the SDH alternative may bring the annual operating cost of leased connections down by 70% to 80% for full SDH lines, easily justifying the modest investment needed in SDH terminal multiplexers. With partially filled lines, the savings are naturally less than this. The second step, already directly taken by a number of mobile network operators, has been to acquire their own SDH terminal equipment and lease dark fiber to carry the signal. In areas where leased dark fiber is rare, the solution has been a rollout of owned or partially owned fiber-optic cable infrastructure between key locations in the networkincluding the central switching sites and a number of other strategically important hub sites. 7.4.2
Dedicated Leased Service
Dedicated service is a circuit between fixed end points. The circuit is leased (or owned by an end user) from a common carrier. A dedicated line is obtained from the carrier on a monthly basis for unlimited usage between the individual points to be connected [5]. The service is available 24 hours a day, 7 days a week, 52 weeks a year, for exclusive use by that user. This service may also be referred to as a leased line, private line, or nailed-up circuit, and its primary features include the following: • Point-to-point service; • Exclusive use 24-7; • Always available; • Payment for mileage.
Transmission-Network Planning and Design
295
T1 digital communications were introduced in the early 1960s to reduce the amount of copper cable needed to carry the same number of telephone conversations as analog communications. The term T1 circuit is commonly used to identify a multiplexed 24-channel, 1.544-Mbps digital data circuit providing communications between two facilities or from a local service provider. T1 refers to the transmission of a DS1-formatted signal onto a copper, fiber, or wireless medium for deploying voice, data, or videoconferencing services. The T1 is part of an extensive digital hierarchy that starts with 24 DS0s at 64 Kbps. These individual DS0s are used to provide voice or digital data to support point-topoint or network applications. By combining multiple DS0s, a high-speed interface can be provided to support a synchronous interface to a LAN router or voice PBX. For distances longer than one mile, a repeater is placed every mile to regenerate the signal. E1 is the European equivalent of the American T1. Although both E1 and T1 use 64-Kbps channels, they differ in many aspects; E1 is a point-topoint dedicated 2.048-Mbps communications circuit that carries 32 channels while T1 has 24 channels. Out of these 32 channels, 30 channels transmit voice and data. Unlike T1, E1 always provides clear 64-Kbps channels. Timeslot 16 (TS16) is used for signaling and carrying line supervision (such as whether the telephones are on-hook or off-hook) while timeslot 0 (TS0) is used for synchronization, channel control, and framing control. There are two options for the physical media: • 120-ohm twisted-pair cabling, typically foil shielded. This is called a
balanced interface and uses a DB-15 or 8-pin modular connector. T1 always uses twisted-pair. • 75-ohm coaxial cable. This is called an unbalanced interface because the two conductors do not have equal impedance to ground, and uses a BNC connector. E1 can be either coax or twisted-pair. T3 is a physical transmission medium, which varies from DS3, which is the actual data, video, and voice that is transmitted over the medium. DS3 can be transmitted over microwave radio, fiber optics, or 75-ohm coaxial cable (in-house wiring with distance limitations of 450 ft). The line coding for a T3 line is bipolar three-zero substitution (B3ZS). The framing formats available are M12/M23, M13, or C-bit parity. DS3 can be delivered as channelized or nonchannelized. Channelized DS3 is delivered as 28 individual DS1s and 672 individual DS0s. Each DS1
296
Transmission Systems Design Handbook for Wireless Networks
may come from a remote location. The telephone companys central office will do the subdivision of the DS3 to each site. The nonchannelized DS3 does not involve DS2 or DS1 multiplexing. This service is delivered as a T3 pipe with the bandwidth being 44.2 Mbps. It is generally used in point-to-point applications (one customer sending data to one remote site). Any subdivision of bandwidth is performed at each customer site rather than the central office. It is expected that in 3G wireless networks not only hub sites, but also some other heavily loaded cell-sites in the wireless network, may require close to this amount of bandwidth (capacity).
7.4.3
xDSL
xDSL is the generic name used to represent a wide variety of digitalsubscriber-line technologies including HDSL, ADSL, and IDSL. HDSL is the most widely available and used xDSL service in North America today. HDSL technology has been developed to allow the transmission of a standard DS1 signal over the outside plant wiring. With HDSL electronics at both the central office and the customers premises, it is possible to extend a full-duplex 1.544 Mbps (voice, data, and video) applications over two pairs of copper wire across private or leased copper facilities to distances of 9,000 ft (26 AWG) or 12,000 ft (24 AWG). All this can be achieved without redesigning the copper loop and without expensive repeaters. The specific rates achievable with DSL depend on factors such as the DSL technology used, the distance between end points, and the wire size. ISDN digital subscriber line (IDSL) is also commonly used for applications that require ISDN BRI signaling in a dedicated mode. IDSL can be extended up to 18,000 ft and can transmit digital data at rates up to 144 Kbps. HDSL technology also provides T1 capability for private networks by utilizing existing copper loop plant originally designed to carry much slower signals (usually POTS). Not only did the copper cable itself have limitations, but also things were done to this cable to make it even more unsuitable for high-speed data transmission. These actions primarily took two forms, like loading and bridge taps. Loading is a procedure where load coils were frequently added to loops longer than 18,000 ft. These load coils were essentially low-pass filters. That is, they passed without attenuation all voice frequencies but effectively blocked frequencies above the voiceband. This is disastrous for data communications, which depend on high frequencies to
Transmission-Network Planning and Design
297
achieve the desired speed of transmission [6]. The xDSL system cannot be supported on a loop with load coils. Bridge tap is any unterminated portion of a loop not in the direct talking path. A bridge tap may be a used cable pair connected at an intermediate point or an extension beyond the customer. For example, a drop wire that provided a second line to a home is left in place even after the second set of CPE is removed. Records of this were not always kept and assigning a particular copper pair to a high-speed data circuit is far from a sure thing. Bridge taps can bring many problems to data transmission and xDSL installations. 7.4.4
Switched Leased Service
Switched service is a circuit for which the end points may vary with each usage. The circuit is provided by a common carrier, which is routed through a switched network, providing circuit switching between public end users. There are two types of switched technology: circuit switched and packet switched. In a switched circuit, a call is established only for as long as needed and then the session is disconnected. Other characteristics of the switched leased services are as follows: • Physical path may vary; • Network is shared; • Customer connects on demand; • Customer pays for usage; • Service provider can oversubscribe capacity. 7.4.4.1 Circuit-Switched Network
A circuit-switched network is one where the user has full access to the connection when needed. The connection is often established by placing a call to another party. The user has exclusive use of the path. Switched 56 is a 56-Kbps digital data service that is purchased from some local exchange carriers or IXCs. It may be deployed on either two or four wires depending on the local carriers capabilities. In both four-wire and two-wire switched-56 applications, 56 Kbps is the standard network operating rate accessed, or access is required on an as demanded basis and 24-
298
Transmission Systems Design Handbook for Wireless Networks
hour-a-day connectivity is not required. The maximum distance is 18,000 ft on a 26-gauge wire without a repeater. ISDN was first established by the CCITT (now ITU-T) in 1980 to integrate an all-digital public switched telephone network. This is accomplished through ISDN that offers a full range of enhanced services supporting voice, data, and video applications through standard interfaces over a single twisted pair of copper. ISDN provides a means of integrating these services and modernizing communications networks to provide information movement and management efficiency. The two types of ISDN service are basic-rate ISDN (BRI) and primary-rate ISDN (PRI). BRI can transmit data up to 128 Kbps. PRI (which is transmitted over a T1 line) can transmit data up to 1.536 Mbps. A local directory number (LDN) is a customers seven-digit ISDN phone number. A service profile identifier (SPID) is a unique identifier that is used to represent the service and feature identifiers of a particular ISDN line or service provider. This number generally is 10 or more digits long and includes the LDN. 7.4.4.2 Packet-Switched Network
In a packet-switched network, data is carried in the form of packets. This data would be given an ID on a per-packet basis and sent across the network in the most efficient way. Frame relay is a digital-packet network that provides all the features and benefits of a dedicated DDS or T1 network, without the expense of multiple dedicated circuits. Frame relay is deployed over the same services used to deploy DDS and T1. In a frame relay network, circuits are connected to a packet switch within the network which ensures that packets are routed to the correct location. Frame relay is optimized for use over higher-speed and very-low-errorrate data circuits. To reduce latency in FR networks, the switches do not perform error correction (other than discarding corrupted frames) or flow control (other than setting forward explicit congestion notification and backward explicit congestion notification bits in the frame header). If the user equipment (UE) does not react to those notifications, then the network discards bits when it gets congested. All other functions of error congestion and flow control are left to the equipment on the customers premises. One of the main advantages contributing to growth in the frame-relay industry is the cost effectiveness of this type of service. Frame-relay service is a cost-effective solution for networks with bursty traffic requiring connections to multiple locations and where a certain degree of delay is acceptable. It also allows a voice circuit to share the same virtual connection as a data circuit, again saving money.
Transmission-Network Planning and Design 7.4.5
299
Higher-Speed Switched and Nonswitched Services
Until recently, the primary method by which businesses and institutions obtained nonswitched private line connections between their locations was to use dedicated 1.5-Mbps T1 lines and dedicated 56-Kbps digital private lines leased from telecommunications carriers, including the local exchange carriers. Some larger businesses and institutions have used higher-speed 45Mbps private lines for point-to-point connections. Recently, new types of digital services, including frame relay, SMDS (a public network service designed primarily for LAN-to-LAN interconnection), and ATM, have been introduced by telecommunications carriers, including local exchange carriers. SMDS is a packet-switched connectionless data service that allows the destination to be specified independently for each packet. Frame relay and ATM are currently nonswitched services that utilize predetermined destinations for traffic; switched versions of these services are under development. All these services offer the advantages of improved sharing of facilities (fibers, terminations on electronic equipment) through statistical multiplexing. These new services, particularly ATM, can also support advanced multimedia applications that require high data rates and low delay variability between communicating end points. 7.4.6
Leased Lines Network Build Out
The major steps in the process of any transmission network include determining the total transmission requirements for each site in the core and access network. After creating a transmission demand matrix (i.e., a matrix of all traffic nodes), identifying the transmission capacity requirements (typically in terms of the number of T1/E1 links required) between nodes would be the next logical step. From the topology and demand matrix, it is important to create a transmission link matrix, a matrix of all transmission nodes showing transmission-capacity requirements between these nodes. This matrix is different from the demand matrix in that it reflects the chosen topology; that is, it identifies the actual transmission links (physical connection), whereas the demand matrix identifies only the demand between two nodes (logical connection). The most important thing in designing the transmission network is to understand the existing network and design the best transmission network with rings, linear systems, and digital-access cross connects (DACS) to satisfy the demand matrix, diversity, flexibility, and topology for the wireless network that is being built. If the network is owned, this information will be well understood; however, if the transmission network is leased, information
300
Transmission Systems Design Handbook for Wireless Networks
on the providers network may be harder to get. Information on the tariffs, distance, and point-of-interconnect information from the provider of leased capacity, as well as local rules and regulations, will be required. It is important to determine the requirements for owned and leased transmission links (e.g., type, size, quantity) from the transmission link matrix. From these it is possible to specify the bill of requirements (BOR), which should specify the equipment quantities and configurations and leasedline quantities and configurations. These activities take place during the planning and design stage of the leased-lines network build-out. Other activities are listed below: System Planning
1. Identify private and public carriers and facilities. 2. Identify regulatory issues. 3. Establish standards and objectives. 4. Establish nondisclosure agreement (NDA) with carriers. 5. Conduct intercity interconnect analysis. 6. Establish procedures for circuit ordering. 7. Establish trouble reporting and escalation procedure. System Design
1. Analyze search ring information (from RF plan) and create a preliminary network design. 2. Perform primary (candidate) site-engineering analyses. 3. Select carrier(s) in each market. 4. Finalize carrier agreements. System Deployment and Testing
1. 2. 3. 4.
Prepare site package. Order leased facilities. Track carriers progress. Prepare ATP.
Transmission-Network Planning and Design
301
System Optimization
1. 2. 3. 4.
Interface with construction contractor for site engineering; Coordinate leased-circuit construction; Perform construction walk-through; Prepare site completion package and as-built documentation.
It is very important to notice that even when the transmission network is completely leased, it still requires a great deal of engineering effort from the wireless operators side. Transmission engineering cannot be left to the carrier(s) to make decisions on the leased-circuits routing, installation, project management and testing without any supervision from the wireless operator. One of the leading causes of delay in getting T1/E1 lines installed is an unprepared equipment room [2]. It is important to coordinate with carriers and cell-site construction teams the requirements for the equipment room or shelter. This could include ducts or cable ways for the lines, ac and dc power, adequate space, heating and air-conditioning, solid grounding, and a very common requirement by carriers for a plywood board on the wall for terminal blocks, NIUs, punch-down blocks, and so on. 7.4.7
Owned Versus Leased Transmission Networks
Every wireless-network project is different and requires detailed business case typeanalysis of owned verses leased transmission facilities. Microwave networks are usually the way to go if the wireless operator opts to build its own transmission network. Some of the advantages of microwave radios are listed below: • MW system meets superior reliability, higher security, and more
demanding performance and quality standards;
• User has total control over site access and restoration time; • From the performance perspective, carriers provide best effort and
average (not specific) availability for all the customers;
• Easy expansion, or future relocation, or both; • Overhead channel and protection (path, hardware, or both); • MW radio has an operational life long after the leased-line payback
time (24 years).
302
Transmission Systems Design Handbook for Wireless Networks
Some of the disadvantages of microwave networks include the following: • High capital investment on day one (unless the MW equipment
supplier is financing); • Microwave system design required;
• Frequency coordination and spectrum licensing required; • Microwave network operations and maintenance required.
Let us assume that the 4T1 microwave-system cost, fully engineered, installed, and tested, is $56,000 per hop (see Table 7.1). Not only is the total cost after four years still the same, but it provides for system expansion for an additional three T1s, and thus the cost per T1 is actually only $14,000. On the other hand, a leased T1 has recurring costs of $500 per month (typically, in the United States), and after four years the cost for the leased T1 would grow to $26,000, or almost double that of the microwave T1 circuits. In many countries where leased T1/E1 lines are not readily available or cost a lot more, it would be even easier to prove the advantage of building a microwave network.
7.5 SynchronizationStratum, BITS, and GPS 7.5.1
Introduction and Historical Overview
Digital network connectivity is dependent upon the availability of a reliable synchronization source to provide a timing reference to the network elements. This chapter discusses the hierarchical clocking scheme, and its applications within public and private networks. Networks that use asynchronous Table 7.1 Leased Lines Versus Microwave Comparison MW 4T1 ($)
Leased 1T1 ($)
Up-front charges
56,000 (turnkey)
2,000
Monthly charges
0
After 12 months
56,000
8,000
After 12 months per T1
14,000
8,000
After 48 months
56,000
26,000
After 48 months per T1
14,000
26,000
500
Transmission-Network Planning and Design
303
digital multiplexers, synchronous multiplexers (SONET), or both, are briefly discussed. Synchronization networks provide timing signals to all synchronization network elements at each node in a digital network. These timing signals are traceable to a highly accurate primary reference source (PRS) clock of Stratum 1 quality [7]. The aim is to ensure that all outgoing transmissions from a digital network node have the same average frequency. Buffer elements are used at important transmission interfaces to absorb differences between the average local frequency and the actual short-term frequency of incoming signals, which may be affected by phase wander and jitter accumulated along the transmission paths. A sync network has two major partsinteroffice and intraoffice. The interoffice network consists of a primary and a secondary DS1 link, carrying timing between offices in a hierarchical relationship. Intraoffice timing distribution is based on the concept of a Building Intergrated Timing Supply (BITS) master clock, providing timing to all other digital equipment in the office. Telephone companies and long-distance carriers started using digital technology in the 1960s to improve service and lower the costs associated with the transmission of analog services. The first digital system installed was known as the T1 transmission system, and it was used primarily for pair gain, or the reduction in the number of pairs used to carry voice traffic. Private companies usually have constructed telephone systems to obtain more efficient communications between locations. Originally, these networks were small copies of the analog transmission and switching systems used by telephone companies. The typical data network was constructed using analog modems in a separate point-to-point network. This scheme often led to two separate networks, one for switched voice and one for data transmission. Unfortunately, the voice quality of these analog private networks proved to be less than desirable due to the low bandwidth and losses of the analog channels. Data transmission over the analog network suffered from error bursts, which resulted in lower speeds. The new digital system provided better voice quality, fewer wires between offices, higher reliability, and improved data transmission. Initially, T1 was an asynchronous system. Each pair of end terminals ran at its own clock rate, and each terminal used its receive timing to demultiplex the incoming signal. The transmit and receive sides were independent of one another. Later, when digital channel units were used, one end terminal was designated as the master and had its own timing reference. The other end terminal was a slave and derived timing for its transmit side from the
304
Transmission Systems Design Handbook for Wireless Networks
data being received (looped timing). This arrangement worked as long as the end terminals were no more complex than a channel bank. The growth of T1 (DS1) networks to include higher multiplexing rates and long-distance DS1 connectivity introduced various synchronization problems. Digital switches with DS1 port interfaces exemplified the shortcomings of an asynchronous system. If the two switch clocks were not at the same frequency, the data would slip at a rate dependent on the difference in clock frequencies. A slip is defined as a one-frame (193 bits) shift in time difference between the two signals in question. This time difference is equal to 125 microseconds. Slips were not considered a major impairment to trunks carrying voice circuits. The lost frames and temporary loss of frame synchronization resulted in occasional pops and clicks being heard during the call in progress. With advances in DS1 connectivity, however, these impairments tended to spread throughout the network. To minimize this problem, a hierarchical clock scheme was developed, whose function was to produce a primary reference for distribution to switching centers in order to synchronize the toll switches. Local switching in that era was primarily analog, so that synchronization was not required at the end offices. Later, digital switches and direct digital services or networks (DDS or DDN) became common at the end offices, providing digital services to customers. This meant that timing had to be distributed to local levels.
7.5.2
Strata
As shown in Figure 7.5, the resulting stratum hierarchy evolved into four levels. Stratum 1 is the primary reference. Stratum 2 is used at toll switches, while Stratum 3 is used at local switches. Channel banks and end terminals that use simple crystal oscillators are known as Stratum 4 devices. Recently, SONET networks have created the need for a clock stratum level better than Stratum 3, which is called Stratum 3E. When the Bell System broke up into the local service providers and long-distance carriers, the timing hierarchy became less well defined. Now each local company could no longer take its timing from the long-distance carrier, but had to engineer a system, either a hierarchy or otherwise, to distribute timing to their offices. This made everything more difficult because failures in the transmission systems may cause islands, or areas without a reference to Stratum 1. In such a case, the island is in holdover at whatever stratum level has been provided. Even if a particular network is still traceable to Stratum 1, the traffic of concern may be coming from such an island and
Transmission-Network Planning and Design Stratum 1 Telco (primary reference source)
Toll switching center
Toll switch
Stratum 2 Telco
305
Toll switch
Toll switch to other local switches
Stratum 3 Telco
Stratum 4 Telco or end user
Local switch
Local switch
to other digital terminating locations Channel banks
Local switch
T1 multiplexers
DACS
Digital switches
Figure 7.5 Digital network hierarchy.
will therefore slip at some rate. Having a Stratum 1 source is, in itself, no guarantee of a slip-free network. Voice equipment tends to reacquire frame synchronization quickly, resulting in a pop or click, which is not usually a problem. Data circuits lose some number of bits depending on the data rate being transmitted, and on whether or not FEC is being used. Some multiplex equipment that provides add and drop services interrupt all output trunks while a new source of synchronization is acquired. Such interruptions, if due to circuit noise, may render a network temporarily useless, as the slip causes further slips downstream (error or slip multiplication). A clock system provides a stable frequency source during circuit impairments. The connected equipment will not be affected until the clock holdover drift results in a slip. A stable clock will change a network that experiences problems two or three times a day to one that maintains timing through a major trunk outage. The network will continue to operate without impairment until the outage is repaired, as long as the repair time is comparable to the time of the first frame slip. Since occasional slips will always occur, the best one can do is to minimize their rate of occurrence. Through careful network engineering of the clock systems, near
306
Transmission Systems Design Handbook for Wireless Networks
perfect timing may be achieved at a reasonable cost with excellent reliability and maintainability. The ANSI standard titled Synchronization Interface Standards for Digital Networks (ANSI/T1.101-1987) was released in 1987. It was revised in Committee T1X1 for release as ANSI/T1.101-1998. This document defines the stratum levels and minimum performance requirements. The requirements for the stratum levels are shown in Table 7.2, which provides a comparison and summary of the drift and slip rates for the strata clock systems [8]. It is important to note the following in Table 7.2: 11 1. Stratum 2 is typically less than 2.5 × 10 /day; 2. Stratum 2 drift results in more than two weeks to a frame slip; 3. Stratum 3E enhanced; The typical performance of which is less than one slip in 36 hours or 9 × 10 10/day.
To calculate slip rate from drift, one assumes a frequency offset equal to the above drift in 24 hours, which accumulates bit slips until 193 bits have Table 7.2 Clock Strata Requirements
Stratum Accuracy Adjustment Range
Pull-In Range
Stability Time to First Frame Slip
N/A
72 days
1
1 × 10−11
2
1.6 × 10−8 Must be capable of synchronizing 1 × 10−10/day (1) 17 days (2) with clock with accuracy of +/−1.6 × 10−8
3E(3)
34 × 10−7 −6
N/A
4.6 × 10−6
5 × 10−9/day
Must be capable of synchronizing 1 × 10−8/day with clock with accuracy of +/−4.6 × 10−6
7 hours (4)
3E
1.0 × 10
3
4.6 × 10−6 Must be capable of synchronizing 3.7 × 10−7/day with clock with accuracy of +/−4.6 × 10−6
6 minutes (255 in 24 hours)
4E
32 × 10−6
Must be capable of synchronizing Same as with clock with accuracy of accuracy +/−32 × 10−6
Not yet specified
4
32 × 10−6
Must be capable of synchronizing Same as with clock with accuracy of accuracy +/−32 × 10−6
N/A
3.5 hours
Transmission-Network Planning and Design
307
been accumulated. Drift rates for various atomic and crystal oscillators are well known and are not usually linear or not necessarily continually increasing. Stratum 1 is defined as a completely autonomous source of timing that has no other input, other than perhaps a yearly calibration. The usual source of Stratum 1 timing is an atomic standard or reference oscillator. The minimum adjustable range and maximum drift are defined as a fractional fre−11 quency offset of 1 × 10 or less. At this minimum accuracy, a properly calibrated source will provide bit-stream timing that will not slip relative to an absolute or perfect standard more than once every 4 to 5 months. Atomic standards, such as cesium clocks, have far better performance. A Stratum 1 clock is an example of a PRS as defined in ANSI/T1.101. Alternatively, a PRS source can be a clock system employing direct control from coordinated universal time (UTC) frequency and time services, such as GPS navigational systems. The GPS may be used to provide high-accuracy, low-cost timing of Stratum 1 quality. A Stratum 2 clock system tracks an input under normal operating conditions and holds to the last best estimate of the input reference frequency during impaired operating conditions. A Stratum 2 clock system requires a −8 minimum adjustment (tracking) range of 1.6 × 10 . The drift of a Stratum 2 −8 with no input reference is less than 1.6 × 10 in 1 year. The short-term drift −10 of the system is less than 1 × 10 in 24 hours. If one interprets this specifica−10 tion as a drift of 1 × 10 each 24 hours, this amounts to a frame slip rate of approximately one slip in 7 days when the Stratum 2 clock system is in the hold mode. Stratum 3 is defined as a clock system that tracks an input as in Stratum 2, but over a wider range. A Stratum 3 clock system requires a minimum −6 adjustment (tracking) range of 4.6 × 10 . The short-term drift of the system −7 is less than 3.7 × 10 in 24 hours. This amounts to approximately 255 frame slips in 24 hours while the system is holding. Some Stratum 3 clock equipment is not adequate to time SONET network elements. Stratum 3E, which was defined in Bellcore documents, is a new standard created as a result of SONET equipment requirements. Stratum 3E tracks input signals within 7.1 Hz of 1.544 MHz from a Stratum 3 or better source. The drift with no input reference is less than 1 × 10−8 in 24 hours. Stratum 4 is defined as a clock system that tracks an input as in Stratum −5 2 or 3, except that the adjustment and drift range is 3.2 × 10 . Also, a Stratum 4 clock has no holdover capability and, in the absence of a reference, free runs within the adjustment range limits. The time between frame slips can be as little as 4 seconds. Stratum 4E is a proposed new customer premises clock
308
Transmission Systems Design Handbook for Wireless Networks
standard which allows a holdover characteristic that is not free running. This new level, intended for use by customer-provided equipment in extending their networks, is not yet standardized. 7.5.3
General Timing Planning Rules in Transmission Networks
Inadequate timing may produce problems in any digital network so the objectives have to be set very early in the planning process: • The number of frame slips per day/month/year to be allowed under • • • •
impaired network conditions; The timing sources available from the carrier or local telephone company to help solve the problem; The degree of redundancy required; The locations of the timing systems and the stratum levels desired; The management and maintenance of the system.
One major problem encountered after designing a timing network is evaluating its performance. Standard tests for large switching systems and clock distribution systems require the use of a cesium or rubidium standard. A much simpler method of verifying performance is to use test equipment that displays bit slip, but reference is required in order to use such equipment effectively. A Stratum 1 clock may control Stratum 2, 3E, 3, 4E, or 4 clocks while a Stratum 2 clock may drive Stratum 2, 3E, 3, 4E, or 4 clocks. A Stratum 3E clock may drive Stratum 3E, 3, 4E, or 4 clocks. A Stratum 3 clock may drive Stratum 3, 4E, or 4 clocks. A Stratum 4E or 4 clock is not recommended as a source of timing for any other clock system. Because of the narrower capture and adjustment range of the higherstrata clock systems (2 is higher than 3, and so on), driving a Stratum 2 clock from a Stratum 3E or 3 clock is not recommended. In fact, it will not work under some transmission-impaired conditions. Also, extreme care must be taken in network applications where more than one Stratum 1 source is used to ensure that these sources are accurate and traceable to some other standard. Another standard commonly used to check a Stratum 1 clock sources accuracy is the GPS system. A GPS receiver can also be used directly as a source of Stratum 1 quality (commonly used for the cell-site equipment timing in cdmaOne and CDMA2000 networks).
Transmission-Network Planning and Design
309
Stratum 1 clock administration, operation, and maintenance can be a costly effort. Atomic sources may not have long maintenance-free operating intervals and may experience failures without giving an indication that the source is off frequency. In addition, if a Stratum 1 source of timing is shown to be inaccurate, the network must be able to accept another networks timing until the problem is corrected. Thus, GPS is attractive in order to assure accuracy and minimize cost. Initially, private digital networks were arranged for point-to-point, single trunk. A channel bank at one end provided the source of timing for the distant end terminal bank, known as the slave. The slave terminal derived or extracted timing from the transmitted data. With no intervening transmission equipment to provide a source of timing, this arrangement rarely experienced synchronization problems. When more than one trunk was connected point-to-point, all equipment at one end used the same clock, which was established through a clock-distribution system. The other end was slaved as before, and timing problems were not encountered. The public-switched network gets more complex every day, comprising numerous long-distance carriers, local operating companies, as well as the competitive access providers. In addition, networks are no longer point-topoint, or PMP. They now include add and drop, as well as ring configurations. Designating the location of master and slave terminals becomes quite difficult when point-to-point circuits are used for add and drop services, switched services, or services passing through certain customer multiplexing equipment. However, reliable network timing can be achieved if one location is designated as the master, and Stratum 2 or 3E clock systems are installed at all other sites. While todays point-to-point services and tariffs are typically untimed, telephone companies and long-distance carriers can install equipment to administer, operate, and provide service observation (such as DACS). Without network timing, the timing is that of the customers equipment (leased lines), and may create jitter and wander problems with SONET transmission equipment in the network. The advantage of having the local or long-distance carriers pass the DS1 circuit through a DACS is that the trunk is timed by the carrier. This then provides a source of timing directly traceable to a Stratum 1 reference. In a network with carrier imposed timing on the trunk or trunks, all private network sites should be configured as slaves, and equipped with Stratum 3E clocks to provide timing to the equipment.
310
Transmission Systems Design Handbook for Wireless Networks
A Stratum 2 clock system should not be used in networks using carrier-derived timing. This arrangement may not work when the transmission system timing is in its holdover condition at the Stratum 3 level at the serving central office or long-distance terminal. In all cases, careful examination of the network and equipment is required. A Stratum 1 clock system may always be used, since it should be identical to the network providers clocks. Only in the case of satellite networks will a Stratum 1 remote clock lead to problems. Clock-system arrangements are based on intended use. A long-distance carrier may elect to provide and administer Stratum 1 clock sources at all major toll switch locations. Another less costly method is to provide one or two Stratum 1 sources, followed by a distribution of the timing information on dedicated DS1 paths to Stratum 2 systems at each switching center. It is possible to use GPS to control rubidium sources. These sources would then distribute timing to Stratum 2 clock systems. In some cases, providing each location with a Stratum 1 GPS system is the most efficient solution. Local telephone companies may choose one DS1 path from one incoming group as a primary reference, and another DS1 path from another incoming group as a secondary reference. These two references feed a Stratum 2 system at the toll level and a Stratum 3 system at the local-switching system level. Such a system provides reliable timing at the local level. As an alternative, many local telephone companies, because of problems they have experienced while using someone elses timing, are beginning to consider the use of Stratum 1 systems of their own. Another method of distribution timing is via older analog microwave links. Many of these furnish a source of synchronization, which is transmitted as subcarriers for synchronizing the single-side band (SSB) and FDM equipment. These pilots may be converted to 1.544-Mbps DS1 framed 1s signals for use in distribution and timing of newer digital equipment, which may be using supergroups on the existing microwave to provide digital services. A private network may require a Stratum 3E clock at each location to provide a local source of timing. The Stratum 3 clocks are referenced to the telephone company by bridging one of the incoming DS1 signals to extract timing information. 7.5.4
Interoffice Distribution
Timing information is distributed to offices through a hierarchical series of levels, starting with a PRS. Clocks are grouped into stratum levels, based on
Transmission-Network Planning and Design
311
their ability to maintain accurate timing if their reference fails (holdover mode). Stratum 1 is the most accurate; it is by definition a stand-alone PRS, which does not have an external reference. Stratum 4 clocks are the least accurate; they have no holdover requirements and are typically found in DS1 terminal equipment, such as D4 channel banks. A given clock must be able to track a reference from a free-running clock of the same or higher stratum level. BITS clocks in central offices are Stratum 2, Stratum 3, or Stratum 3E, with Stratum 2 being used in the larger offices where holdover drift can affect hundreds or thousands of outgoing trunks. A sync network is designed so that a clock always receives timing from a clock of equal or higher stratum level. This ensures that if an upstream clock enters a hold mode, the downstream clocks will be able to track it. Stratum 2, 3E (3E Enhanced), and 3 clocks are provided with primary and secondary timing reference inputs, with automatic switching between the two if either one fails. Timing distribution is typically done with traffic-carrying DS1 signals over paths selected for best availability. However, using traffic-carrying DS1s for this purpose is discouraged when using SONET transmission equipment. The following are the major rules used in designing a synchronization network: • An office BITS clock can receive its reference only from another
office or offices of the same or higher stratum level. A higher stratum is preferred, provided reliable diverse DS1 paths from that office exist, but is not necessary.
• For a given BITS clock, a DS1 facility with the highest availability
should be selected for the primary reference input from an upstream clock. A facility with the next highest availability should be selected for the secondary reference, preferably on a diverse route from a different upstream clock. Its historical failure record, installation or rearrangement activity, facility length and type, protection switching, and the number of repeaters or multiplexers on the path determine the availability of a path. There may be specific equipment types that must be avoided.
• No timing loops are allowed in the sync network for any combina-
tion of primary and secondary facilities. The potential for loops exists when either primary or secondary reference signals are passed between clocks of the same stratum. Loops are avoided within an
312
Transmission Systems Design Handbook for Wireless Networks
office by distributing timing from the BITS clock in a star network. SONET rings, at present, cannot operate with redundant timing, as they are by definition a loop. • There are no fixed rules for the number of BITS clocks that can
be used in cascade. Any clock can track any other clock of equal or higher stratum level, and it will filter out most jitter and shortterm wander introduced along the timing reference path. However, long-term wander (over periods of hours) and possible phase transients will accumulate along the cascaded paths from the PRS to a given clock. Furthermore, the failure of a reference path will affect all downstream clocks. For these reasons, the number of cascaded clocks should be minimized as far as possible and be consistent with the use of whatever highly reliable reference paths are available in the network. As an example, it would be preferable to time a Stratum 3 clock from another Stratum 3 clock upstream, which is in turn timed from a Stratum 2 clock through reliable facilities, rather than use a direct path of questionable reliability to that Stratum 2 clock. The objective should be to maximize the overall availability.
• According to [9] it is recommended that no more than two Stratum
3 or 3E offices be timed in tandem, and where possible, the timing distribution should be limited to only one Stratum 3 office. This reference applies to large telecommunications synchronization networks.
7.5.5
Intraoffice Distribution
The BITS clock system is the preferred method of distributing timing within an office (Figure 7.6). Redundant hardware and automatic switching between primary and secondary reference inputs provide a high degree of availability. The BITS clock supplies timing directly to all digital equipment in the office requiring synchronization, usually by means of DS1 framed 1s, or a 64-Kbps composite clock. Many network elements have primary and secondary timing ports; the signals to these should be taken from different output cards of the BITS system. The BITS clock may also provide primary and secondary timing signals, through interoffice sync network paths specially selected for high availability to other BITS clocks in downstream offices.
Transmission-Network Planning and Design
313
GPS
DS1 BITS
64 Kbps composite clock
DS1
DS1
Digital switches
DCS
DS1 SONET
Channel banks
DDS
Traffic Timing Figure 7.6 BITS clock system.
In the wireless network, the BITS clock is placed at the BSC/MSC location (switch office) and receives the reference clock from the PSTN (T1/E1 line) or GPS receiver installed at the premise. 7.5.6
SONET Network Timing
SONET networks are designed to operate in a maintenance mode when not timed. In this mode, the jitter may be larger than that of a properly timed network. It has been noted by many that for voice traffic, and even for digital traffic between switches and digital cross connects, no impairments are noticed, even when the SONET system is untimed. However, it is recommended that all SONET networks be timed to minimize phase noise (jitter and wander). With SONET, the multiplex equipment itself must be timed. If the timing of the dropped DS1s is different from the multiplex timing, there will be unsatisfactory jitter and wander performance, which may or may not affect the connected equipment. By timing all network sources of DS1 signals from the same reference and making sure that the reference chosen is the same as the SONET
314
Transmission Systems Design Handbook for Wireless Networks
reference, problems can be minimized. Some of the more important considerations for synchronizing SONET network elements are as follows: • External timing for SONET NEs from a BITS clock is the preferred
mode, where BITS is available. • When BITS timing is not available, other SONET timing modes (line, through, or loop timing) should be chosen in such a way as to avoid timing loops and to minimize the lengths of timing paths. Rings may only be fed in one direction with a single reference. • If a string of more than one SONET ADMs exists between BITS timed offices, no secondary references should be used. This also applies to ring configurations; timing should be passed in one direction only, between line-timed ADMs. • Caution is advised when using DS1s carried on SONET for synchronization distribution; these DS1s are subject to phase transients (pointer adjustments) that may not meet short-term stability requirements. It may be necessary to pass timing this way, because customers may have no other choice of a timing source. A typical private network timing is shown in Figure 7.7. It is necessary to know if the carrier can supply a source of Stratum 1 traceable timing. For purposes of this example, the carrier identifies two T1 trunks at locations A and B, which will be timed by the network and which can be traced to the carriers Stratum 1 source. At location A, a synchronization timing system connects to one of the incoming trunks timed by the network. The outputs of the synchronization system drive the clock inputs of the multiplex equipment. Another synchronization timing system at location B is connected in the same manner as at location A. 7.5.7
Synchronization: Issues in PCS Networks
New digital network technologiesCDMA, TDMA, and GSMhave stringent requirements for network synchronization and timing accuracy. In PCS and cellular systems, these requirements appear both in the wireless and in the wireline parts of the network, as well as at the points of interconnection with other public networks required to complete a call. The over-the-air portion of every call requires accurate timing and the most obvious example is found in the CDMA world. To achieve the timing
Transmission-Network Planning and Design
MUX
315
Local Telco CO
Location A Timing Private network
MUX Location B
Longdistance carrier
Local Telco CO
Timing
Figure 7.7 Private network timing.
accuracy required for CDMAs soft handoffs, the IS-95 standard calls for synchronization using the GPS. The precise timing allows more calls to be put through the network, because a very narrow window can be used for timing the handoff between cells, thus freeing network facilities to handle other calls. Problems at the wireless-to-wireline interconnect could appear when the digital information being transmitted leaves the airlink to pass through a switching office. Unless the call is to another subscriber from the same network, it normally enters the PSTN and experience has shown that problems can originate here in several ways. If the wireless network uses its own clock, and if the timing has variations, a slip can occur at the interface between the wireless and the wired networks, and the call can be lost. Wireless networks often take their timing from the PSTN, however, and this can be the source for a second kind of problem. If the timing of the PSTN is inaccurate, it will affect the performance of the wireless network. Moreover, the problem may not show up at the interface to the PSTN from which the timing was taken,
316
Transmission Systems Design Handbook for Wireless Networks
but at a later point, perhaps at the interface to another operators network. A problem such as this can be subtle and hard to identify. Another risk of taking timing from a PSTN involves holdover. If the PSTN has a fiber network, signals must pass through numerous repeaters and regenerators. This constant regeneration of the signals can cause difficulties in some clocks, causing them to enter a state of holdover. For those wireless networks that use the PSTN timing to generate its radio frequencies, the effect can be a drifting of their frequencies into adjacent channels, causing channel interference and problems with regulatory agencies. The ideal solution for the wireless-network operator is to implement a highly accurate timing system based on synchronization with the GPS. This solution eliminates the uncertainties and holdover problems associated with reliance on timing from the PSTN, and it offers independence and control of the quality of the network synchronization. It has to be noted that this solution, GPS timing, is preferred in North America, but could be difficult or impossible to implement in some other parts of the world because of strategic, political, or other reasons. 7.5.8
Cell-Site Timing in Wireless Networks
As wireless communications systems move from analog to digital operation, their timing subsystems become both more complex and more critical. Operational timing in each cell of a wireless network is based on the GPS location and time standards. Timing references are downlinked from a satellite by the GPS receiver at an RBS. The clocks are then used by the RBS for timing mobile station calls and T1/E1 channel support. CDMA (cdmaOne and CDMA2000) is dependent on precision timing for its performance advantages over current analog systems. Accurate timing enables advanced features like soft handoffs, which allow a CDMA cellular user to move freely throughout a service area and be transferred from one base station to another without interruption in service or deterioration of signal quality. In the final stages of the encoding of the radio link from the base station to the mobile, CDMA adds a special pseudorandom code to the signal that repeats itself after a finite amount of time. Base stations in the system distinguish themselves from each other by transmitting different portions of the code at a given time. In other words, the base stations transmit time offset versions of the same pseudorandom code. In order to assure that the time offsets used remain unique from each other, CDMA stations must remain synchronized to a common time reference. All sector air interface transmissions are referenced to a common systemwide timing reference that uses the GPS time,
Transmission-Network Planning and Design
317
which is traceable to and synchronous with UTC (in the United States, the offical UTC is kept by the U.S. Naval Observatory). GPS and UTC differ by an integer number of seconds, specifically the number of leap-second corrections added to UTC since January 6, 1980. The start of CDMA system time is January 6, 1980, 00:00:00 UTC, which coincides with the start of GPS time. CDMA system time keeps track of leap-second corrections to UTC, but does not use these corrections for physical adjustments to the CDMA system time clocks. Customarily, cellular base stations receive their timing signals from the worldwide GPS, an array of satellites that beam timing signals accurate to within 300 ns to receiving stations located around the globe. Should the link fail, a backup system must be in place to maintain timing accuracy; if there is no backup, the cell goes down and communication is lost, which is an unacceptable condition. The GPS time standard is downlinked from the satellite as long as it is in range of the GPS receiver at the RBS. There are, however, intervals of time when a satellite is not in range (overhead), and no GPS time standard is available. This period without any GPS time update is called the GPS holdover time.
7.6 Transmission-Network Optimization 7.6.1
Daisy Chaining and Traffic Grooming
By definition, the term daisy chain in telecommunications refers to a wiring method where each telephone jack in a building is wired in series from the previous jack. Daisy chain is not the preferred wiring method, since a break in the wiring would disable all jacks downstream from the break. The same analogy goes for the wireless network, which has a large number of cell sites; some of the sites will be connected directly to the switch location (to BSC) and some will be connected through other cell sites (daisy chained). A problem on a certain T1/E1 circuit will cause problems not only on that immediately affected cell site, but on all of the cell sites downstream from it. It is important to note that the transmission-network planners have to be aware of the risk when the decision is made to use daisy chaining in order to save on transmission facilities. Grooming, or segregation as it is sometimes called, is done when E1/T1 lines contain a mixture of different types of service circuits. The (X) circuits are segregated to one E1/T1 line, while the other circuits (O) are
318
Transmission Systems Design Handbook for Wireless Networks
cross-connected to another E1/T1 line (Figure 7.8). DACS or mini-DACS equipment is usually used to perform this kind of functionality. 7.6.2
Voice Compression
7.6.2.1 Voice Coding Overview
One of the principal technologies enabling explosive growth in the field of digital communication is voice coding, in which an analog speech signal from a microphone is digitally sampled via an A-to-D converter and then efficiently compressed into a digital bit stream for transmission or storage. A corresponding voice decoder receives this bit stream and decompresses it back into a series of digital speech samples suitable for playback through a D-to-A converter and a loudspeaker. Voice coders can take a number of different forms each of which involves tradeoffs in terms of bit rate (i.e., degree of compression), complexity, and voice quality, as well as robustness. Voice coders are normally divided into two broad classes referred to as waveform coders and model-based speech coders. In a waveform coder the objective is to reproduce at the decoder the original speech samples on a sample-by-sample basis. A simple PCM waveform coder accomplishes this by quantizing each speech sample to one of a fixed number of levels. Assuming 8 bits (256 levels) are used per sample and the signal is sampled at 8 kHz, the overall data rate is 64 Kbps. More involved adaptive differential pulse code modulation (ADPCM) waveform coders apply prediction with differential quantization to reduce the data rate to 24 to 32 Kbps. In any case, the process of quantizing the speech samples adds quantization noise, which is usually audible as distortion in the decoded signal. The primary advantage of waveform coders is that if the data rate is kept sufficiently high the amount of DACS or mini-DACS XOXOXOXOX XXXXXXXXX XOXOXOXOX
XOXOXOXOX
Figure 7.8 Digital access cross connect.
OOOOOOOOO
Transmission-Network Planning and Design
319
distortion can be kept reasonably low. Hence waveform coders have historically been prevalent at rates over 16 Kbps. Another advantage of traditional waveform coders is that their complexity is typically modest, allowing them to be more readily implemented on early DSP devices. In contrast to waveform coders, model-based speech coders or vocoders use a parametric model to approximate short (1040 ms) segments of speech. For each segment, a set of model parameters are estimated and converted into a bit stream. The decoder converts this bit stream back into model parameters and then uses these parameters to synthesize a speech signal that is perceptually close to the original. In this approach no attempt is made to recreate the original speech samples, instead only the perceptual content as approximated by the model parameters is maintained. The use of a parametric model allows vocoders to operate at lower data rates (under 8 Kbps) than waveform coders; however, they require an accurate speech model to obtain good performance. Early vocoders such as the channel vocoder, homomorphic vocoder, and Linear Predictive Coding (LPC) vocoder all demonstrated the ability to produce intelligible speech at low to medium data rates. A good example is the 2,400-bps LPC-10 vocoder used as a U.S. government standard for secure (i.e., encrypted) telephony, but resulting in poor voice quality. Over the last decade, continued work has improved the performance of voice coders. The challenge with waveform coders has been to try to maintain adequate voice quality while lowering the bit rate. Generally this effort has focused on techniques commonly referred to as code-excited linear prediction (CELP), in which vector quantization is combined with adaptive linear prediction. This approach borrows several concepts from model-based coders in that an all-pole model is used to approximate the speech spectrum and a long-term predictor is used to represent the pitch (i.e., local periodicity) of the speech signal. However, in a CELP coder, an error signal or residual is computed to compensate for the shortcomings of the linear predictive model. This residual is quantized using vector quantization which typically requires a search for the best vector from a large codebook of candidates. While this approach has made headway, yielding good algorithms at 8 Kbps, the performance generally degrades rapidly at lower rates. In addition, the vector search employed in CELP coders (and its many variants) has significantly increased the complexity of these algorithms. In contrast to waveform coders, the challenge in model-based coders has been to improve the speech model to allow higher voice quality at low bit rates. One approach that has made a significant contribution is the
320
Transmission Systems Design Handbook for Wireless Networks
multiband excitation (MBE) speech model. In this model, which is fundamentally different from the linear-predictive methods found in traditional vocoders as well as CELP, speech is modeled with a fundamental frequency, a set of spectral coefficients, and a set of frequency-dependent voicing decisions. The inclusion of multiband voicing information plus new algorithms to analyze and synthesize speech has resulted in new MBE-based vocoders that can provide very high-quality speech at rates between 2 and 5 Kbps. Achievement of high-quality speech at such low bit rates is facilitated by the lack of any residual signal in the MBE-based approach. Instead, increased emphasis is placed on high-fidelity estimation and quantization of the model parameters so that voice quality can be maintained without the need for such an error signal. 7.6.2.2 Voice Compression in Wireless Networks
Development of advanced model-based coders has had a significant impact in a number of fields, including wireless communications, voice storage, and digital telephony, which require high quality combined with efficient bandwidth utilization. In a wireless system, this added coding efficiency is typically used to increase the number of users which can be supported across a fixed bandwidth. The rapid growth in personal communications has led to a critical need for increased capacity in voice communication networks. This need typically takes the form of a broad requirement to handle more simultaneous voice connections or messages in some constrained bandwidth or bit rate, without degrading power, weight, and range or voice quality. The typical solution to this requirement is convert to a digital communication network and to then employ voice compression to reduce the amount of digital data (i.e., bits) which must be handled. One of the applications where increased capacity has been required is in wireless networks. In this case the large increase in subscribers has overstretched the capabilities of the original analog cellular systems deployed in the United States and around the world. The combination of voice compression and digital communications is key component in most modern mobile voice communication systems and in many nonmobile systems as well and the basic reasons are cost and performance. In a digital system using voice compression, the number of users, which can be supported in some available bandwidth, is greatly increased relative to analog or uncompressed digital systems. The result is that the cost of developing and operating the system can be spread over many more users which typically leads to lower charges per user.
Transmission-Network Planning and Design
321
In addition, lowering the data rate in a wireless application generally leads to smaller, less expensive mobile equipment which uses less power and, with additional FEC, is more robust to bit errors found in a typical mobile environment. Transmission engineers have been using voice compression as a tool to increase capacity of the transmission (backhaul) networks for years. Voice circuits between cell sites and base station controllers, quite often carried over the leased T1/E1 facilities, can benefit from the addition of voice-compression hardware, which is usually built into the RBS and BSC utilizing the manufacturers proprietary compression scheme. In 3G wireless networks, IS-127 enhanced variable rate codec (EVRC) vocoder allows increased network capacity or cell site coverage without sacrificing voice quality. EVRC offers 13-Kbps voice quality at an 8-Kbps rate per IS-127. With EVRC, the vocoder rates are decreased during low speech activity to preserve bandwidth, and increased during high speech activity. The MOS for EVRC is 4.01, and as reference, the 64-Kbps PCM has a MOS of 4.27. EVRC has an average rate of 5.16-Kbps. Dynamic allocation of vocoders is supported in the hardware-software architecture of the BSC. The new selective mode vocoder (SMV) will also be supported in the future as the standards continue to mature. SMV promises EVRC-like voice quality with a capacity increase of ∼27%. Higher-capacity gains are also possible with SMV (up to ∼49%), but at the expense of slightly lower voice quality when compared with EVRC. It must be noted, though, that additional external voice compression can be used on the transmission network successfully only in case of channelized (individual) DS0s within the T1/E1 circuit that does not have compression already applied to it.
7.6.3
Signal Propagation Delay
Signal propagation delay, or latency, describes the delay of a transmission from the time it enters the network until the time it leaves the network. Low latency means short delays while high latency means long delays. Latency may occur in the handset or in the network. Latency that occurs in the handset or between the handset and the base station is called access latency. Latency that occurs from the base station through the network is called network latency. Low latency is essential for real-time transmissions. These include live voice conversations (but not voice mail messages, which are time insensitive) and live twoway video (but not entertainment video clips, which also are time insensitive).
322
Transmission Systems Design Handbook for Wireless Networks
Latency is not a phenomenon only of mobile networks. It is an outcome of all the networks, terminals, and devices through which transmissions may pass and the bottlenecks (and, therefore, delays) they may encounter. At home, users of broadband Internet connections experience latency as delays in downloading Web sites during peak traffic hours (often in early evenings and during inclement weather). Such delays are due to overloading bandwidth at the network periphery. More important are delays due to overloading bandwidth at the network core. A user in New York may download a Web site that is hosted in San Diego. Depending on traffic loading and transmissions costs at the moment, the download may travel from San Diego to Los Angeles to Denver to Houston to Chicago and finally to New York. It may use fiber networks owned by a number of different carriers. At each switching point, and in particular at the juncture of each network, it will encounter delay. Each of these delays increases the latency. This means that even if a mobile network is configured to provide low latency, the operator cannot guarantee low latency for end users who use their mobile devices to access other networks or who use their terminals in a noisy, and thereby latency inducing, RF environment. Also, differential delay between E1/T1 timeslots that may be routed over different paths (for fractional services) can cause protocol time-outs, retransmissions, and disruptions in video circuits as well as inhibit voice transmissions. Many types of transmission equipment such as multiplexers, packet assemblers, microwave radios, and DACSs add small amounts of delay due to their internal buffering while satellite links add significant delay to a signal. That is the reason satellite links are never used for backhaul in typical wireless networks. Testing signal delay is done during an out-of-service E1/T1 facilities bit error test. By establishing a loopback of the test signal at the desired point in the network, the test set can be used to measure the round-trip (RT) delay. It is also useful for fractional (N × 64 Kbps) E1/T1 services to measure the differential delay between the individual 64-Kbps timeslots, particularly when one application uses several contiguous or noncontiguous timeslots. For example, CDMA networks are very sensitive to delays and some vendors recommend that backhaul delay between the cell site and the BSC be as follows: Cell-site1 to BSC < 12 ms Cell-site 2 to BSC < 12 ms ∆ Delay = (Cell site 1 to BSC) − (Cell site 2 to BSC) < 10 ms
Transmission-Network Planning and Design
323
While designing wireless networks, transmission engineers have to be careful since CDMA networks, as well as other packetized data-transmission systems, do not tolerate noncontiguous timeslots. 7.6.4
Example of Optimized Network Design
A plan is in place to deploy multiple cell sites to support six rural service areas (RSAs) in city A. Each cell site is initially 8 to 12 channels over a T1 facility, served by a single MSC. All calls to and from landline use the switched network. To cover the RSAs without incurring significant long-distance phone charges, a multiple T1 private network is planned to be extended from the MSC to LEC COs in each of the six RSAspromising to result in a very costly overlay network. One solution is to use mini-DACS at the cell sites to groom and fill the cellular service with the switched-network backhaul access, making an overlay network unnecessary (Figure 7.9). The equipment consists of T1 multiplexers at each cell site with dual 4-wire channel cards and fractional T1 channel cards, as needed, to connect locally to the T1 and to the LEC CO. The MUX installed at the cell site has the flexibility of multiplexing channels onto a network T1 to the MSC location. These channels can be
T1 microwave
LEC central office Radio cabinet
Fractional T1
T1 Telco
MSC
T1
Te lc
Radio cabinet
LEC central office
oo
rF
ibe
r
Radio cabinet
LEC central office
Fractional T1
Long haul T1 network Local T1 access
Figure 7.9 Groom and fill network design example.
Fractional T1
324
Transmission Systems Design Handbook for Wireless Networks
allocated to voice and control data to support the cellular service, or can be allocated as a partially filled local-access T1 to the nearest CO to support back-haul phone calls locally. This partially filled T1 is often referred to as a fractional T1. By allocating the network channels to either the cellular voice and data or the fractional T1 channel allocation, the digital bandwidth can be shared between the various services. This configuration is called groom and fill. At the MSC location, the cellular and switched voice channels are connected using the mini-DACS. The mini-DACS usually can terminate up to 16 T1s in a small chassis, reducing space, power, and overall cost of the solution. The MSC has voice tandem trunks that connect on a T1. These trunks are connected to the mini-DACS together with the cellular T1s. The T1s to each of the cell sites also connect to the mini-DACS. The mini-DACS is then programmed to allow any DS0 channel originating from the MSC to connect to any DS0 of the long-haul T1 that is connected to the cell site. This grooms the network at the MSC, while the mini-DACS redistributes the DS0 channels at the cell site. An RSA cell site can, for example, be set up to support 12 DS0s for cellular and 12 DS0s for backhaul switched-voice access. At the MSC, the mini-DACS programmed on a DS0 basis fills the connecting long-haul T1 with channels connected to various T1 trunks of the MSC. The cellular base station trunks use 11 DS0s for cellular voice and one DS0 for cellular data control to the base station. The remaining 12 DS0 channels are crossconnected to an MSC voice trunk T1. At the cell site, the mini-DACS is programmed to connect the 11 cellular voice channels to six dual-channel fourwire VF channel cards, each connected to cellular transceivers. One channel is available for cellular data at 56/64 Kbps, which connects to a BSC system. The remaining 12 channels are programmed to provide T1 access to the local CO over a fractional T1 facility.
7.7 Transmission Network: Design Examples 7.7.1
Small PDH Microwave Transmission Network
Figure 7.10 shows an example of a small wireless network that has a total of four cell sites, out of which one is colocated with the switch (BSC). All the sites contained RBS, and they were connected to the switch by means of microwave radios. Connection to the local telephone company (PSTN) was achieved over a new short run of fiber-optic cable laid just for this purpose.
Transmission-Network Planning and Design 4-ft dish 6-ft dish 2 ODUs 2 coax. 1 waveguide 2 racks
Site 4
325
Site 3 7 GHz MW system 16xE1 1+1
15 GHz MW system 8xE1 1+1
4-ft dish 2 ODUs 2 coax. 1 racks
Switch site Site 2 4-ft dish 2 ODUs 2 coax. 1 racks
15 GHz MW system 8xE1 1+1
Site 1 Fiber-optic cable cca 800m
PSTN
Figure 7.10 PDH microwave network example.
If we assume a requirement of 1E1 per cell site (RBS), and calculate required capacity over every individual link (1E1 between Sites 3 and 4, 2E1 between Sites 3 and 1, and 1E1 between Sites 2 and 1), it is obvious that the system was designed for increased capacity and expansion in the future. The cost difference between installing 4E1 and 8E1 microwave systems is negligible in comparison with adding additional capacity or deploying a new microwave system later on. It makes sense to install an 8E1 microwave system from day one if there is even a remote chance that the capacity might be increased sometime in the future. Also, Site 3 is planned to be used as a hub site for any future expansion and since it is located on the top of the hill overlooking the city, it has the LOS with almost every potential new site location in the city. Thus, a 16E1 microwave system connecting Site 3 and the BSC location is not overkill. 7.7.2
Complex Transmission Network
Shown in Figure 7.11 is an example of the large network (over 300 cell sites) and as such, only a small part of the network is shown. This transmission
326
Transmission Systems Design Handbook for Wireless Networks
End link New microwave SDH ring
Spur hub
End link
Existing fiber-optic ring
SDH ring node
New microwave SDH ring
BSC MSC Fiber-optic POP Cell site MW SDH site (ideally colocated with the cell site) Microwave repeater
Figure 7.11 Transmission system design example for the complex network.
network is designed to meet service demands, but with the most economical routing in mind. It was designed and implemented in the one of the biggest cities in the world using multiple circuit providers, different technologies, and a combination of leased copper and fiber facilities and the new microwave systems. Survivability and reliability of the network are achieved by means of transmission loops (ring configuration) or a combination of star and ring configurations that provide the following:
Transmission-Network Planning and Design
327
• Increased system reliability due to less equipment; • Protection against catastrophic site failure; • No loss of E1/T1 traffic due to a single failureprovides alternate
routing of E1/T1 traffic automatically;
• Alternate routing of orderwire and alarm system automatically; • Easy and rapid maintenance, since individual E1/T1 failures are
switched and can be tested without taking the E1/T1 out of service;
• Isolated paths for maintenance and fault location purposes.
Predesigned protection, which refers to the fact that recovery from network failures is based on preplanned schemes and usually relies on resources dedicated to protection purposes (fibers, wavelengths, switches, etc.). A better and more efficient system is dynamic restoration, in which the discovery of spare capacity is done dynamically, as needed [10]. In SONET networks, intelligent DACS and controllers are used as the main components to realize dynamic restoration. Both protection schemes have their advantages and disadvantages and have to be chosen based on network requirements and topology. It is important to notice that even leased-lines (facilities) networks require careful planning and routing, the philosophy of the entire network (present requirements and future expansion) design cannot be left to the discretion of the telephone company providing the T1/E1 leased lines. This is especially important when multiple service providers and multiple equipment vendors are used for the project. Study of the existing transmission facilities and laying the RF network over it determine the number and type of the required and newly constructed facilities. As usual, in the network of this size, site acquisition is the most important bottleneck to take into consideration when trying to schedule the build out.
7.8 Overview of RNC Dimensioning in the 3G Wireless Network The purpose of this section is to describe briefly how to dimension the RNCs in a UMTS network for WCDMA (FDD mode). The intention here is to describe the dimensioning process without actually going into the details of the RNC hardware, which is vendor specific. This is a very generic, highlevel design that only covers some very basic assumptions.
328
7.8.1
Transmission Systems Design Handbook for Wireless Networks
Traffic Classes in the UTRAN Network
According to the UMTS Terrestrial Radio Access Network (UTRAN) system description, the ALL5 will be used for carrying signaling traffic and O&M traffic, and the AAL2 will be used to carry user traffic over Iub (packet and voice data traffic). The signaling traffic can be further divided into control signaling traffic and O&M signaling traffic. Since the voice data and the packet data may travel through different routes in the network (mainly at Iu and Iur interfaces), for dimensioning purposes we should further divide the AAL2 traffic into voice-data traffic and packet-data traffic. From the dimensioning point of view, therefore, the traffic in a UTRAN network can be classified as follows: • Packet-data traffic; • Voice-data traffic; • Control and signaling traffic; • O&M (network management) traffic.
We can further divide the packet data into two categories, one being real-time packet data and the other non-real-time packet data. Real-time traffic is usually generated by some kind of data application, such as a multimedia conference application, real audio player, and so on. Non-real-time packet-data traffic is usually generated by such applications as Web browsers, e-mail, and FTP. Burstyness is the main characteristic of this kind of traffic. Voice data will probably have peaks of voice-data traffic and packetdata traffic at different times. The UTRAN network nodes ATM traffic multiplexing ability can fully utilize this peak time difference (multiplexing gain). The amount of O&M traffic in a UTRAN network is dependent on how the O&M is implemented in the system, and what functionality it will cover. Usually, an O&M system (network management system) for a packet network just performs the following functions: • Set: initializes a specific management function in a specific point; • Get: gets some wanted statistical data to further analyze the perform-
ance of the specific network element. This element may be a piece of software, an interface or node, link, and so on.
• Trap: automatically reports (alarm) when critical events happen.
Transmission-Network Planning and Design
329
When dimensioning a network, traffic generated by the above O&M functions usually does not take too much bandwidth and can be performed by the system administrator at nonpeak hours. The network operator or administrator usually chooses nonpeak hours to do the corresponding operations to analyze the network performance. Counting might be a very important function for the O&M subsystem. The use of ATM in UTRAN networks provides a way to allow the O&M system to perform byte-based counting of user traffic. So, another function might be to count how many bytes, kilobytes, or megabytes the user has transferred over the network, so they can charge the user. All the control and signaling traffic will be transferred over a logical network. Permanent configured signaling VCs between all the RNCs and RBSs will form this logic network. From a transmission-network dimensioning point of view, the control and signaling of traffic volume are important. How the control and signaling traffic will be treated when they travel through a network is irrelevant for this discussion. Since the control and signaling traffic usually has higher priority than the user traffic, network nodes usually implement some kind of QoS mechanism to differentiate them. Since a system usually does similar housekeeping for every user, the volume of the signaling traffic from an RBS to an RNC should be roughly proportional to the number of subscribers in the corresponding cell. 7.8.2
Description of RNC Interfaces
The RNC has four interfaces toward different parts of the UMTS network. They are shown in Figure 7.12 and explained as follows: • The Iub interface is the interface between the RBS and RNC. The traf-
fic on this interface is user data, signaling, and network management.
• The Iur interface is the interface between different RNCs. The main
part of the traffic on Iur is soft handoff traffic. In addition, signaling and management are carried over the Iur interface.
• The Iu interface is the interconnection point between the RNS and
the CN. The main traffic on the Iu interface is user data, but signaling is carried over the Iu interface as well. The CN often is divided into a CS domain and a packet-switched (PS) domain. For this reason, Iu often is divided into a circuit-switched part (Iuc, Iu-CS) and a packet-switched part (Iup, Iu-PS).
330
Transmission Systems Design Handbook for Wireless Networks Core network
lu
lur RNC
Mur
lub Mub
RBS1
RBS2
Mobile terminals
NMS called Ranos
RBS3
Physical connection Logical connection
Figure 7.12 UTRAN network interfaces.
• Mur interface is the interface between RNCs and Radio Access Net-
work Operations System (RANOS) and handles the O&M traffic.
• Mub interface is the interface between base stations and RANOS and
handles the O&M traffic.
• Node B is a logical node responsible for radio transmission and
reception in one or more cells to and from the UE. It terminates the Iub interface toward the RNC.
• Soft handoff is the state where a UE is connected to two or more
cells in different base stations.
Transmission-Network Planning and Design
331
• The radio network subsystem (RNS) typically consists of an RNC
and the RBSs connected to this RNC.
• Serving RNS is a role an RNS can take with respect to a specific con-
nection between a UE and UTRAN. There is one serving RNS for each UE that has a connection to UTRAN. The RNS is in charge of the radio connection between a UE and the UTRAN. The serving RNS terminates the Iu for this UE.
• Drift RNS is a role an RNS can take with respect to a specific con-
nection between a UE and UTRAN. An RNS that supports the serving RNS with radio resources when the connection between the UTRAN, and the UE needs to use cell(s) controlled by this RNS is referred to as a drift RNS.
• Controlling RNC is a role an RNC can take with respect to a spe-
cific set of RBSs. There is only one controlling RNC for any RBS. The controlling RNC has the overall control of the logical resources of its RBSs.
• Inter-RNC soft handoff is soft handoff with more than one RNS
involved. An inter-RNC soft handoff can occur if a user travels from an area controlled by one RNC to an area controlled by another RNC (from one RNS to another).
The RNC locations are dependent on many factors, such as transmission cost, existing infrastructure, and location of CN nodes. The RNCs can be located or distributed near the base stations or centralized in a few major switch locations. Whether centralized or distributed, RNC topology must be determined from case to case. The reason for distributed RNC topology is to reduce the need for transmission resources. It is, however, not true that distributed RNC topology automatically leads to reduced need of transmission resources and the following should be considered: • The sum of Iur and Iu traffic is often higher than Iub traffic. Especially
in urban areas, the traffic is high and the number of base stations supported by each RNC low. A low number of base stations per RNC leads to a large proportion of the UE involved in RNC handoff.
• Statistical multiplexing can be achieved in the RNC, but there are
alternatives that might be more cost-effective. For example, the RBS
332
Transmission Systems Design Handbook for Wireless Networks
is ATM-based and it is possible to get efficient statistical multiplexing using the base stations as hubs. • The RNC capacity is used more efficiently with centralized RNC
resources than if the resources are distributed around many locations.
• The location of MGWs is of importance for the location of RNCs. • In the future, most of the traffic most likely will be IP-based. If some
of the functionality contained in the MGWs is moved into the RNCs, it will be possible to connect the RNC directly to ISPs. This would reduce the traffic between the CN and the UTRAN considerably.
• Many RNC locations might give extra costs for buildings and
personnel.
7.8.3
User Traffic Modeling
So far we have discussed cell capacity, service classes, and their possible impacts on transmission-network dimensioning. In real life, many more factors must be taken into consideration while designing the network. For example, in a CDMA system, users may conduct a voice call, and at the same time run several data applications (e.g., while checking e-mail, browsing the Web, talking with a friend). A common scenario is that when some users are using speech service, some other users are using data service. This indicates that the traffic volume generated by a user and all the users in a cell is a variable, which depends on the environment, the portion of users using different services, and so on. This means that a parameter describing the average user traffic behavior is required so that reasonable average traffic can be deduced from it to indicate the bandwidth requirement of a corresponding transmission link(s). This parameter(s) is called the user traffic model. A user traffic model is a collection of information that describes the average and common user-traffic behavior in a system. These parameters should be decided according to the operators suggestion, the service supported by the system, experimental statistics, former experience (which could be difficult to define in new 3G systems), and so on. They may include the following: • Vm: Mb per busy hour (or Erlang) voice-data traffic generated by
user;
• Pm: Mb per hour packet-data traffic generated by a user;
Transmission-Network Planning and Design
333
• Tvm: Connection holding time in minutes for voice-data connection; • Tpm: Connection holding time in minutes for packet-data connection; • Td: Data traffic peak hour(s); • Tv: Voice traffic peak hour(s); • S : Percentage of users using speech service; • C d: Percentage of users using circuit-data service; • P d: Percentage of users using packet-data service.
It is important to realize that these parameters might be given in another form. For a specific traffic class, it can be expressed with megabit (or Erlangs) per user per busy hour but also in sessions per day per user and average session length to describe the above Vm or Pm. Usually, the operator will provide a kind of user traffic model. Sometimes, the network planner will provide help to the wireless operator to define user traffic model. Different operator might provide their user traffic model in different forms. The user traffic model provided by the operator is likely to be a mixture of many other parameters, which may look very confusing. It is imporatant to understand that the user traffic model is the first step to calculating the per-site traffic throughput correctly and we should focus on the useful information, and skip the rest. With these parameters, we can roughly decide the average user traffic behavior in a system or a specific coverage area. For example, it is possible to estimate the user call frequency and mean traffic generated per user during busy hour, and so on. With these parameters, the number of cell sites, number of subscribers, a signaling and management traffic model, it is possible to roughly calculate the possible traffic load on a link between an RBS and an RNC. Usually, this is the starting point of dimensioning the transmission-network part of a UTRAN radio access network. The following formula can be used to convert voice traffic Erlang into bandwidth requirement on a specific ATM link: Bandwidth (Mbps) = (Erlang(A, blockingpob))(53/48/0.9) Here, 53/48 is an ATM overhead, and 0.9 is a 90% link load. The Erlang(A, blockingpob) represents the number of 64-Kbps links needed.
334
7.8.4
Transmission Systems Design Handbook for Wireless Networks
Traffic Calculation Guidelines
The calculation of the traffic on each RNC interface can be done based on the traffic model, RF design, and some basic assumptions. The main input for the dimensioning of RNCs is the number of base stations, user data, and RNC type. The number of base stations and user data should be given for the area served by each RNC. User data on the Iu Interface is one of the parameters used for RNC dimensioning. It is of importance to notice that only the traffic in one direction (uplink or downlink) is used for the dimensioning of the RNC. The user data on the uplink (UL) and downlink (DL) should be calculated and the highest figure will be used for the RNC dimensioning. U = U v + U cs + U ps where U = User data; Uv = User voice; Ucs = User circuit-switched data; Ups = User packet-switched data. The user data is a combination of different services. The services can be either circuit data or packet data and have a number of different bit rates. The amount of voice normally is expressed in Erlangs. In order to calculate the amount of voice expressed in bytes per second, knowledge of the voice codec and the voice activity is needed. Voice in WCDMA uses a codec called adaptive multirate (AMR). AMR supports eight different source codec bit rates ranging from 4.75 to 12.20 Kbps. In addition, a codec mode for periods of silence is available. During periods of silence, one five-octet-long silence descriptor (SID) is sent every 160 ms. A commonly used model for voice activity is that each subscriber talks 50% of the time. During periods of talk, 5% of the time is considered as silence. This gives an actual time of talk of 45% and of silence 55%. An investigation shows that the voice activity in reality is around 60%. The reason for a higher figure is the influence of background noise from traffic, other people, and so on. For this calculation we can assume 50% voice activity. Table 7.3 shows the average bit rate for different AMR modes. For the calculations, 50% voice activity is used. The SID descriptor is considered as user data and, thus, included in the calculations.
Transmission-Network Planning and Design
335
Table 7.3 Average Voice Bit Rate AMR mode (Kbps)
Average User Bit Rate (Kbps)
04.75
2.50
05.15
2.70
05.90
3.07
06.70
3.47
07.40
3.82
07.95
4.10
10.20
5.22
12.20
6.22
Calculation of user circuit data is as follows: U cs, c = ∑U cs,c where Ucs,c = User data for circuit-switched data with bit rate c; c = All used circuit-data bit rates. Calculation of user packet data is as follows: U ps = (1 + Ω r )∑U ps, p where Ωr = Overhead caused by transmission (10%); Ups,p = User data for packet-switched data with bit rate p; p = All used packet-data bit rates. The retransmission rate of 10% is commonly used, although due to customer demand or other factors, different values might be used. 7.8.4.1 Calculation of Iu Traffic Load
The traffic load on the Iu interface is made up of user data, signaling, and frame overhead. All traffic is transported using ATM, and the user circuit
336
Transmission Systems Design Handbook for Wireless Networks
data and voice use AAL2, while user packet data uses ATM AAL5. The overhead is different for different services and, therefore, can make dimensioning less precise. The traffic load on the Iu interface is as follows:
(
T u = 1 + Ω sign
)(T
v
+ T cs + T ps
)(1 / l)
where Tu = Total Iu traffic load; Tv = Iu voice-traffic load; Tcs = Iu circuit-switched-data traffic load; Tps = Iu packet-switched-data traffic load; Ωsign = 10% (signaling). In order to cope with traffic variations, a load factor ( l) has been introduced. The load factor is dependent on the traffic load and low traffic load leads to larger traffic variations. The traffic variations on the Iu interface are smaller than the traffic variations on the Iub interface. Usually a load factor of 80% is used. Iu Voice Traffic Load
Table 7.4 shows the average bit rate for different AMR modes. For the calculations, 50% voice activity is used. The SID descriptor is considered as user data and, thus, included in the calculations. The multiplexing efficiency is considered to be 100%. Iu Circuit-Switched Data Traffic Load
T cs = (1 + Ω fr )U cs where Ucs = User data for circuit-data service; Ωfr = 21% for ATM, AAL2, UP, GTP frame overhead (fixed rate 64Kbps circuit-switched data). The frame overhead for fixed-rate 64 Kbps and for variable rate 57.6 Kbps is 21% and 28%, respectively. Higher bit rates are not yet standardized, but it can be assumed that the bit rate for fixed-rate 64 Kbps can be also used for higher bit rates. Again, 100% multiplexing efficiency is assumed.
Transmission-Network Planning and Design
337
Table 7.4 Average Iu Voice Bit Rate Including Frame Overhead
AMR Mode (Kbps)
Average User Bit Rate VA = 50% (Kbps)
14.75
4.62
15.15
5.07
15.90
5.30
16.70
5.75
17.40
6.20
17.95
6.65
10.20
8.01
12.20
9.13
Iu Packet-Switched Data Traffic Load
The frame overhead on the Iu interface is heavily dependent on the size of the user data packets. T ps = (1 + Ω fr )U ps where Ups = User data for packet-data service; Ωfr = 72% for ATM, AAL5, UP, GTP frame overhead. The frame overhead is only valid if the average length of packets is 125 octets. Larger packet lengths give a smaller percentage overhead and vice versa. The user data protocols are assumed to be TCP/IPv4 or protocols of similar length. Again, 100% multiplexing efficiency is assumed. 7.8.4.2 Calculation of Iub Traffic Load
The traffic load on the Iub interface is made up of user data, signaling, O&M, and frame overhead. All traffic is transported using ATM with user data using AAL2. Soft handoff causes a large part of the overhead. Making a correct calculation of the user data traffic from one or a few base stations is currently a problem. The average traffic per RBS can be used for calculation of the traffic on the RNC level, as many RBSs are connected to one RNC. The
338
Transmission Systems Design Handbook for Wireless Networks
assumption is that the connected RBSs will have peak traffic at different times and the combined busy hour (BH) traffic probably will be near the average BH traffic due to statistical multiplexing. Iub Total Traffic Load
(
)(
T = 1 + Ω sign + Ω O& M T v + T cs + T ps
)(1 / l)
where T = Total Iub traffic load; Tv = Iub voice traffic load; Tcs = Iub circuit-switched-data traffic load; Tps = Iub packet-switched-data traffic load; Ωsign = 10% for signaling; ΩO&M = 2% for O&M. In order to cope with traffic variations, a load factor has been introduced. The load factor ( l) is dependent on the traffic load. Low traffic load leads to larger traffic variations, and normal traffic load is at least 1 Mbps per RBS. For most applications, a load factor of 80% is used. Iub Voice Traffic Load
T v = (1 + Ω SH )U v where Uv = User voice including frame overhead (see Table 7.5); ΩSH = 30% for soft handoff. Iub Circuit-Switched-Data Traffic Load
T cs = (1 + Ω SH )(1 + Ω fr )U cs where ΩSH = 30% for soft handoff factor; Ωfr = 23% for ATM, AAL2, UP frame overhead;
Transmission-Network Planning and Design
339
Table 7.5 Average Iub and Iur Voice User Bit Rate
AMR Mode (Kbps)
Average User Bit Rate VA = 50 %, Uplink (Kbps)
Average User Bit Rate VA = 50%, Downlink (Kbps)
14.75
15.61
5.10
15.15
16.06
5.55
15.90
16.29
5.78
16.70
16.74
6.23
17.40
17.19
6.68
17.95
17.58
7.13
10.20
19.22
8.71
12.20
10.35
9.84
Ucs = Circuit-switched data on Iub The frame overhead is calculated for fixed-rate 64 Kbps and variablerate 57.6 Kbps circuit-switched data. It is assumed that all CS RABs have a similar-size frame overhead. Again, 100% multiplexing efficiency is assumed. Iub Packet-Switched-Data Traffic Load
T ps = (1 + Ω SH )(1 + Ω fr )U ps where ΩSH = 30% for soft handoff; Ωfr = 24% for ATM, AAL2, PDCP, UP frame overhead; Ups = Packet-switched user data on Iub. The frame overhead varies with the packet size. The frame overhead is approximately 24% if header compression is used and the average length of the packets is 125 octets. If the packet length is larger, the overhead will be higher due to less impact from header compression. The user data protocols are assumed to be TCP/IPv4 or protocols of similar length. Multiplexing efficiency of 100% is assumed.
340
Transmission Systems Design Handbook for Wireless Networks
7.8.4.3 Calculation of Iur Traffic Load
On the Iur interface, user data is transported for users that are in soft handoff with more than one RNC involved. The user data is of the same type as on the Iub interface with the same type of overhead. The traffic load on Iur interface is made up of user data, signaling, O&M, and frame overhead. All traffic is transported using ATM and all user data use the AAL2 protocol. The amount of the Iur traffic load depends on how large the probability is that an UE is in soft handoff with two or more RNSs involved. If an RNS does not have any borders towards another RNS, there is no traffic on the Iur interface. For the case in which the RNS has a homogeneous cell pattern and is surrounded by other RNSs on all borders, an RNC handoff factor has been introduced. Calculation of Traffic Load on the Iur Interface
(
)(
T = 1 + Ω sign + Ω O& M T v + T cs + T ps
)(1 / l)
where T = Total traffic load on the Iur interface; Ωsign = 10% for signaling; ΩO&M = 2% for O&M; Tv = Iur voice traffic load; Tcs = Iur circuit-switched-data traffic load; Tps = Iur packet-switched-data traffic load. In order to cope with traffic variations, a load factor (l) has been introduced. The load factor is dependent on the traffic load, and for normal traffic, a load factor of 80% is used. Low traffic load leads to larger traffic variations. The Iur interface normally has smaller traffic variations than Iub interface. Iur Voice Traffic Load
T v = aΩ SHU v where a = Inter-RNC soft handoff factor;
ΩSH = 30% for soft handoff;
Transmission-Network Planning and Design
341
Uv = User voice including frame overhead (see Table 7.5). Iur Circuit-Switched-Data Traffic Load
T cs = aΩ SH (1 + Ω fr )U cs where a = Inter-RNC soft handoff factor;
ΩSH = 30% soft handoff factor; Ωfr = 23% for ATM, AAL2, UP frame overhead; Ucs = Circuit-switched user data on Iub. The frame overhead is calculated for fixed-rate 64-Kbps and variablerate 57.6-Kbps circuit-switched data. Again, 100% multiplexing efficiency is considered. Iur Packet-Switched-Data Traffic Load
T ps = aΩ SH (1 + Ω fr )U cs where a = Inter-RNC soft handoff factor;
ΩSH = 30% for soft handoff; Ωfr = 24% for ATM, AAL2, PDCP, UP frame overhead; Ups = Packet-switched user data on Iub. The frame overhead varies with the packet size. The frame overhead is approximately 24% if header compression is used and the average length of packets is 125 octets. If the packet length is larger, the overhead will be higher due to less impact from header compression. The user data protocols are assumed to be TCP/IPv4 or protocols of similar length. Multiplexing efficiency of 100% is assumed. The simplified inter-RNC soft handoff factor is calculated as follows: a = 2.5 / n RBS where nRBS = Number of RBSs connected to current RNC.
342
7.8.5
Transmission Systems Design Handbook for Wireless Networks
Example of Traffic Calculations
The user traffic model shown in Table 7.6 is very straightforward. It gives the Erlangs and BH data volume, so it is not very difficult to calculate the traffic volume generated per subscriber during busy hour. Once the network planner fully understands the user traffic model, it is easy to calculate the average traffic volume generated on a per-subscriber basis during busy hour: Voice traffic per subscriber: 8 × 0.5 × 15 = 60 bps Circuit-data traffic per subscriber: 64 × 1 × 4 = 256 bps Packet-data traffic per subscriber: 500 bps So, the average pure user data rate generated per subscriber during BH is 60 + 256 + 500 = 816 bps Note that the data rate is the pure user data rate; this means that it does not include any protocol overhead. Based on the raw data rate, in the next step, we can calculate data throughput per site. To calculate the traffic throughput per site, we need to know the number of subscribers per site. Table 7.6 Example of User Traffic Model Circuit Services Call type
Voice
Data
Average data rate (Kbps)
8
64
Activity factor
0.5
1
BH traffic per user
15 mErlang
4 mErlang
Packet Services Uplink
Downlink
Average data volume
20 kB/day/sub
200 kB/day/sub
BH data volume
50 kB/day/sub
500 kB/day/sub
Average packet size
128 bytes/packet
128 bytes/packet
Average number of connections
2 times/day
2 times/day
Number of handoffs
2 times/connection
2 times/connection
Transmission-Network Planning and Design
343
This usually can be calculated from the total number of subscribers the system is going to support and the number of the RBSs (provided by the radio planner). If the Ttp is traffic throughput from a site, then we can calculate T tp = T + Ω + Ω O&M where T = Pure user traffic in a site multiplied by the number of users in a site; Ω = Protocol overhead (assume an added 20%); ΩO&M = Operation and maintenance of traffic for the site (assume 128 kbps). We can add 20% for the ATM protocol overhead. The O&M traffic also remains as an open issue but a good approximation is to add 128-Kbps O&M traffic to each site. With a given user traffic model, one usually can calculate the mean (average) traffic generated by the number of users in a given area (cell or site). Usually, this is the first step in dimensioning the radio access network. However, this mean traffic volume sometimes does not reflect the whole picture of the real traffic. In TDMA networks, the capacity of a radio timeslot is fixed. This means that the traffic generated by one user is limited by this capacity. However, in a UTRAN (CDMA-based) network, since different services correspond to different data rates, one user may at the same time run several applications. The traffic burstyness characteristics require peak traffic to be taken into consideration when dimensioning the transmission network. One interesting scenario that happens frequently is that the mean data rate (site traffic throughput) calculated according to some user traffic model is lower than the data rate required by a single service. For example, the network might be designed to support 384-Kbps data service, but according to the user traffic model and the number of users covered by one site, when one calculates the mean traffic volume generated by the site, the result is lower than 384 Kbps. If we directly use the calculated mean traffic volume to dimension the transport link, in this case, the resulting transport link may not even be able to support a single user running a single 384-Kbps application with required quality. To solve this problem, peak traffic might be considered while dimensioning transmission links.
344
Transmission Systems Design Handbook for Wireless Networks
7.9 Alternative Solutions in Transmission Networks 7.9.1
Considering Dark Fiber and Dark Copper
Sometimes wireless operators, instead of leasing E1/T1 circuits or fiber-optic facilities, can lease copper or fiber lines without terminal equipment (dark fiber or copper), then design the system and provide their own terminal equipment. This solution for providing transmission facilities is obviously more cost efficient, but requires a lot more time and resources on the wireless operators side to do the engineering, equipment procurement, and installation, as well as operation and maintenance. Electrical utility companies, cable companies, and railways are good candidates from which to lease dark copper or fiber. Sometimes offering services in return, wireless operators could make a business case and build a partnership that would be of interest to all parties involved. Electrical utilities design their networks with high reliability and survivability as well as very low network delay. Their telecommunications networks are required to carry not only data traffic for LAN connectivity and administrative data needs but also teleprotection traffic. Teleprotection traffic is a reason that network delays must be very low for correct operation of protective relaying equipment. 7.9.2
Partnership with Utilities
Most data today travels over a network that was built for voice traffica process that is unreliable and inefficient because the loss of a single bit stream of data will adversely affect the entire communication. System providers are offering several different devices to transmit and receive data, but regardless of how data is transmitted, the wireless infrastructure will have to be improved to ensure consistently reliable communications. For ISPs and other data-centric businesses, it is critical that wireless providers work quickly and effectively to broaden their coverage areas. They also must provide an infrastructure that is more suited to wireless data transmission in order to get to market more quickly and deliver their own services and applications more reliably. Required improvements include upgrading existing cell sites with new electronics and placement of new towers and antennas to ensure coverage across the nation. However, these activities often require wireless providers to solve zoning regulation issues, civic resistance, and technical obstacles. By creating strategic partnerships with local utilities, which have a core competency of infrastructure management, communications providers can overcome these
Transmission-Network Planning and Design
345
problems. While the wireless communications provider maintains the role of network manager and supervisor, the utility leverages its existing relationships, equipment, and expertise to construct the infrastructure for delivery of wireless communications. Whenever possible, wireless providers are using existing buildings and communications infrastructure (colocation) to build out new networks. Adding antennas and upgraded electronics to towers already in place is both cost effective and more easily accepted by the public. However, existing infrastructure is not always available to cover new areas. There are a number of ways wireless providers can work with local electric utilities to place antennas and associated electronics on electric transmission and distribution facilities and, thereby, expand the network, improve coverage, and increase reliability. 7.9.2.1 Electric Transmission Towers
Large electric transmission towers provide a corridor between generation stations and substations. They tend to be in remote, out-of-the-way locations and run through less expensive territories. Typical voltages on these transmission towers range from 138,000 to 500,000V. Their location and 150-footplus height allow for placement of larger antenna arrays and excellent antenna elevation. These towers are ideal as antenna locations for propagating wireless signals over large areas, but precise engineering and extreme care must be used when placing RF or microwave antennas near transmission conductors to avoid the dangers of electric arcing. From a strictly microwave perspective, these towers as well as power lines are not an obstacle to install a microwave system even if they obstruct the LOS of the microwave system. Signal loss due to the partially obstructed Fresnels zone would be only 1 to 2 dB, depending of the type of the electrictransmission tower and its construction. 7.9.2.2 Distribution Poles
Colocating antennas on wooden distribution poles enables the wireless provider to enter neighborhoods that otherwise would not be served. The lower pole heightstypically 50 to 90 ftmean that signal propagation will be reduced, requiring installation of more antennas, new poles, and other structures. When pole locations and physical constraints prevent the use of existing towers and distribution poles, there may be no alternative but to erect a new tower. Utilities own property in their coverage areassubstations, offices, and maintenance and storage sites. In addition, they often hold right-of-way easements that provide access to areas otherwise unattainable. Developing a relationship with the local electric utility could offer the service
346
Transmission Systems Design Handbook for Wireless Networks
provider hundreds of potential sites. Wireless service providers attempting to gain similar access must negotiate individual leases with individual ownersa process that is time consuming, frustrating, and expensive. Another option is negotiating a master lease with a company that owns multiple sites in the neighborhood. School districts typically resist antennas on their properties, but local mass-transit (rail and bus) companies, city governments, libraries, and churches are often open to the idea of leasing space for a tower or other antenna mounting structures. 7.9.2.3 Technical Issues
Some utilities have resisted the idea of adding wireless infrastructure development to their workload because of the potential effects on the electric transmission and distribution (T&D) system. Installing antennas on electric transmission and distribution equipment affects wind resistance and torsion stresses on the towers. The design of these structures must account for conditions such as conductor weight, wind loading and ice loading, bracket design, attachment methods, and dead-weight loads. In some cases, utilities simply may not be willing to take these risks with their infrastructures. Research has shown, however, that proper design and clearances between wires and antennas produce no ill effects on either the communications network or the T&D system. It is critical that electrical and physical constraints on individual installations be carefully considered. Electric arc effects can allow flashover between the grounded antenna and T&D wires, so antennas must be kept 20 ft from 13,000V distribution lines and a minimum of 28 ft from 500,000V transmission lines to provide appropriate clearances and safe working distances. Distances must also be maintained between antennas that are comounted on a tower. Analyses must carefully review initial weight, wind loading, and torsion effects. Weld methods, bracket types, and even the types and sizes of bolts must be appropriate for the installation. Perhaps the most critical element is the towers foundation. Depending on the utilitys philosophy, safety factors of 10% to 50% are factored in to prevent tower failure under all but the most unusual circumstances. If the existing safety margin is inadequate, the carrier and the utility often negotiate the cost of replacing the foundation and associated structure. Installation and maintenance of wireless antenna systems are more painstaking than T&D line work because of the precision required to position and orient the antenna. Additional training will be required, but utility line workers are generally receptive because they know that broadening their skill set increases their employee value. Economies of scale have been realized in
Transmission-Network Planning and Design
347
wireless maintenance activities because they are integrated into maintenance practices for the electrical infrastructure. It is important to understand that electric transmission towers are dangerous locations prone to a phenomenon called ground potential rise (GPR). The communications cables (E1/T1 circuits) are susceptible to damage from lightning surges, since they can develop high shield-to-pair voltages even with low lightning currents on the shield; these voltages could cause lengthy and potentially catastrophic outages. GPR usually occurs during a power fault when the fault current returns to the power neutral source through the earth. Communications cables are considered exposed to GPR when the possibility exists that the local ground (at the cell site) differs from remote ground by 300V or more. Optical isolators (optocouplers) are generally used to treat each circuit (including T1/E1 circuits) going into the power station to protect the circuits and facility, as well as personnel, from electric hazard associated with GPR. Design of any communications circuits into the power location has to be in accordance with ANSI/IEEE 367, Recommended Practice for Determining the Electric Power Station Ground Potential Rise and Induced Voltage from a Power Fault, and ANSI/IEEE 487, Guide for Protection of Wireline Communications Facilities Serving Electric Power Stations. The only other alternative would be a letter, on letterhead and appropriately signed, from the power utility owner, stating that at no time will the GPR at the site or sites ever exceed 1,000V peak-asymmetrical, as calculated per ANSI/IEEE 367. The power utility owner must be aware that issuance of the letter constitutes an assumption of liability for injury or damage brought about by electrical fault conditions. Installation of the microwave systems using electrical utility poles and towers has to be carefully examined, since they may not fulfill requirements for the twist and sway of the microwave antenna mounting structure, especially on the higher microwave frequencies. One major benefit of partnership with the local utility is the utilitys experience and established credibility with the community and town zoning boards. Enhancing communications services that will benefit local residents and businesses, while diminishing the need for new roads to reach new towers and minimizing installation of new structures, is a win-win situation. The communications network improvements may be invisible to the neighborhood but the benefits are not. Wireless providers may prefer to erect new towers and install antennas on targeted buildings as they grow their networks. When entering a new territory, it is much faster and easier to conduct a single negotiation by partnering with one local utility than hundreds of negotiations to get the same area coverage. When wireless providers seek to
348
Transmission Systems Design Handbook for Wireless Networks
expand their network infrastructure and broaden coverage, partnering with utilities speeds deployment, reduces costs, and increases efficiency. Utility expertise in zoning, construction, and maintenance, as well as access to existing antenna-ready structures, can be put to good use without expanding the carriers payroll. Using existing structures to locate antennas allows faster deployment of equipment and improves maintenance cost-efficiency while the wireless carrier enjoys fast, cost-effective deployment of a new network. The ISP or wireless operator receives a higher QoS, more reliable coverage and transmission, and faster time to market. 7.9.2.4 Power Line Carrier and Optical Power Ground Wire
Voice and limited data for supervisory control and data acquisition (SCADA) over the power lines [power line carrier (PLC)] has been used by electrical utilities for years. SCADA systems are used extensively by power, water, gas, and other utility companies to monitor and manage distribution facilities. Very narrowband in its nature, it was not suitable for any serious application outside transmission of the monitoring and control signals. Over the last few years and with the introduction of digital PLCs, its bandwidth is approaching 1E/T1. Aerial fiber-optic cables can be suspended on the telephone poles just like any other copper cable, wrapped around existing power lines, or be a part of the ground wire on top of the high-voltage transmission tower [optical power ground wire (OPGW)]. OPGW is described in more detail in the chapter dealing with fiber-optic equipment. Utilities may be willing to lease out some of the capacity of their fiber-optic system, and usually every tower has a junction box with access to the fiber-optic cable. 7.9.3
Optical Laser Communications
The primary application for high-capacity laser communication system, sometimes called laser transmission systems (LTS), is extending fiber-optic backbones to the customers premises. The products optical fiber interfaces must be fully compatible with the existing carriers SONET/SDH backbones, and the system can meet both the voice and data bandwidth needs of the business customer. Typically, in North America only 10% to 15% of the carriers customers and only 3% to 5% of commercial buildings are on fiber with highbandwidth capabilities. These numbers are even lower in Europe and other parts of the world. In general, for a building, a minimum of 10 to 15 T1s
Transmission-Network Planning and Design
349
(DS3 break-even point) of existing revenue was required for a carrier to even consider the long and expensive process of deploying fiber. Laser (wireless) optical networking is an innovative technology that improves upon the concept of free-space optics. A free-space optical link consists of two optical transceivers accurately aligned to each other with a clear line of sight. Typically, the optical transceivers are mounted on building rooftops and use the light near the infrared region of the spectrum. The optical transceiver consists of a laser transmitter and a detector to provide full duplex capability. Free-space optics enables very fast deployments of broadband access services to buildings; the time-consuming and expensive process of getting permits and trenching city roads is completely avoided. This approach is, however, problematic in that atmospheric conditions have a significant impact on the optical link performance. Availability of a free-space optics link is generally determined by the link length and fog patterns in a specific location. The laser optical network uses mesh configuration of short, redundant links between optical transceivers. The mesh network functionality is provided by compact nodes mounted on the rooftops of various buildings and connected by LOS. Beam attenuation, depending on bandwidth configuration, can cover distances up to 2.5 mi (4 km) with transmission rates up to 622 Mbps (OC-12 equivalent). Infrared-based laser optical transmission systems are not subject to any licensing requirements in the United States and probably the rest of the world. Because this transmission system is immune to electro-magnetic interference (EMI) and does not cause RF interference, it is also virtually interceptionproof, exceeding methods employed in microwave and copper cable products. Safety issues are regulated by international standards IEC 825-1 and EN-60825-1 regulating health issues and especially eye safety of optical sources used in laser communications systems. This classification is based on concerns for the potential risk to human health (such as eye damage) caused by high-power optical transmission systems. LTS could become much more interesting in 3G wireless systems, where huge capacity links to every cell will be required. Heavily loaded sites will be in dense urban and urban areas where the distance between cell sites is short and can range from a few hundred meters to a few kilometers; they are therefore good candidates for alternative backhaul media solutions.
350
Transmission Systems Design Handbook for Wireless Networks
References [1] ITU-T G.174,Transmission Performance Objectives for Terrestrial Digital Wireless Systems Using Portable Terminals to Access the PSTN, June 1994. [2] Lee, W. C. Y., Mobile Cellular Telecommunications: Analog and Digital Systems, New York: McGraw-Hill, 1995. [3] ITU-T G.803, Architecture of Transport Networks Based on the Synchronous Digital Hierarchy (SDH), March 2000. [4] ITU-T G.176, Planning Guidelines for the Integration of ATM Technology into Networks Supporting Voiceband Services, April 1997. [5] Sherman, K., Data Communications: A Users Guide, Third Edition, New York: Prentice Hall, 1990. [6] Flanagan, W. A., The Guide to T1 Networking, Telecom Library, Inc., New York, 1990. [7] GR-2830-CORE, Primary Reference Sources: Generic Criteria, Bellcore, Issue 2, December 1995. [8] Larus Corporation, Digital Network Timing and Synchronization, San Jose, CA, 1997. [9] Dixon, R. C., Spread Spectrum Systems with Commercial Applications, New York: Wiley, 1994. [10] Zhou, D., Survivability in Optical Networks, IEEE Network, November/December 2000, pp. 1623.
8 Transmission Equipment 8.1 Digital Microwave Radio 8.1.1
PDH and SDH Microwave Radios
The PDH microwave radio provides a transmission medium for digital traffic of standard capacities typically ranging from 1.544 Mbps (1T1) to 45 Mbps (1DS3) in North America and from 2.048 Mbps (1E1) to 34 Mbps (16E1) based on ITU standards. Sometimes high-capacity MW backbones in North America are built for 3DS3 and require at least 28 MHz of bandwidth. These radio links may be established between any two points with the line of sight and depending on the frequency, geographical region, and rain statistics the typical link distance can be up to 25 mi (40 km). For the longer microwave link hops, additional measures have to be taken to ensure required reliability of the system (e.g., space or frequency diversity). In todays wireless networks, these PDH microwave systems are used for the low- to medium-capacity links (e.g., RBS-BSC backhaul connectivity) in wireless networks. When it comes to high-capacity transport of information, network planners have two key technologies at their disposalfiber-optic systems and digital microwave radio. The new generation of digital microwave systems, based on SONET/SDH, is able to meet the requirements for the highcapacity backbone transmission systems. SDH/SONET radios provide economical solution when existing infrastructure (towers and shelters) can be reused, and when rights of way or adverse terrain make fiber deployment very costly or time consuming. In existing 2G wireless networks, these high351
352
Transmission Systems Design Handbook for Wireless Networks
capacity microwave systems are only used for the connection of the hub sites to the fiber-optic ring or directly to the BSC. Also, they can be used for the PSTN connection in third-world countries where existing telecommunications infrastructure is limited. In the future 3G wireless systems it is expected that more cell sites will require this kind of connectivity due to the increased backhaul capacity requirements. To maximize the benefits of the SONET/SDH microwave radio, the radio must be capable of complementing a synchronous fiber-optic network. This means that in order for the microwave radio to integrate with fiberoptic network elements, its design must address a number of parameters, including capacity and growth, network management, maintaining pace with SDH/SONET standards evolution, interface, and performance. Providing microwave radio with the optical interface will allow the microwave network to integrate with the fiber-optic network without the use of multiplexing equipment (unless drop or insert of the traffic is required). Todays SONET/SDH radio technology is capable of delivering bandwidth-efficient 8 bits/s/Hz of bandwidth. For example, 512-state QAM technique can pack two STM-1 streams into a single 40-MHz channel using a single carrier [1]. By adding channels in a 1:N configuration, system capacities of up to 14 protected STM-1s can be achieved within one frequency band (e.g., in the upper 6-GHz band, eight bidirectional channels are available). By deploying a dual-band configuration (such as lower 4- and 5-GHz bands), system capacities of STM-16 and greater are achievable. Although theoretically possible, this can be done only if the spectrum governing bodies allow more than one channel to be used by the same user. SDH/SONET microwave radios have to meet very stringent error performance objectives defined in ITU and ANSI standards. Sophisticated and powerful countermeasures, including FER and ATPC, are used to combat propagation anomalies.
8.1.2
Standard Microwave Radio Configuration
Standard microwave radio configuration consists of the entire microwave and digital modem part being placed indoors, microwave antenna mounted outside on the tower, and a waveguide connecting the transceiver of the radio with the antenna. This solution is shown in Figure 8.1 and it is acceptable for the lower frequencies but quickly becomes unacceptable as frequency increases. This is due to the high losses in the transmission lines (coaxial or waveguide) which become unacceptably high at higher frequencies. Most
Transmission Equipment
353
PCS antennas
Microwave antenna
Equipment shelter
A standard microwave system includes an antenna, MW radio waveguide, installation and mounting accessories, and pressurization
Figure 8.1 Standard-configuration microwave system.
commonly used waveguides today for terrestrial microwave point-to-point systems are Helliax elliptical waveguides. This configuration is still used today for lower-frequency bands below 10 GHz and for high-capacity (backbone) microwave systems. Whenever possible, split configuration is replacing the standard configuration and this is particularly true for the cell-site microwave systems in wireless networks. 8.1.3
Split Microwave Radio Configuration
8.1.3.1 General Description
To reduce losses between the transceiver and antenna, the outdoor unit (ODU) containing all the RF modules can be mounted near the antenna
354
Transmission Systems Design Handbook for Wireless Networks
(Figure 8.2). The ODU is connected to the indoor unit (IDU), containing baseband circuitry, modulator, and demodulator, by means of one single coaxial intermediate frequency (IF) cable. The distance between indoor and outdoor equipment can usually be up to 300m (1,000 ft). SDH microwave radio will be described as an example of the highcapacity, split-configuration microwave system. Baseband interfaces for the equipment are STM-1 electrical or optical, 2-Mbps wayside channel and 64Kbps insert channels. The basic block diagram for a digital microwave system (split configuration), including the main blocks, is shown in Figure 8.3. The block diagram includes marked interface points that serve as reference points for several technical parameters used in the text. Some main technical characteristics of the European version of the split-configuration STM-1 microwave radio in 18- and 23-GHz bands are presented in the following sections. 8.1.3.2 Equipment Main Characteristics
Table 8.1 shows two very common frequency bands, traditionally used in wireless networks18 and 23 GHz. Channeling plans are according to ITU-R recommendations. MW Antenna Outdoor module
Indoor module
Single coaxial cable
Figure 8.2 Split-configuration MW radio.
Transmission Equipment IDU OUT
IN
355
ODU
Modulator and baseband
Transmitter
RF Rx filter
Branching network
Feeder
Modulator and baseband
Receiver
RF Rx filter
Branching network
Feeder
Figure 8.3 Principal block diagram of the MW radio.
Nonprotected systems (1+0) consist of one IDU and one ODU interconnected with a single coaxial cable. In the case of failure of any of the electrical or mechanical components, the entire microwave hop would fail. In hot standby (protected) configuration (1+1), the IDU, the transceiver unit, and the coaxial cable between IDU and ODU are duplicated. The two transceivers share the same branching unit. A switch at radio frequency level, included in the branching unit, allows for switching between the two transmitters. A splitter and switching unit placed between the two IDUs is added. The unit splits and switches between the STM-1 interfaces. For each terminal one antenna is used. Also, there is a version of the protected system with two ODUs and only one antenna. These radios were also designed with ease of installation in mind. The equipment is designed to enable very easy and quick installation within a few hours. It is designed for split-mount installation but can also be all-indoor mounted. One coaxial cable between IDU and ODU is used for 1+0 systems and two cables for hot standby systems (1+1). Installation of the splitTable 8.1 Frequency Bands, 18 and 23 GHz Frequency Frequency Band (GHz) Range (GHz)
ITU-R Adjacent Channel Tx-Rx Duplex Recommendation Spacing (MHz) Spacing (MHz)
18
17.719.7
ITU-R F.595-3
55
1,010
23
21.223.6
ITU-R F.637-2
56
1,232
356
Transmission Systems Design Handbook for Wireless Networks
terminal configuration, assuming that all of the infrastructure is available (tower, shelter, racks, pipe-mounts), should not take more than one day per hop and two people to do the work. Indoor installation consists of one small module for nonprotected and three modules for protected systems (two modems plus the switching module). Figure 8.4 illustrates the usual telecommunications rack, layout with the fuse or breaker panel on the top, space for the small rectifier and battery backup at the bottom of the rack, and microwave radio and the DSX panel mounted in the middle of the rack.
Rack units 44 (RUs; 1RU=1.75 in) 40
Dual fuse panel 1+10 IDU (nonprotected)
35
Channel A Channel B 1+1 IDU
30
Modem A Prot. unit Modem B
25
20
1+1 protection
DSX panel
15
10
6
Figure 8.4 Microwave equipment rack.
Rectifier and battery backup
Transmission Equipment
357
Mechanical Characteristics
The ODU consists of the antenna (Ø = 0.45m, 0.6m, or 1.2m), a branching unit, and one or two transceivers, depending on the selected configuration. The unit may be attached to a vertical pipe-mount (Ø = 115 mm or Ø = 75 mm). A pipe-mount arrangement for tower, wall, and roof mounting will depend on the location where the ODU and antenna are mounted. The IDU can be installed as a stand-alone unit or it can be mounted within an ETSI standard cabinet (ref. ETS 300 119) or 19-inch standard cabinet (ref. IEC 297-2 and IEC 297-3). Sometimes IDUs can be wall-mounted as well. Dimensions and weight are as follows: • Dimensions (1 + 0 Terminal) •
IDU: 483 mm (W) × 250 mm (D) × 50 mm (H)
•
ODU: 135 mm (W) × 310 mm (D) × 430 mm (H)
• Weight for a complete 1 + 0 terminal •
IDU: approximately 4.0 kg
•
ODU (without antenna): approximately 9.0 kg
Power Supply
The equipment operates from a battery supply between −40.5V and −57V, nominally −48 VDC according to ETS 300 132-2. The primary dc power is supplied to the IDU through a main fuse and a filtering function, which includes input filter to attenuate the common mode noise. The power to the ODU is supplied from the indoor unit via the intermediate frequency (IF) coaxial cable. Power consumption is as follows: IDU: <30W ODU: <48W 1 + 0 system: <78W 1 + 1 system: <170W 8.1.3.3 Transmission Interfaces
The IDU is equipped with one main STM-1 data interface. The main data interface is in the standard version STM-1 electrical. As an option this interface may be STM-1 optical. In the following sections the main characteristics of the interfaces are listed.
358
Transmission Systems Design Handbook for Wireless Networks
Transmission Interface CharacteristicsSTM-1 Electrical Electrical Interface According to ITU-T Rec. G.703 Bit rate:
155.520 Mbps ± 20 ppm
Line code:
CMI
Impedance:
75Ω unbalanced
Return loss (8240 MHz):
≥ 15 dB
Pulse amplitude:
1.0V 0.1V
aximum attenuation of input signal at 78 MHz:
12.7 dB
Connector type:
DIN47297, 1.0/2.3 mm
Input jitter and wander tolerances are according to ITU-T Rec. G.783. Detailed parameters are given in ITU-T Rec. G.958. Transmission Interface CharacteristicsSTM-1 Optical Optical Interface According to ITU-T Rec. G.957; S-1.1 Bit rate:
155.520 Mbps ± 20 ppm
Operating wavelength range:
1,2611,360 nm
Source type:
MLM
Mean launched power:
Maximum: −8d Minimum: −15 dBm
8.1.4
Spectral width RMS:
Maximum: 7.7 nm
Minimum extinction ratio:
8.2 dB
Attenuation range:
012 dB
Minimum receiver sensitivity (BER < 10−10):
−28 dBm
Minimum overload:
8 dBm
Connector type:
SC
Microwave Antennas
The microwave antenna is a highly directional, parabolic, dish-shaped radiator that is connected to the microwave transmitter (through RF branching networkRF circulators, filters, couplers and switches) via coaxial cable or waveguide. Microwave antennas are available in many sizes to satisfy the requirements of a particular application [2]. Standard dishes, open-grid (for low wind loading), or high-performance, and single or dual polarized
Transmission Equipment
359
antenna models are available. Many antennas come with galvanized steel mounts based on EIA standards RS195B and RS222C. Optional highperformance antenna versions include an RF shroud to improve side-lobe performance and a planar radome to protect the antenna against ice or snow accumulation. Generally speaking, the larger the antenna diameter, the higher the antenna gain relative to the isotropic antenna, and the smaller the beamwidth (Figure 8.5). As the antenna beamwidth decreases, antenna alignment and, thus, stability become more critical. Furthermore, weight and wind loading are greater with large antennas. As a consequence the antenna mounting structure must be several times more rigid (twist and sway) for each successive increase in antenna size. The selection of antenna size should be based upon the results of path analysis and calculations. The antenna size should be determined before a frequency coordination study can be performed and before applying for the license to operate the microwave system. Figure 8.6 shows some of the most common terrestrial microwave antennas. Optimum antenna alignment occurs when both transmitting and receiving antennas are precisely aimed at each other in both azimuth and elevation. Azimuth is the angle in the horizontal plane with respect to true north and elevation is the angle in the vertical plane with respect to the horizontal plane. The main parameters of interest when choosing the microwave antenna are as follows: • Operating frequency band; • Radiation pattern; • Gain; 2
Ga(dBi)=10 log10h[4pAa/l ]
Where Ga = Antenna directional gain h = Aperture efficiency Aa = Antenna aperture area l = Wavelength
Figure 8.5 Parabolic-antenna directional gain.
360
Transmission Systems Design Handbook for Wireless Networks
a)
c)
b)
d)
Figure 8.6 Basic microwave antenna types: a) Standard parabolic antenna, b) antenna and outdoor MW radio unit, c) shielded antenna, and d) grid antenna.
• Polarization (single- or dual-polarized); • Half-power beamwidth; • Wind load; • Front-to-back ratio; • Cross-polarization discrimination (in dB, the difference between the
peak of the copolarized main beam and the maximum crosspolarized signal over an angle twice the 3-dB beamwidth of the copolarized main beam);
• Isolation (between inputs of single-band, dual-polarized antennas).
Additional options for most microwave antennas that could be ordered from their manufacturers include the following:
Transmission Equipment
361
• Input flanges; • Antenna color; • Radoms (reflector protectors) and their colors; • High-wind survival options; • Corrosive environment; • Packing type and quantity; • Reflector types; • Special-purpose antennas; • Special accessories, such as struts and ice shields. 8.1.5
Transmission Lines
Elliptical waveguides are commonly used for terrestrial microwave systems. They are precision-formed from corrugated high-conductivity copper and have an elliptical cross section. The corrugated wall gives the waveguide increased crush strength, light weight, and relative ease of handling. Generally speaking, elliptical waveguide installation is a very tricky job and has to be performed by trained and experienced microwave technicians. An elliptical waveguide may be ordered with the connectors attached and tuned when the precise length is known in advance. If the connectors are to be installed in the field it may be necessary to sweep and tune the waveguide and connectors to realize optimum return loss. The waveguide must be supported with special hangers at regular intervals to prevent stress, movement, and excessive pulling forces due to the cumulative weight of the waveguide. Hangers are usually installed at 3-ft (1m) intervals. Waveguide (as well as coaxial cable) must be grounded using special grounding kits. It is common practice to ground the top of the waveguide run, the bottom of the run, and at the radio terminal. The waveguide or air dielectric coaxial cable must be pressurized with equipment designed to dry the pressurized air. This action is necessary to displace any water vapor from the waveguide or cable that will significantly increase attenuation. The pressurization equipment may be either a mechanical compressor or a container of compressed gas (must be replaced regularly) equipped with a pressure regulator. There are several types of mechanical compressors available with various capacities, operating from ac or dc line voltage and operating in manual or automatic drying cycles. Pressurization equipment is normally attached to the indoor end of the waveguide or coaxial cable, through a manifold.
362
8.1.6
Transmission Systems Design Handbook for Wireless Networks
Environmental and Quality Issues
Some microwave radio products are widely deployed in geographic areas recognized for extreme environmental conditions as the Middle East, Asia, Eastern Europe, South America, and North America. All the radios have to be thoroughly tested and cycled in temperature prior to shipment. The following text describes the standard tests performed on MW radios prior to shipment, additional tests performed, and a brief description of quality processes. MW Radio Standard Test Process
The MW radio, in case of split configuration, is composed of two primary elementsan IDU and an ODU. Prior to shipment, both are tested according to the following list: • In-circuit testing (ICT) performed by the Printed Circuit Board
Assembly (PCBA) supplier;
• Bench testing, functional testing performed at ambient temperature
(25°C);
• Temperature cycling of IDU (0°C to 50°C).
Equipment is powered on and monitored according to the following: • At ambient temperature (25°C); • Cycled to the minimum temperature (rate of 1°C/min) and held for
4 hours;
• Cycled to the maximum temperature (rate of 1°C/min) and held for
4 hours;
• Cycled to ambient temperature (rate of 1°C/min); • Temperature cycling of the ODU (−30°C to 55°C). Additional MW Radio Tests
Some microwave radios that are to be deployed in very harsh environment, could be tested with additional high-temperature/low-temperature thermal shock tests. In order to verify the good operation of the radios under test, BER measurements are performed during the entire test. Output power readings are measured using the RSL display of the radio or external power meter.
Transmission Equipment
363
• Number of hops can be subjected to high-temperature conditions
(70°C) for duration of 30 days (there should be no failures); • Number of hops can be cycled from −30°C to 55°C, 2 hours at each extreme, for 40 cycles (there should be no failures). Environmental Stress Screening
Environmental stress screening (ESS) is an effective method for improving the reliability of electronic systems through the elimination of latent (hidden) physical defects from electronic components and assemblies. Examples of these defects include the following: • Faulty workmanship (e.g., soldering, lead forming, fastening); • Internal cracks or dislocations; • Thermal or mechanical stress points; • Damaged packages or leads; • Delamination. 8.1.7
Standards and Recommendations
A partial list of international standards that apply to the split-configuration microwave radios in 13, 15, 18, 23, and 38 GHz is shown in Table 8.2. Different countries may have different requirements for the imported telecommunications equipment (different from the internationally accepted standards) so it is necessary to check all the requirements prior to shipping of the equipment. It is important to remember that certification and homologation of the equipment could be a lengthy process, which in some countries can take from a few months to more than a year.
8.2 Fiber-Optic Equipment 8.2.1
SONET and SDH
SDH and SONET were designed to allow for flexibility in the creation of equipment for the telecommunications needs. The following are some of the key transmission products: • Fiber-optic line equipment;
364
Transmission Systems Design Handbook for Wireless Networks Table 8.2 Table of Compliance with International Standards
Requirement
Standard
EMC
Europe ETS 300 385, EMC Standard for Digital Fixed Links and Ancillary Equipment with Data Rates at Around 2 Mbps and Above Radiated Emissions EN55022 Class B dc Conducted Emissions EN55022 Radiated Susceptibility ENV 50140 (3 V/m) Electrostatic Discharge EN60801-2 (+/−4 kV contact, +/−8 kV air discharge) Electrical Fast Transient IEC 801-4 (+/−1kV) Conducted Susceptibility ENV 50141 (3 Vrms) IEC 1,000-4-2 ESD (+/−8 kV contact, +/−15 kV air) 1,000-4-3 Radiated Susceptibility (15 V/m) 1,000-4-4 Electrical Fast Transient (+/−1 kV) United States Radiated Emissions FCC part 15 Subpart B Class B Bellcore GR-1089-CORE Canada Radiated Emissions ICES-003 Class B
Electrical safety
Europe EN60950, Safety of Information Technology Equipment, Including Electrical Business Equipment EN60215, Safety Requirements for Radio Transmitting Equipment United States UL 1950, Safety Requirements for Radio Transmitting Equipment
Environmental
United States Bellcore (called Telecordia today) GR-63-CORE (excluding altitude) Temperature, Humidity Tests Fire Resistance Test Equipment Handling (Drop) Test Earthquake (Zone 4) Airborne Contaminant Test
Transmission Equipment
365
Table 8.2 (continued) Requirement
Standard
Environmental
Acoustic Noise Test Illumination Test Bellcore GR-487-CORE section 3.27, Wind-Driven Rain Intrusion Test (110 mph) Bellcore GR-487-CORE section 3.31, Wind Resistance (100 mph) MIL -STD- 810C Method 506.1 Procedure I, Rain Test MIL-STD-810E, Solar Radiation Test
EMC
Europe ETS 300 385, EMC Standard for Digital Fixed Links and Ancillary Equipment with Data Rates at Around 2 Mbps and Above Radiated Emissions EN55022 Class B dc Conducted Emissions EN55022 Radiated Susceptibility ENV 50140 (10 V/m) Electrostatic Discharge EN60801-2 (+/4 kV contact, +/15 kV air discharge) Electrical Fast Transient IEC 801-4 (+/1 kV) Conducted Susceptibility ENV 50141 (3 Vrms)
Electrical safety
Europe EN60950, Safety of Information Technology Equipment, Including Electrical Business Equipment EN60215, Safety Requirements for Radio Transmitting Equipment
• Microwave equipment; • Multiplexers (ADM, hub, ring, terminal); • DACS.
Optical line systems, and to a lesser extent, microwave radio systems, provide the transmission-bearer backbone for the SDH/SONET networks. A multiplexer is a physical-layer device that combines multiple data streams into one or more output channels at the source. Multiplexers demultiplex the channels into multiple data streams at the remote end and thus maximize the use of the bandwidth of the physical medium by enabling it to
366
Transmission Systems Design Handbook for Wireless Networks
be shared by multiple traffic sources. Terminal multiplexers provide access to the SDH/SONET network for various types of traffic using traditional interfaces, such as 2 or 1.5 Mbps. An ADM can offer the same facilities as terminal multiplexers, but they can also provide low-cost access to a portion of the traffic passing along a bearer [4]. Most designs of ADM are suitable for incorporation in rings to provide increased service flexibility in both urban and rural areas. ADM ring design also employs alternative routing for maximum availability to overcome fiber cuts and equipment failures. What is the APS and what is the difference between 1+1 protection and 1:1 protection? The APS as specified in ITU-T G.783 is applied for line-switching procedures using SDH K1/K2 byte and its restoration time is within 50 ms. In the 1+1 configuration, the same signal is transmitted over the working and standby lines. In the 1:1 configuration, the idle signal or extra traffic signal is transmitted on the standby line. Both 1+1 and 1:1 protections perform 100% restoration, while the 1:N protection performs partial restoration. Hub multiplexers provide flexibility for interconnecting traffic between bearers, usually optical fibers. A hub multiplexer is connected as a star, and traffic can be consolidated or services managed while standby bearers between hubs provide alternate routing for restoration. Several rings of ADMs can converge on a single hub, providing interconnection of traffic between those rings and connection into the existing network. Usually, multiplexers are designed to be flexible and can be configured in a way that they can provide different functionality based on the network requirements. As discussed in earlier chapters, typical DWDM multiplexer technical specifications contain client-side data as well as some additional information about the different wavelengths available in this unit: Optical channel capacity: Client capacity: Protection switching: Nodes in ring: Client STM-16/OC-48 interface line rate: Wavelength: Output signal level: Minimum receiver level (BER 1 10-10):
2,488.32 Mbps (STM-16/OC-48 equivalent) STM-16, maximum four channels Optical channel 1:1 BWPSR Maximum 16 2,488.32 Mbps 1,266 ∼ 1,360 nm −10 to −3 dBm Min. −18 dBm
Transmission Equipment
Minimum overload (BER 1 10-10): Output channel wavelength (nm):
Power requirements: Dimensions:
367
Min. −3 dBm 1,538.98, 1,539.77, 1,540.56, 1,541.35, 1,542.14, 1,542.94, 1,543.73, 1,544.53, 1,545.32, 1,546.12, 1,546.92, 1,547.72, 1,548.51, 1,549.32, 1,550.12, 1,550.92, 1,551.72, 1,552.52, 1,553.33, 1,554.13, 1,554.94, 1,555.75, 1,556.55, 1,557.36 −40 ∼ −57.5 VDC 442 mm (H) × 584 mm (W) × 279 mm (D)
What follows is a partial list of some of the ITU-T recommendations relevant to fiber-optic equipment: G.664Optical safety procedures and requirements for optical transport systems. G.691Optical interfaces for single-channel STM-64, STM-256 systems, and other SDH systems with optical amplifiers. G.692Optical interfaces for multichannel systems with optical amplifiers. G.703Physical/electrical characteristics of hierarchical digital interfaces. G.704Synchronous frame structures used at 1,544, 6,312, 2,048, 8,448, and 44,736 Kbps hierarchical levels. G.707Network node interface for the SDH. G.781Synchronization layer functions. G.783Characteristics of SDH equipment functional blocks. G.784SDH management. G.803Architecture of transport networks based on the SDH. G.813Timing characteristics of SDH equipment slave clocks (SEC). G.823The control of jitter and wander within digital networks which are based on the 2,048-Kbps hierarchy. G.825The control of jitter and wander within digital networks which are based on the SDH. G.826Error performance parameters and objectives for international, constant bit-rate digital paths at or above the primary rate. G.828Error performance parameters and objectives for international, constant bit-rate synchronous digital paths.
368
Transmission Systems Design Handbook for Wireless Networks
G.841Types and characteristics of SDH network protection architectures. G.842Interworking of SDH network protection architectures. G.957Optical interfaces for equipment and systems relating to the synchronous digital hierarchy. G.958Digital line systems based on the synchronous digital hierarchy for use of optical fiber cables. 8.2.2
OPGW
One type of optical fiber network is an optical communication network via OPGW, which makes use of the power transmission line network (Figure 8.7), or more precisely, the use of power ground wire. OPGW is designed for installation on transmission and distribution lines to carry voice, data, and video communications. Because of the outstanding properties of optical fiber, optical communication systems are used in many areas of application, especially in electric power supply, public communications, and road and railway traffic, where the distinctive features of the fibers are utilized effectively. Therefore, optical fibers are increasingly in use for overhead transmission lines (Figure 8.8). Advantages of the OPGW as a transmission medium are as follows: • It provides large transmission capacity. • It provides long-distance nonrepeatered transmission.
Figure 8.7 OPGW cable.
Transmission Equipment 1 Tension assembly set 4 Tower fixing clamp (R, Y1, Y2)
Vibration damper (stock-bridge type)
369
2 Suspension assembly set
Air ball armor rods
3 Vibration damper
Join box
Tension tower
Tension tower
Suspension tower
Tension tower
Figure 8.8 OPGW installation.
• It is compact in size and light weight. • Optical fibers are housed in spiral grooves of an aluminum spacer to
be protected against lateral pressure. • Optical fibers are coated with heat-proof materials to get high thermo-resistant characteristics. • Since an optical-fiber unit is placed inside an aluminum tube, optical fibers are protected against severe environmental conditions and external lateral force. • Since a diameter of OPGW is nearly equal in size to conventional groundwire, the towers do not require any reinforcement.
The cable core (Figure 8.9) consists of either a single central loose tube or a loose tube bundle stranded around a central strength member. The fibers in the loose tubes can be single-mode or multimode. A tube-filling compound provides both water blocking and prevention of mechanical damages. The cable core elements are protected against mechanical damages by a pultruded fiber reinforced plastic (FRP) tube. The FRP sheath provides excellent protection against tensile and compression forces that may be incurred during installation and service. It also resists against rodents, termites, and bullets. Coupled with the aluminum outer sheath, it lies more uniformly, concentrically, and tightly around the cable core elements than
370
Transmission Systems Design Handbook for Wireless Networks
Optical fiber a)
FRP Heat-resistant polymer, filled Coating layer
6-core unit b)
AL spacer
AL tube
Figure 8.9 OPGW cross section: a) 6-core unit, and b) optical unit.
welded aluminum tube. The cable is capable of withstanding harsh environmental conditions. The standard cable is sheathed and comes with different thickness of aluminum on demand for special applications.
8.3 Wireline Equipment 8.3.1
Digital-Access Cross Connects
The use of DACS (also known as DXC, DXCS, or DCS) for efficient network operation is established and very common in the wireless industry. They are classified as DCS p/q, where p is the hierarchical order of the port bit rate and q is the hierarchical order of the traffic component that is switched within that port bit rate. Wireless service providers use these electronic switching platforms to maximize the use of transmission facilities and reduce costs. Application of the 1/0 DACS is shown in Figure 8.10. This is the equipment used for subT1 or E1 circuit grooming and cross connecting.
Transmission Equipment West
371
Digital radio
MUX (E1/T)
Fiberoptic terminal
Digital access PCM (E1/T) channel MUX (E1/T) cross-connect system (DACCS) bank
System administrator
(E1/T) MUX
(E1/T)
Digital radio
East
Voice data
Digital PABX
The digital cross-connect (1/0 system) is a fully electronic switching device that interconnects DS0 64 Kbps subchannels of E1/T1 circuits.
Figure 8.10 DACS application.
Wireless service providers must support easy and in-service growth, flexibility to changes in traffic patterns, a high level of protection, and optimized bandwidth utilization. All this must be accomplished with low operation and maintenance costs and with crucial time limitations. A cellular/PCS network can have thousands of cell sites spread over large geographic areas with extensive numbers of links connecting the cells to the BSC/MSC. Whether the cellular service providers own their infrastructure or lease lines from the local carrier, optimal utilization of the bandwidth results in enormous savings in the overall operation costs of the network. A typical DACS application in the wireless network is shown in Figure 8.11. There are two main types of DACS: nonblocking and blocking cross connects. Nonblocking means that any input and any output can always be connected to each other. Blocking is when there is a chance that it wont be able to connect input and output. In transmission networks, it is required that any path can be switched without influencing an existing path. So for the cross connect, nonblocking is preferable to blocking. A DACS should allow nonblocking connections between any of its ports and provide major flexibility points for network management. A DACS system provides a wide range of services required in the cellular network. Those are add and drop and grooming capabilities, switching rates from 64 Kbps up to 155 Mbps, and line interfaces ranging from E1/T1
372
Transmission Systems Design Handbook for Wireless Networks
1xE1 1xE1/T1
PSTN 5 2xE1/T1
Cell site 10
Cell site 6
1xE1/T1
PSTN 4
1xE1/T1
1xE1/T1
3xE/T11 Cell site 7
Cell site 5 4xE1/T1
Calling area 1 BSC
1xE1 Cell site 11
6xE1/T1 DACS
PSTN 1
2xE1/T1 Cell site 3
1xE1/T1 Cell site 4
3xE1/T1 6xE1/T1
1xE1
PSTN 2
4xE1
PSTN 3
1xE1/T1 Cell site 8
1xE1/T1
3xE1/T1
Cell site 1
1xE/T11
Calling Area 2 Cell site 9
MSC Legend xE1
DACS
Microwave system or leased lines Capacity requirements BSC-MSC trunks (microwave or leased lines) BSC-MSC trunks (microwave or leased lines) BSC-PSTN trunks (leased lines or microwave) Digital access cross-connect system ( 1/ 0 or 3/1/0 for large networks)
Notes: 1. The BSC and cell site 11 are colocated Notes: 2. The MSC may or may not be colocated with the BSC.
Figure 8.11 DACS application with single BSC.
Cell site 2
Transmission Equipment
373
through E3/T3/STS-1, up to STM-1/OC-3/STM-4/OC-12, while maintaining the synchronization, framing, and control needed for each digital trunk. This way, the operator can efficiently fill or groom DS0s into the T1/E1 pipe. DACS also supports voice traffic along with data traffic, and can be implemented in both mesh and ring topologies. It is important for every wireless operator to be able to have DACS that will facilitate easy and inservice growth, a high level of protection, and extensive performance monitoring and testing in order to reduce troubleshooting time and guarantee quality service. Continuous performance monitoring is usually available on all interface ports with immediate alerting upon failure detection or performance degradation. Remote provisioning assures faster response to customer orders and rapid provisioning of new services, as well as changes in existing services. Support of the T1.403 standard enables in-band remote performance monitoring (on T1 lines) and advanced troubleshooting operations at remote sites. Some of the new features these days enable network service providers to deliver thousands of lines of dedicated IP, frame-relay, and voice services on a single, flexible platform. The solution supports the grooming and concentration of fractional and full T1/E1 rates for dedicated frame relay, IP over frame relay, and IP over sync PPP, which are the prevalent modes of delivering data services to business users. It can also offer concentration and adaptation of voice or other constant bit rate services to ATM transmission networks. Some additionally provide a gateway capability between SONET and SDH networks, enabling global carriers to optimize their bandwidth utilization. Cellular networks are rapidly expanding, and wireless operators are deploying an increasing number of cell sites and microcells. This rapid growth of cellular services has made the task of managing network and facility performance both a critical and tremendous job for wireless operators. For example, degraded performance on a T1 line impacts cellular customers, with problems such as dropped calls and poor reception. A network outage means loss of revenue until service is restored or switched to a spare T1 line. Effective network management of unmanned sites requires the ability to remotely detect and isolate alarms and conduct equipment tests. Remote management demands that the equipment be visible from all parts of the systemthe cell site, MSC, NOC, and on any SS7 links. This requires products that easily communicate with each other and with standard networkmanagement interfaces, such as SNMP and OAM&P. In addition to the management issues, accommodating growth in a cellular network often means deploying more outdoor microcells. Equipment must withstand
374
Transmission Systems Design Handbook for Wireless Networks
severe outdoor temperatures, mount compactly in tight areas, and provide front access. The mini-DACS is used for small-capacity applications (416 T1/E1s) that require add and drop, multiplexing, or full 1/0 cross-connect functionality (Figure 8.12). As an example, it can act as a cell-site multiplexer terminating a T1/E1 circuit and also provide access to a single 64-Kbps data channel. Mini-DACS can also provide cell-site grooming (TDMA-based networks) and other functions in wireless applications. In short, it provides full crossconnect functionality in any bandwidth-management application. Sometimes ADMs and hub multiplexers, which include time-slot interchange, also can be used as small nonblocking DCSs. A ring of several ADMs can be managed as a distributed cross connect, but typically they will experience some blocking, which must be anticipated in transmissionnetwork planning.
Radios
Cell site 1
FT1
Radios
Cell site 2
FT1
Hub site 1 Cell site 4
Hub site 2 Cell site 5
Radios
Radios T1
Radios
Mini-DACS
Cell site 3
Mini-DACS
MSC T1 3/1/0 DACS
FT1 FT1 - Fractional T1
Figure 8.12 Mini-DACS application.
Transmission Equipment 8.3.2
375
CSU/DSU
The channel service unit/data service unit (CSU/DSU) is actually two devices typically packaged as a single unit. The CSU performs protective and diagnostic functions for a telecommunications line; the DSU connects a terminal to a digital line. The CSU/DSU is required for both ends of a T1 or T3 connection, and the units at both ends must be set to the same communications standard. In the wireless network every radio base station has a T1 link to the BSC, using some type of CSU/DSU. Without the intelligent CSU, if a site in a wireless network went off the air, NOC would get an alarm at the network center but have no idea what the problem was. It is important that CSU has intelligence and the ability to transmit diagnostic information so that it can be determined whether T1 is failing or if the problem has to do with the cell-site equipment. As a result, the operations group knows whether to call the telephone company that is providing the T1 circuit or dispatch a technician to fix the cell-site equipment. The two most important tests performed by CSU/DSUs are network loopback and payload loopback (Figure 8.13). The network loopback test verifies the operation of the T1 network and is available only on full bandwidth. It loops the data received from the T1 Network loopback DTE
Local CSU/DSU
Data port DSU
T1 network
Network interface
Remote CSU/DSU CSU Data port
Network interface DSU
CSU Payload loopback DTE
Local CSU/DSU
Data port DSU
Network interface CSU
Figure 8.13 CSU/DSU loopback.
T1 network
Remote CSU/DSU CSU Data port
Network interface DSU
376
Transmission Systems Design Handbook for Wireless Networks
network back to the network. The data is regenerated before it is looped back; however, the unit does not perform additional processing of the data. This minimizes the impact of the unit during the test so that network problems can be isolated. The payload loop test verifies proper operation of the unit and the T1 network. It loops the payload data received from the T1 network back toward the network. Before it is looped back, the data is regenerated and a new framing pattern is inserted. Thus, the proper T1 framing of the CSU/DSU and network can be verified. The individual channel payload loops a selected fraction of the T1 signal toward the network. In this mode, the selected data is corrected for BPV, CRC, and framing errors, and the FDL is regenerated before the data is looped back to the network. The loop-up remote test puts the remote unit into network loopback using the industry standard set codes. Once in loop-up remote, test patterns can be sent to verify the BER performance of the bidirectional T1 network signal. To put the remote unit into network loopback, the local CSU/DSU momentarily transmits the industry standard loop-up code to the remote CSU/DSU. The loop code and network parameters for the local and remote units must match. The loop-down remote test is used to terminate the remote loopback. Sometimes in the case of a multivendor T1 circuit (the end-to-end T1 circuit is provided by more than one carrier), sending the loop-up command can trigger intermediate CSU/DSUs to go into loopback mode. It is important to identify all of the sections of the T1 circuit between end points and then troubleshoot one at a time. This could be a problem, and the troubleshooting of such a circuit is a very time-consuming and labor-intensive process. 8.3.3
DSL and ADSL
The innovation of digital technology has given tremendous advantage to the utilization of copper lines. Traditionally, PCM systems were mostly used for junction working between two exchanges. These systems were capable of working up to 20 to 24 regenerators with power-feeding arrangements from the local exchanges; however, this technology got isolated after the induction of fiber-optics-based digital technology, which gives very high bandwidth, span, and quality benefits over the conventional cable-based PCM systems. Digital Subscriber Line (DSL) refers to a variety of systems that are designed to provide more capacity on the existing embedded copper loop plants in the access network. DSL technologies offer equipment catering to a variety of services and are categorized in two broad categories: symmetrical transmission (HDSL) and asymmetrical transmission (ADSL).
Transmission Equipment
377
HDSL provides a bidirectional, high-data-rate service that enables provision of services where high data rates are required in both directions [5]. HDSL provides a bidirectional data rate of 1.544 Mbps or 2.048 Mbps over the unmodified subscriber line (copper), for a maximum distance of about 3.6 km from the exchange. More than 60% of leased T1 lines in North America have HDSL. HDSL is used in Europe (8-wire E1) and North America (4-wire T1). Some companies are already developing HDSL2 (2-wire T1) also called SDSL (single-line DSL). HDSL2 is a technology that transmits a T1 on 2-wire (one pair) 24 AWG copper wire without repeaters for 12,000 ft. In 1961 the Bell System deployed the first digital T1 circuit. The T-Carrier System is a two-way transmission path, one cable pair for each direction of transmission. This T1 replaced analog carriers that were being deployed. It is also important to note that in this traditional method of deployment, repeaters are required at about every 6,000 ft for the intermediate repeater sections. The first repeater from the central office is usually less than 3,000 ft away, and the repeater closest to the customer location is usually also less than 3,000 ft away. In the early 1990s HDSL circuits were developed. These circuits helped the carriers to deploy T1 circuits faster with less engineering. HDSL is called repeaterless T1 because it can reach 12,000 ft on 24-gauge wire. This requires a transceiver unit (HTU-C) at the telephone company CO and a transceiver unit at the end point. The transmit pair and receive pair do not need to be binder-group separated as in traditional T1. If more distance is needed, one or two HDSL range extenders (HREs) or HDSL doublers can be deployed for 24,000 or 36,000 ft of loop length. The HTU-C powers the HTU-R and HREs. HDSL is also useful in a campus or dry-wire application. In the late 1990s HDSL2 used the Trellis Coded pulse amplitude modulation (PAM) line code. This yields three bits per baud. Trellis-coded PAM along with partially overlapped echo-canceled transmission (POET) allows HDSL2 to deliver all 24 channels using two wires, while maintaining a loop distance of 12,000 ft on 24 AWG. In comparison, the first version of HDSL uses the 2B1Q line code, which yields 2 bits per baud. HDSL2 and HDSL can both tolerate bridged taps, but neither can tolerate load coils. A single bridged tap can be no more than 2,000 ft, and the total length of all bridged taps cannot exceed 2,500 ft. As mentioned earlier, HDSL2 and HDSL can reach a distance of 12,000 ft on 24-AWG cable nonloaded, nonbridged tapped cable. This distance is 9,000 ft if the cable is 26-AWG.
378
Transmission Systems Design Handbook for Wireless Networks
Internet access, competition from cable modems, and interest from the personal computer and modem industries are strong drivers of ADSL technology. ADSL is, essentially, a modem that employs a very sophisticated coding scheme. This coding scheme permits transmission over copper pairs at rates as high as 6 Mbps for distances of 9,000 to 12,000 ft. Speeds of this magnitude bring to mind television signals; a 6-Mbps channel can easily handle a television movie. ADSL succeeds because it takes advantage of the fact that most of its target applications (video-on-demand, home shopping, Internet access, etc.) function perfectly well with a relatively low upstream data rate (hence the term asymmetric). LECs are now using ADSL as an access technology for their television businesses and for Internet access. An example of ADSL performance with crosstalk noise is shown in Table 8.3. An example of ADSL lite performance with cross-talk noise is shown in Table 8.4. Some of the new technologies, such as DSL, also present new and different challenges in terms of power protection. DSL is a unique application in which there is a digital and an analog signal on the same pair of wires but they require a completely different approach in protection. DSLs are designed to carry large amounts of traffic between phone companies central offices and customer premises. Because DSL uses the phone companies ordinary copper wire for communication, carrying voice traffic simultaneously with data traffic is imperative. At the customer premises, both the data network and conventional phones connect to an integrated access device (IAD), transferring both voice and data over the DSL signal. At the CO end, the DSL gateway extracts the voice traffic from the data traffic and sends it over the conventional telephone system, again using SS7. Thus, the DSL gateway can be a replacement for a standard Class 5 switch (or coexist as a peer), which means it must have an SS7 interface. This scenario offers a very attractive service to companies that want to compete with the local telephone operating company. By transferring telephone traffic Table 8.3 ADSL Performance with Cross-talk Noise Downstream (Kbps)
Upstream (Kbps)
Loop Length (km)
Loop Diameter (mm)
4,096
320
2.8
0.4
2,048
128
3.5
0.4
576
128
4.2
0.4
Transmission Equipment
379
Table 8.4 ADSL Lite Performance with Cross-talk Noise Downstream (Kbps)
Upstream (Kbps)
Loop Length (km)
Loop Diameter (mm)
1,536
256
2.8
0.4
1,536
96
3.5
0.4
512
96
4.2
0.4
over the DSL data path, the company can capture the local, as well as longdistance, phone traffic and provide bundled data service to differentiate itself from competitors. To achieve this, Loop Emulation Service (LES) using AAL2 for narrowband services was developed to provide an effective transport mechanism to carry voice, voice-band data, and fax traffic over xDSL. The architecture has three key components: (1) AAL2-enabled customer premises equipment, (2) a digital subscriber loop access multiplexer (DSLAM) providing ATM cell multiplexing, and (3) a VoDSL gateway. The solution delivers tollquality voice as well as the feature set of the conventional wireline services, such as call waiting, caller ID, and so on. The CPE is typically an IAD, which packetizes the analog and digital voice traffic into AAL2 packets. These packets along with signaling events are sent along a single xDSL over to a DSLAM, where the voice traffic is aggregated onto a permanent virtual connection (PVC) to the VoDSL gateway. The gateway completes the call by performing not only the ATM-to-TDM conversion and vice versa, but also provides the GR303/TR08-based signaling interworking that is needed between the xDSL interface and existing PSTN switch interfaces. In the future, there will be two changes in the access architecture. First, the standalone VoDSL gateway functionality will be integrated into the DSLAM, eliminating the need for the CLEC or ILEC to deploy and maintain more equipment. Second, new signaling and control capabilities will be introduced enabling a distributive-switching model based on Media Gateway Control Protocol (MGCP). 8.3.4
Echo Cancellers
Echo cancellers are predominately used in three applications: TDM longdistance networks, wireless networks, and Voice over X (VoX) networks. Long-distance networks were the first to use echo cancellers and continue to
380
Transmission Systems Design Handbook for Wireless Networks
make up the majority of users in the worldwide telecommunications network. Wireless networks still see exponential growth and are deploying echo cancellers at a rapid rate. Voice calls over IP networks are now expected to double each year for the foreseeable future and will also require echo-cancellation devices to be utilized on the voice path. 8.3.4.1 Wireline/PSTN Long-Distance Networks
In long-distance networks, echo cancellers are used to remove the echo made noticeable primarily by the delay inherent in transmission facilities. Two echo cancellers per transmission path are generally used in long-distance networksone at the far end and another at the near end (see Figure 8.14). An echo canceller placed at the far end of a transmission path protects the nearend caller from echo, while the near-end echo canceller protects the far-end caller from echo. 8.3.4.2 Digital Wireless Networks
In a typical wireless-to-wireline phone call, two types of echoes exist. Hybrid echo on the PSTN end of the phone call is caused by the reflection resulting from the four-wire-to-two-wire impedance mismatch. For echo to be noticeable, the human ear must detect some delay between the source signal (in this case, the spoken word) and the echo signal. In typical local-loop applications, this echo is not noticeable because the delay is so short that the human ear E C
CO
LD switch
LD switch
LD switch E C
E C
CO
CO
CO-Central office LD-Long distance
Figure 8.14 Echo cancellation in wireline networks.
No echo No echo
E C
CO
Transmission Equipment
381
does not separate the original speech from the echo. Typical long-distance applications induce delay primarily due to propagation and thus require hybrid echo cancellers for correct operation. In a wireless network (Figure 8.15), however, propagation is a secondary issue because there is always processing delay introduced into the propagation path through the network. This delay is typically composed of speech coding, processing, interleaving, and radio path delays. All digital wireless networks involve a highly complex infrastructure, which can introduce up to 300 ms of round-trip delay, more than sufficient to produce a noticeable echo and degraded speech quality from the hybrid mismatch. This echo is specifically dealt with in the mobile network, typically at the MSC, through the specifications that require cancellation looking toward the PSTN. Unlike long-distance networks, digital wireless networks use echo cancellers to remove echo that is made audible primarily by processing delays caused by speech-compression equipment and not by transmission-facility delays. All digital wireless networks use some form of compression to reduce the amount of radio bandwidth required to carry the conversation. Compression induces delays that can cause echo to be heard on calls made by the mobile telephone user. Echo cancellers are used in all digital wireless networks in order to ensure that the digital mobile telephone user does not hear echo when placing a call into the local telephone exchange. Also, unlike long-distance networks, digital wireless networks normally use only one echo canceller on each facility. These echo cancellers point into the PSTN and protect the mobile subscriber from hearing echo caused by reflections in the two-wire local loop.
PSTN
MSC
E C
E C
BSC
RBS
RBS
Figure 8.15 Echo cancellation in mobile networks.
382
Transmission Systems Design Handbook for Wireless Networks
Some echo protection may be provided for the PSTN subscriber by echo control circuitry built into the mobile telephone. The quality of this echo control can be limited and may vary from one phone manufacturer to the next. Better mobile-to-mobile voice quality can be maintained by placing high-quality echo cancellers in between the MSC and the BSC. The evolving digital PSTN requires a loss plan to ensure that appropriate transmission levels exist at the various A/D conversion points (see ITU-T Recommendations G. 223, V.2, and M.1050). With such a plan, PCM overload distortion is avoided and signal levels allow the echo canceller to operate as per its design intent. Guidance for transmission levels can be found in the G.100 series of ITU-T recommendations for PSTNs that utilize analog accesses and for connections from digital cellular networks. Encoders should be consistent with ITU-T Recommendation G.711. For PSTNs with digital access, guidance for terminal design can be found in ITU-T Recommendation P.310. 8.3.4.3 Echo Cancellation in VoX Networks
Echo cancellers are required in VoX-type networks because of the compression algorithms and packet delays associated with VoX calls. There are various components of packet delay in VoX networks including processing delay, queuing delay, and transmission delay. There can be additional delays in getting the packet from the network interface to the application and eventually to the caller. The operation of the echo canceller has a major impact on voice quality and the performance of an echo canceller in a VoX network should be no less than that of an echo canceller in any other network (Figure 8.16). In TDM networks, echo cancellers are typically installed at the ends of long-distance trunks (one end only). The equipment at both sides of the ATM network should perform echo cancellation. In ATM networks, echo cancellation may be appropriate over shorter-distance circuits due to the packetization delays associated with VoATM. Compression algorithms such as G.729 and G.723 have delays that already exceed the 25-ms round-tripdelay threshold and, therefore, require echo cancellation regardless of network type (TDM, FR, or ATM). Once echo cancellation is in place, perception of voice quality degradation does not occur until delays extend into the 150200-ms range. Echo cancellation is beneficial on a channel when it is carrying voice, but is often undesirable when the channel is carrying other types of information. Tone detection can be used to automatically turn off echo cancellation on a channel for the duration of data transfer (such as for a modem call) over the channel.
Transmission Equipment
E C
PBX
VoX switch
VoX switch
VoX switch
E C
CO
383
E C
CO
E C
CO
Figure 8.16 Echo cancellation in packet-switched networks.
Jitter occurs because packets travel different paths over the network to their destination and arrive at variable intervals. Holding the packets in a buffer until they all arrive can remove the jitter; however, these buffers cause more delay. Voice packets must also be dealt with in a different manner than data packets, as they cannot be lost when an IP network is carrying peak loads. To prevent voice-packet losses, prioritization protocols have been developed that give a higher priority to voice packets and put them in front of data packets. This reduces the delay of voice packets on congested nodes; however, echo cancellers are still required. 8.3.4.4 Acoustic Echo and Handset Specifications
Another echo phenomenon, acoustic echo, has become apparent in mobile networks with the proliferation of mobile phones and consumer demand for toll-quality voice (or the closest resemblance to it) in mobile networks. Acoustic echo is defined as the coupling of received voice transmission in a mobile handset or hands-free set into the microphone of that set. When acoustic echo occurs, it is the PSTN user who is discomforted. For simplicity, this discussion will refer to a mobile user and a PSTN user, even though the general case could include two mobile users. Acoustic echo is a much more complex signal than hybrid echo. The simple case of a hands-free set illustrates the example most clearly: The received signal emits from the hands-free speaker set and reflects off multiple surfaces inside an automobile. The reflections return the signal, at various time delays and amplitudes, into the microphone, and are returned to the
384
Transmission Systems Design Handbook for Wireless Networks
speakers ear. In addition, the tail circuit is nonlinear due to transcoding. Since these reflected signals are typically delayed 180 ms, the speaker hears a perceptible echo of his or her speech. Portable handsets provide a source of coupling between the earpiece and mouthpiece, either through direct coupling as handsets get smaller, in reflections off the user, or even in reflections from surrounding environs. This coupling is prominent in situations when the user increases the handset volume, as in high-background noise environments and for low-volume speakers. Wireless standard developers did not ignore acoustic echo when developing standards. As with any echo, the closer to the source it is dealt with, the more effective the solution. For acoustic echo, the source is the handset. Two aspects of handset specifications deal with the echo loss performance characteristics of handsets; one is the handset performance and the other is testing specifications. The weighted terminal coupling loss (TCLw) figure of 46 dB is the specified performance parameter for handset acoustic echo return loss (AERL). New specifications are intended to improve testing procedures via an artificial voice test stimulus in conjunction with a proposed head and torso simulator (HATS), which incorporates free air transmission; these test methods would attempt to more closely represent actual user conditions. Another important problem in mobile systems is background noise. Apart from the acoustic echo that is fed back from the mobile, the speech from the mobile will be mixed with any noise present within the vehicle. This will produce acoustic hands-free echo. Echo cancellers generally have problems eliminating echo if there is a high level of background noise.
8.4 Cabling 8.4.1
NEC Cable Categorization
The National Electrical Code (NEC) is a set of guidelines describing procedures that minimize the hazards of electrical shock, fires, and explosions caused by electrical installation. NEC types are acronyms consisting of a prefix describing cable type (e.g., coax, CATV, and fiber optic) and a suffix indicating the type of flame test it has passed and where it can be installed. Almost everyone involved with wire and cableOEM engineers, wire product engineers, distributors, installers, and architectsis affected by the NEC and must incorporate NEC guidelines into their work. NEC covers wire and cable installed in factories, office buildings, hotels, motels, apartment buildings, residences, and all cables that pass through any floor, wall, or ceiling, or that travel in ducts, plenums, and other
Transmission Equipment
385
air-handling spaces. Each individual municipality, city, county, or state can decide whether or not it wishes to adopt the 1996 NEC as law. Local authorities have jurisdiction to enforce their own codes and they have the right to accept or refuse any installation in accordance with their local laws. One of the organizations local inspectors rely on to test wire and cable is Underwriters Laboratories (UL) in the United States and Canadian Standardization Association (CSA) in Canada. In the past, AWM cable was sometimes incorrectly used to wire buildings, never its intended use. AWM cable is intended for internal wiring of factory-assembled, listed appliances such as computers, business machines, ranges, washers, dryers, radios, and televisions. In some cases, AWM cable may be used for external connection. In these situations, the user should be aware that AWM cable temperatures and voltage ratings may differ from NEC ratings. The CSA/UL cable designation (and its meaning) would be one of the following: • CMP Cable meets CSA FT6 or UL 910; • CMR Cable meets UL 1666; • CMG Cable meets CSA FT 4; • CM Cable meets UL 1581, Sec. 1160 (Vertical Tray); • CMX Cable meets UL 1581, Sec. 1080 (VW-1); • CMH Cable meets CSA FT 1.
The switch office in the wireless network and other important telecommunications facilities have to use only properly rated cables for certain applications and critical locations where these cables are installed. For example, ventilation shafts must not have any cables that can produce smoke and endanger the safety of the people in the building. Similarly, cables should not spread the flame between different equipment rooms. The CSA flame tests (FT Rating) are defined in CSA C22.2 No. 0.3 as follows: FT 1 Vertical Flame Test (per C.S.A. C22.2 No. 0.3-92 Para 4.11.1) A finished cable shall not propagate a flame or continue to burn for more than one (1) minute after five (5) fifteen (15) second applications of the test flame. There is an interval of fifteen (15) seconds between flame applications. The flame test is performed in accordance with Para 4.11.1 of Canadian Standards Association (CSA) Standard C22.2 No.
386
Transmission Systems Design Handbook for Wireless Networks
0.3. In addition, if more than 25% of the indicator flag is burned, the test cable fails. FT 4 Vertical Flame Test Cables In Cable Trays per C.S.A. C22.2 No. 0.3-92 Para 4.11.4 is similar to the UL 1581 Vertical Flame Test, but is more severe. The FT 4 test has its burner mounted at 20° from the horizontal with the burner ports facing up. The UL-1581 Vertical Tray has its burner at 0° from the horizontal. The FT 4 samples must be larger than 13 mm (0.512 in.) in diameter. If not, then the cable samples are grouped in units of at least three (3) in order to obtain a grouped overall diameter of 13 mm. The UL-1581 Vertical Tray does not distinguish on cable size. The FT 4 has a maximum char height of 1.5m (59 in.) measured from the lower edge of the burner face. The UL-1581 has a flame height allowable up to approximately 78" measured from the burner. FT 6 Horizontal Flame and Smoke Test (per C.S.A. C22.2 No. 0.3-92 Appendix B) This test is for cables which must pass a Horizontal Flame and Smoke Test in accordance with ANSI/NFPA Standard 262-1985 (UL-910). The maximum flame spread shall be 1.50 meters (4.92 ft). The smoke density shall be 0.5 at peak optical density and 0.15 at maximum average optical density. 8.4.2
Digital Cross Connects
An access demarcation is a physical location through which content passes without alteration, whether that content be a telephone conversation, electric service to a building, or data traffic sent from one computer to another. The access demarcation is where the pipe (the medium through which content or service is delivered) owned by the service provider meets the pipe owned by the individual or business subscriber. Joining the two pipes together creates the connection necessary for delivery of content and services to a subscriber. The demarcation point is the point of interconnection between the telephone companys facilities and the wireless operators wiring. This could be fiber-optic equipment or T1/E1 lines. The demarcation point (demarc) should be located on the wireless operators side of the telephone companys protector, or the equivalent thereof in cases where a protector is not required. The access demarcation point acts as a property boundary; the service provider owns and is responsible for the equipment on one side of the demarc,
Transmission Equipment
387
and the subscriber (or building owner) owns and is responsible for the equipment that conveys and utilizes the service on the premises. Digital signal cross connect (DSX) products are used to connect one piece of digital telecommunication equipment to another, simplifying equipment connections and providing convenient test access and tremendous flexibility for rearranging and restoring circuits. DSX panels are available in numerous configurations and sizes to fit a wide variety of applications, but they all perform the same function. It terminates equipment and provides temporary jack access for centralized testing, cross connection, reconfiguration, and restoration of various digital circuits. It can be used in voice, data, or video networks for patching during equipment installations or breakdowns, network expansion, or traffic-pattern adjustments. At T1 (1.544 Mbps) or E1 (2.048 Mbps) rates, DSX-1 panels connect network equipment such as office repeater bays, channel banks, multiplexers, digital switches, loop carriers, and loop switches. At the T3 (44.736 Mbps) rate, DSX-3 panels provide terminations for the high-speed (DS3 rate) side of the M13 multiplexers. The DSX-3 supports networks operating at the DS3 rate and the STS-1 and STS-3 electrical SONET rates, and it is usually placed between network elements, such as fiber-optic terminals, multiplexers, broadband digital switches, digital crossconnect systems, and digital radio components, in a central office, CEV, hut or cabinet, or at the customers premises. The connecting block (also called a terminal block, a punch-down block, a quick-connect block, or a cross-connect block) is a plastic block containing metal wiring terminals to establish connections from one group of wires to another. Usually each wire can be connected to several other wires in a bus or common arrangement. There are several types of connecting blocks: 66 clip, BIX, Krone, 110, and so on. A connecting block has insulation displacement connections (IDCs), which means you do not have to remove insulation from around the wire conductor before you punch it down or terminate it. The BIX Cross-Connect System was developed in the late 1970s as a replacement to the 66-type punch-down block of the 1960s and early 1970s. BIX connectors, BIX mounts, and the BIX connecting tool are the building blocks on which BIX Cross-Connect Systems are based. BIX Cross-Connect Systems continued to evolve throughout the 1980s. In fact, with little or no change, the same BIX system used for voice applications now supports computer systems and data. In 1992 there was an introduction of wiring standards including the EIA/TIA-568 (CSA T529) Commercial Building Wiring Standard, which defined product performance and system design
388
Transmission Systems Design Handbook for Wireless Networks
criteria specifications for a universal wiring system capable of supporting voice and data applications. The Commercial Building Wiring Standard has since been revised and released in North America as ANSI/TIA/EIA-568-A (CSA T529-95). 8.4.3
DS1 Signal Termination
8.4.3.1 DS1 Signal Precompensation and Signal Levels
In most central offices, there is a DSX-1 cross-connect jackfield located between the channel bank and the office repeater, or between any two dissimilar elements in a whole end-to-end system. The jackfield serves as a maintenance test point for technicians and as a wire-wrap interconnection point between the two elements (Figure 8.17). If a transmit port must send a DS1 signal only 50 to 100 ft to the DSX-1 panel (typical in small equipment rooms), then it can send a normal DS1 signal waveform. Only 50 to 100 ft of cable capacitance will not affect that waveform much if the cable is good, so the waveform looks perfect at the DSX-1 panel when viewed on an oscilloscope. That is the objective, to get a perfect signal waveform presented to the DSX-1 panel. This standard signal is referred to as having a level of 0 dBdsx. A standard signal that has been attenuated by several hundred ft of cable might have a level of 3 dBdsx. Looking at the outbound DS1 signal going from the terminal equipment toward the span line, if it is perfect at the DSX-1 panel, then it gets minimally affected by the 50 to 100 ft of cable capacitance, so the signal still End office +3 −3
DSX-1
Office repeater
Channel bank, digital multiplex or other DS1 equipment This diagram uses one line to indicate one pair of wires
+3 −3
Figure 8.17 Transmission paths in the central office.
Primary protection
Line repeater
Transmission Equipment
389
looks fairly normal at the receiver. Often the assumption is made that the transmit cable distance to the DSX-1 is the same as the receive cable distance from the DSX-1 (on the same side of the DSX-1). If the cable distance to the DSX-1 is much longer, say 655 ft, then a normal signal is going to be attenuated and rounded off due to cable capacitance. In many pieces of equipment, a pre-equalizer is used in the DS1 interface on the transmit port. This pre-equalizer sharpens up the pulse edges to exaggerated amplitudes. This is launched down through the 655 ft of cable and the sharp edges become smoothed down (rolled off) due to cable capacitance. When it arrives at the DSX-1 panel, it should be exactly a perfect waveform (see Figure 8.18). When the signal comes into an office repeater port after passing 655 ft of cable from the DSX-1, it can be successfully received without errors. It might be possible to attenuate the signal even more without degradation, say from 700 or 800 ft, but errors in transmission might also appear (which would be unacceptable). It depends on the specific equipment and cable in use. Some types of inside plant equipment are only capable of transmitting and receiving a signal through 0 to 133 ft to the DSX-1 panel. It is Rising edge overshoot
Normalized amplitude
1.0
Nominal +3.0V
0.5 Standard T1 pulse mask 0
−0.5
324 ns nominal −450
−300
Figure 8.18 T1 pulse mask.
−150
Falling edge undershoot
0 150 300 Time (nanoseconds)
450
600
750
390
Transmission Systems Design Handbook for Wireless Networks
important to have each and every network element meeting the perfect waveform mask, which is an industry standard. The idea is to consider each direction of signal flow. Know what type of cable is in use and roughly what its attenuation characteristics are (22-gauge ABAM cable has a nominal capacitance of 14 to 16 pF/ft at 772 kHz). Also consider that 24-gauge cable has approximately 25% higher attenuation than standard 22-gauge. Know what signal levels should be present at each point and what the receive sensitivity is for each device. Many problems are traced to the use of cable that is not intended for T1 use. Office repeaters are somewhat more complex, since they have one set of capabilities on the equipment side (inside) and one set on the facility side (outside). A typical office repeater has a digital regenerator only on the receive side (receive from facility), and has enough sensitivity to regenerate a signal from a line repeater that is about 3,000 to 4,000 ft away. It has a typical inside cable specification for 0 to 655 ft, depending on the exact pre-equalization (Figure 8.19). 8.4.3.2 Conductors: Gauge, Shielding, and Other Factors
In most areas, 22 AWG is the preferred cable for most types of T1. In some metropolitan areas, limited 24 AWG and 26 AWG are used. Keep in mind the different attenuation characteristics of each. Larger sizes, such as 19 AWG, were never very popular for T1 because of the economics of buying DSX-1 0 dBdsx
−1
Office repeater
Line repeater 0 −21
Line repeater −30 0
Office repeater −20 +1
DSX-1 0
7.5
7.5 0 dBdsx
+1
−20
0
This diagram uses one line to indicate one pair of wires Figure 8.19 DSX signal level example.
−30
0
−21
−1
0
This example uses 30 dB as the maximum repeater section loss and an end section with less loss
Transmission Equipment
391
copper wire versus buying electronic systems. D-shield is a type of cable with one half of the pairs (transmit in one direction) organized into one D-shaped half of the cable, as viewed in cross section. This type of shielding effectively makes one-cable operation almost as good as two-cable operation in terms of cross-talk reduction. Filled cable contains a gel that prevents ground water from easily leaking into the cable. Although this cable has higher initial cost, it tends to produce longer cable life in rainy regions. Note that filled cable has a slightly different cable attenuation figure compared to unfilled cable. T1-class cables are commonly manufactured with multiples of 25-pair bundles. For single-cable operation, the standard choice is the east-west direction of transmission to be in one bundle and the west-east direction to be in a separate bundle. Furthermore, 104-pair cables are organized with two bundles for each direction and four pairs for the various voice-frequency circuits associated with T1 lines (order wire, fault locate). If the cable does not have this bundle organization, then it becomes necessary to test for potential cross talk (distortion factor, or D-factor) from every cable pair to every other cable pair to determine which pairs have the most coupling loss. Although modern line repeaters are very reliable, they are used in a hostile environment (temperature extremes, vibration, lightning, automobile damage, rodent damage) and as a result, line repeaters can fail any time. Poor cable splicing techniques might also lead to intermittent noise after installation. When channel banks or customer services have gone into alarm at the central office, how long the service disruption will remain is a function of the restoration technique. Most user companies have implemented spare span lines within each T1 trunk cable, which means that when a working span begins to fail, the traffic at the ends is rerouted to the spare span line. Some choose to do this manually (via patch cords) at each attended office. Others who place more value on availability of the span line for live traffic have chosen to implement automatic transfer equipment. The APS is located at each end of a span line facility and takes care of signal failure detection and automatic transfer. APS products generally fall into two categories: one-for-one switching (1:1) or one-for-N switching (1:N ). The term 1:1 means that there is one protection line for each normal working line. Another way of stating this is one standby line and one normal line. This is most commonly used in ring networks and in applications involving mixed transmission media (one fiber facility and one microwave facility, for example). In contrast, 1:N means that there is one protection line for a number (N ) of normal working lines. In practice, the integer N could be 1, and the upper limit on N may be dozens. The number N commonly
392
Transmission Systems Design Handbook for Wireless Networks
has a value between 2 and 8. This type is most commonly used on straight T1 cable carrier applications, especially for subscriber loop carriers. 8.4.4
Leased Lines and the Network Interface Unit
The NIU acts as the demarcation point where the customer premises equipment meets the public network. This is the equipment that the local telecommunications provider will install at every cell site where the wireless operator ordered the leased T1 circuit. The T1 signal arrives with a signal level of 0 dB to 16.5 dB. Although the signal jumps over a series of transformers, it is important to note that the smart jack, unlike a repeater, does not regenerate the signal. NIUs are sometimes referred to as smart jacks in that they have intelligent functions that are necessary for the demarcation point between the telecommunications company and the customer premises. NIUs bear a strong resemblance to terminating office repeaters that do not feed span power to the facility. They also bear a resemblance to T1 CSUs. Different NIUs have different feature sets, but diagnostic loopbacks are most commonly found.
8.5 Grounding 8.5.1
Earth Grounding Basics
Grounding can be described as the science of obtaining a low-resistance path for the dissipation of current into the Earth. There are different methods to obtain a ground, but first a discussion of grounding fundamentals is crucial to understanding and designing a grounding system. Grounding is the physical bonding or connection of equipment by a conductor to the Earth. Without a proper low-resistance ground, standard protection devices such as breakers, or transient voltage surge and lightning protection systems, are rendered ineffective. Most communications equipment manufacturers may void their equipment warranties at sites where the ground system performance does not meet their explicit Earth grounding requirements, typically 5 ohms or less. Good grounding also has other benefits, such as enhanced personnel safety, reduction in system noise, and protection from lightning, unwanted voltages, currents, and power surges. Protection for cable and wireless telecommunications equipment begins with a stable, low-impedance Earth ground, which can be effectively assured with an electrolytic Earth grounding system. Standards for equipment performance mandate installation and maintenance of a reliable, lowresistance Earth ground. These standards often cannot be met and certainly
Transmission Equipment
393
cannot be assured for the long term by traditional grounding methods. With appropriate designs, active electrolytic grounding systems can provide excellent results that are superior to those of traditional driven rods. The Earth is composed of many materials that are variously good and poor conductors of electricity, but the Earth as a whole is considered a good conductor. For this reason and as a reference point, the Earths potential is assumed to be zero. When an object is grounded, it too is thereby forced to assume the same zero potential as the Earth. If the potential of the grounded object is higher or lower, current will pass through the grounding connection until the potential of the object and Earth are the same. The Earth electrode is that connection path from equipment to the Earth. The resistance of the electrode, measured in ohms, determines how quickly, and at what potential, energy is equalized. Hence, grounding is necessary to maintain an objects potential the same as that of the Earth. Grounding is critical for the following reasons: The first is human safety. Second, it is meant to protect equipment from voltage surges and transients. A low resistance ground will keep equipment at or near Earth potential, reducing any voltage difference between equipment and Earth. This will prevent an accident or fatality during human contact. Equipment damage (i.e., sensitive telecommunications equipment) from surges caused by lightning and so forth can result in the loss of millions of dollars in damages and downtime. Equipment failures caused by inadequate grounding hinder performance and productivity which increases costs of operations. So, a grounding system design has two main objectives. First, it is meant to provide personal safety and equipment protection by providing a low-resistance path to safely dissipate any unwanted charges or potentials. Second, it is meant to provide a reference point approximately equal to the potential of the Earth for sensitive equipment. To be effective, a grounding system must be stable and reliable in all adverse environmental conditions, be maintenance free, and have a long life expectancy with no recurring costs. A vital element in all electrical system designs, Earth grounding is important for the long-term successful operation of equipment and safety of personnel. Providing a low-impedance path to Earth for high currents from lightning strikes or power surges is an important requirement. Maintaining good Earth grounding is also necessary to ensure all noncurrent carrying metal parts are at ground or zero potential for personnel safety. Grounding also provides a reliable path to Earth for the low currents from radio frequency interference (RFI) or static that will build up on system enclosures and signal return paths. Also, the neutral currents from any three phase imbalances and harmonics are sent to Earth through the grounding system.
394
Transmission Systems Design Handbook for Wireless Networks
Once installed, grounding systems are expected to provide a stable, low-impedance path to Earth. The system must perform reliably and should be cost effective from installation throughout useful life of the site. Generally, attempts to meet these expectations are in the form of copper-clad steel-driven rods, connected to a bar or equipment via mechanical methods. However, sensitive equipment requires tighter standards, often including achievement of a specific low resistance level of 5 ohms or less. In order to ensure that these standards are met and that the Earth ground system functions properly, thorough planning, site evaluation, testing, and specific designs are critical. The single most important element of the site evaluation is accurate soil resistivity measurement. For raw land sites, measuring soil resistivity is the first step in the planning process. With this data, a soil resistivity profile is built. This profile is the key for an accurate ground grid design and is the only way to ensure predictable grounding system performance. A soils profile is the collection of soil resistivity levels at 5-, 10-, 20-, and 40-ft depths, sometimes up to 100 ft. This data is gathered at three or four different locations on the site. This allows the design engineer to determine the length, configuration, and quantity of rods required to achieve the specified ground system resistance. An accurate soil resistivity profile is also important, due to the wide variation of resistivity at different depths that is often encountered. For example, soil resistivity levels at 5 and 10 ft may be different than levels at 20 and 40 ft. Should the soil resistivity be less at shallower depths, an L-shaped configuration installed in a trench may be the best solution. The key to ground grid design and its effectiveness is the soil resistivity data derived from on-site four-point testing. Shortcutting the planning stage is often the reason many grounding installations are found lacking upon project completion with failure of acceptance testing. Both equipment manufacturers and some regulatory agencies take serious interest in the installed ground grid resistance measurements before operational approvals are granted. Soil resistivity testing allows a proper design that will save time, money, and effort. At colocated and existing sites an extra step should be taken. The installed grounding system must be tested and evaluated. This data will be used to determine if a grounding system upgrade is required to protect the new equipment. Lack of planning and poor or nonexistent soil resistivity testing will result in unpredictable ground system performance and ineffective equipment protection. Site and ground grid testing is one step that has been the most often undervalued in the past.
Transmission Equipment 8.5.2
395
Ground System Design Fundamentals
The characteristics of soil determine the design and physical construction of a grounding system if we are to achieve a desired conductivity. This includes grounding electrodes, electrode spacing, and placement. The single most important characteristic that we are concerned about is the soils conductivity or ability to conduct electricity, inversely called soil resistivity. Soil resistivity testing will determine how nonconductive the soil is, and ultimately the grounding system layout, to achieve a specific Earth-ground resistance. Factors that affect soil resistivity are its moisture content, its electrolyte and metal content, and environmental changes in temperature. Soil or Earth resistivity is the soils electrical resistance to the flow of dc and ac currents. The most common unit used is the ohm-meter, which refers to the resistance measurement between opposite faces of a cubic meter of soil. Theoretically, the Earth-ground resistance of any ground system or electrode, R, can be calculated using the general resistance formula: R = r (L /A) where r = resistivity of the Earth (ohm-meter) L = length of the conducting path (meters) A = cross-sectional area of the path (meter2) Soil resistivity is a proportionality constant that relates the resistance of a ground system to the length of the conducting path and its cross-sectional area. It is important to measure soil resistivity since resistivity can vary widely in different soil mediums. For example, typical surface soils can vary in resistivity from a range of 100 to 5,000 ohm-cm. Moreover, knowing what the resistivity is in surface soil is necessary for an effective grounding design. Typically, soil resistivity is measured using the Wenner four-pin electrode method using a ground resistance meter. Four metal pins are placed in contact with the ground in a straight line and equally spaced. A constant current is then injected through the ground via the tester and the outer two electrodes labeled C1 and C2. The potential drop is then measured across the inner two electrodes labeled P1 and P2. The meter provides a direct ohm reading that is used in the following formula to determine soil resistivity: r = 1915 . ×R×A
396
Transmission Systems Design Handbook for Wireless Networks
where 191.5 = constant; r = soil resistivity in ohm-cm; R = ground resistance meter readout in ohms; A = distance between electrodes in feet. The resistivity calculated is the resistivity of the soil between the surface of the ground and a depth equal to the pin spacing. Most soils naturally contain varying amounts of electrolytes that conduct electricity. As a result, the addition of moisture will enhance its conductive properties. However, the addition of moisture to some soils such as granite, sandstone, and surface limestone will have little or no effect in reducing the resistivity. In general, however, the greater the moisture content in soil the lower the resistivity. Temperature, like moisture, can have a significant impact on resistivity. The soil resistivity does not vary much with temperature until the temperatures reach freezing conditions, that is, 32°F; at this temperature the moisture in the soil will freeze up, and the resistivity will increase. The amount of salts in the ground also influences a soils resistivity. In general, the more salts or electrolytes that a soil contains the lower the resistivity. 8.5.3
Types of Grounding
There are several types of grounding systems used in industry today. Some of the most common include driven rods, water pipes, chemical wells, ufer grounds, and electrolytic rods. Driven Rods
Historically, copper-clad steel-driven rods have been the common method of Earth grounding. Simple, relatively easy to install, driven rods have been the mainstay for most grounding applications. A rod or series of rods are driven into the Earth, then bonded to each other with various mechanical connectors. Damage during installation, difficult soil conditions, corrosion, and loosening of bonding connections contribute to the ineffectiveness of driven electrodes over time. Utilizing standard rods for sensitive, critical communications equipment is an invitation to degraded performance and costly down time. Some of the drawbacks of using driven rods include the following: • Easily affected by the environment, aging, temperature, and moisture;
Transmission Equipment
397
• Resistance increases steadily with age; • Easily damaged during installation. Scratches expose the steel metal
to the environment, which will make it susceptible to corrosion.
Driven rods are inexpensive and are adequate for a short time in good soil conditions; however, they will eventually fail in service. Water Pipes
Water pipes or water mains are used as Earth electrodes. Some of the drawbacks of using a water pipe ground are as follows: • They are difficult to test and impossible to maintain; • Plastic inserts or o-rings destroy the circuit integrity; • Cold water pipes produce condensation, which encourages corrosion.
Water pipes should never be used as a single ground source. They are, in fact, an unreliable grounding source which can be destroyed by a simple plumbing upgrade. Instead, water pipes should be used in conjunction with driven rods or a ground grid system in compliance with NEC 250 code. Chemical Wells
Chemical wells are ground wells that are filled with highly conductive chemicals, connected to grounding systems with copper rods. Many old-fashioned chemicals are usually hazardous to the environment (i.e., copper sulfate, magnesium sulfate) and are restricted by the EPA. Ufer Grounds
Ufer grounds consist of copper-wire grids that are incorporated into the building foundation concrete during construction. Ufer grounds are impossible to test and maintain since the conductor, typically 4/0 stranded cable, disappears into the foundation. As a result, time and gradual removal of moisture can cause changes in foundation integrity and ground resistance. Electrolytic Rods
The electrolytic grounding systems should be considered an excellent alternative to driven rods. Their cost effectiveness and proven performance have been demonstrated over the years. Telecommunications equipment operation is extremely sensitive to power quality variations, causing effectiveness
398
Transmission Systems Design Handbook for Wireless Networks
of Earth ground to be vitally important. Electrolytic rods are 100% copper tubes filled with a combination of natural Earth salts called calsolyte. To be effective, active electrolytic rods must have holes drilled near the top and bottom of the rod. The holes near the top are referred to as breather holes and are an inlet for air to enter. The hygroscopic salts in the tube absorb the moisture in the air and form an electrolytic solution. This solution then is deposited into the backfilled soil environment by weeping out through the bottom holes, creating electrolytic roots. The electrolytic roots produced further lower the resistance of the ground by ionizing the surrounding soil. Electrolytic rods are reliable, since they are protected from the corrosive soil environment and are not as prone to corrosion attack, which is a problem with driven rods and other systems. 8.5.4
Grounding for Wireless Cell Sites
With the large capital investment in the extensive network of cellular sites, expanding numbers of PCS sites, and switching facilities, the reliability of the services delivered becomes of ultimate importance. Grounding system design enabled by planning and testing is required to provide effective protection and equipment performance. A system designed for long-term performance makes sense from both an operational and cost perspective. Minimal expense for soil resistivity testing to support the design process, accompanied by postinstallation testing, ensures a system that will meet specified requirements. The design process for a grounding system begins with a site survey of the installation area. A site survey must include soil resistivity analysis at several depths, relevant site plans, topography analysis, and a boring core sample if available. The site survey will determine if any physical barriers such as rock, high-resistivity soil, or power lines will affect the Earth-ground resistance in the installation area. Once the information is obtained a design can be initiated. Grounding for PCS sites requires knowledge of grounding fundamentals. For example, using 10-ft electrolytic rods, the rods must be spaced at least 20 ft apart (2 × L) in order to take full advantage of the rods sphere of influence, etc. Ground conductors must be buried below the frost line and connected to the rods in a counterpoise arrangement to ensure system integrity. And, finally, the PCS grounding system is an example of a grounding system that is stable, reliable, safe, maintenance free, and long lasting with no recurring costs. It is recommended to use electrolytic rods in any grounding system. Limited real estate in most urban sites and constraints from existing grounding systems is what most designers face in the rapidly expanding PCS
Transmission Equipment
399
market. The limited real estate combined with low resistance specifications make grounding these sites very challenging. Electrolytic grounding systems can be used effectively and efficiently to overcome these challenging conditions. The sensitive wireless infrastructure equipment and lightning protection is being located on rooftops in between and around the existing heating, ventilating, and air-conditioning (HVAC) equipment and elevator penthouse. Electrolytic single-rod, single-point design grounding system for these restricted sites has been utilized on hundreds of sites (Southern California, for example). Monopole installations are springing up in crowded metropolitan areas, planters, and in highway right of ways. The base station equipment may be located in underground parking structures, church steeples, gas station signs, or leased wall spaces in buildings giving a whole new dimension to tenant space. Custom-designed rod lengths can address issues of limited real estate. Improved effectiveness over time helps guarantee long-term equipment safety and performance to a degree not possible with traditional driven rods. Where land and access are restricted, custom-designed products and services have to be used.
8.6 Power and Battery Backup 8.6.1
ac Power
There are three distinct ac uninterrupted power supply (UPS) topologiesstandby, line-interactive, and on-line. Standby UPS, also known as off-line UPS, consists of a basic battery powerconversion circuit and a switch that senses irregularities in the electric utility. The load is connected directly to the utility power and power protection is only available when there is an outage. However, some standby UPSs include suppression circuits or power line conditioners to increase the level of protection they offer. Line-interactive UPSs offer a higher level of performance by adding voltage regulation features to conventional standby designs. Like stand-by models, line-interactive UPSs protect against power surges by passing the surge voltage to the load until it hits a predetermined limit. At this point the unit switches to battery, but only after passing most surges through to the load. Line-interactive units can provide moderate protection against highvoltage spikes and high-frequency transients, although they do not provide complete isolation for the load. On-line ac UPSs use double conversionac/dc and dc/acproviding complete isolation from most types of power problems.
400
8.6.2
Transmission Systems Design Handbook for Wireless Networks
dc Power
Transmission equipment is usually required to operate off the −48V or +24V power supply. Battery backup must provide sufficient reserve capacity to allow for the uninterrupted operation of equipment for at least 4 hours for the cell sites and at least 8 hours for the switch office. Customers may specify more or less capacity, as local conditions such as travel time, standby generation of power, and power reliability may vary with location. The equipment must be capable of successfully operating within published specifications and satisfy all required operational specifications with power variations that do not exceed −56V dc or fall below −40V dc at any given time. In newer and more advanced power systems, alarms will notify the user when the voltages get to within 20% of these extreme values. In the case of battery drain, it will signal when the equipment is within 1 hour of losing backup power. Both the constant voltage and the constant current charge methods are suitable for charging recombinant lead-acid batteries. Of the two, the most common method used is simple constant voltage, which allows the battery to seek its own current level in overcharge (or float charge). For cyclic applications in which batteries undergo regular charge and discharge, the charge voltage per cell ranges from 2.4 to 2.5V. Although a float charge of 2.25 to 2.35V is advisable for batteries in standby usage, it is important to consult the battery manufacturers literature for proper charge-voltage recommendations. The following general design procedure is used to calculate necessary capacity for the battery and the rectifiers: Load current = power consumption/system voltage Battery capacity (4 hrs) = (installed power/system voltage) × 4 × 1.2 Required battery charging current = 0.1 battery capacity (4 hrs) Rectifier rating = load current + required battery charging current The last formula shows that the rectifier has to be dimensioned to be able to recharge empty batteries and at the same time carry the full load of the system indefinitely. Ideally, the system should be easy to upgrade and expand (modular approach) to accommodate changes and growth. In wireless systems, valve-regulated lead acid (VRLA) batteries (gel-cell) are usually used. These batteries are sealed and maintenance free throughout battery lifetime; they do not require any forced ventilation and can be placed in the equipment room or outdoor cabinet. The rectifier rack can also house the batteries for small-size systems.
Transmission Equipment 8.6.3
401
Batteries
Power loss in a wireless communications system can be disastrous for a wireless operator business and also for an organization relying heavily on such service. The battery is by far the most unreliable part of any system that depends upon it. This is true because if any one of many external variables is incorrect, the natural deterioration process within each cell is accelerated. When the utility power fails and the battery instantly assumes the critical load, the choices and investments that have been made in emergency power equipment will be realized. The deterioration accelerating factors include the following: • Too high cell temperature; • Too high cell float voltage; • Too low cell float voltage; • Too many discharge cycles.
The first step to minimize the problem is to become aware of and then track the critical variables with an automated continuous-monitoring system. The second step is to correct and optimize any of these critical variables where possible. The third step is to immediately notify the user if the critical measured parameters are outside of defined limits. The final step is to automatically identify unfavorable trends and notify the users of a developing problem before it becomes serious. Historically, backup power for telephone networks has been provided using flooded batteries. Over the past several years, there has been a trend toward other types of batteries such as valve-regulated batteries (also called gel-cell batteries). The term valve-regulated identifies a battery equipped with mechanical safety vents that can open under excessive overcharge. Generally speaking, using VRLAs, which require less attention, offers remote monitoring and reduces the number of sites that must be visited for troubleshooting and for routine maintenance. This, in turn, reduces the cost of these activities. Flooded batteries are generally too large to fit into the limited space available for wireless site applications. Service and maintenance of flooded batteries in remote sites would also be very expensive, since the service technicians would have to visit many sites on a regular basis in order to check electrolyte levels. VRLA batteries do not give off any gases under normal conditions, so special ventilation provisions are not needed. Relative to flooded batteries, VRLA batteries also offer excellent power density, do not require any regular electrolyte
402
Transmission Systems Design Handbook for Wireless Networks
maintenance, and can be mounted in any orientation. These characteristics make VRLA batteries ideal for wireless applications. Following are the installation and maintenance rules for VRLA batteries: 1. Heat reduces battery life. Batteries rely on chemical reactions to generate electricity, and heat speeds up those reactions. Operating temperatures outside the range −15°C (5°F) to +50°C (122°F) should be avoided for float and standby applications. To obtain the rated service life, operating temperature should be kept at 20°C (68°F); a battery will lose nearly half of its life expectancy with each 10°C temperature rise above its norm. 2. In an uncontrolled environment it is important to use temperaturecompensated float voltages, which will decrease automatically whenever battery temperature rises, thereby keeping the batteries within their specified operating temperature. 3. Lead-acid batteries, when integrated with the equipment, should be installed on the lowest level of the enclosure (e.g., outdoor cabinets and RF microcells). 4. Sufficient air space between batteries must be provided, usually 0.2 to 0.5 in around each battery. 5. Batteries should not be installed near any equipment that emits sparks. Under normal operating conditions, hydrogen gas and oxygen are produced, but in the case of abuse (e.g., overcharging, undercharging, or exposure to extreme heat), some batteries can generate ignitable gases as well. 6. Adequate ventilation in confined spaces is not absolutely required but it is recommended. 7. Batteries must be securely fastened if the installation could be subject to heavy vibration or mechanical shock. 8. If two or more groups of batteries are connected in parallel to increase battery string capacity, they should all present the same impedance to the load. This means that all the wires, cables, and bus bars must have the same loop resistance. 9. Ripple current flowing in the battery should never exceed 0.1°C. Higher ripple currents can cause heat generation and reduce charge efficiency, thus reducing battery life expectancy. 10. Generally, it is not advisable to fast-charge sealed lead-acid batteries. Unless regulation is perfect, fast charging at rates exceeding
Transmission Equipment
403
33% of the batterys ampere-hour capacity can easily outstrip the batterys oxygen recombination capability and may generate temperatures high enough to dry out the electrolyte or damage the integrity of the batterys plate separators. 11. Batteries of various capacities, of different ages, from several manufacturers, or that were taken from other applications, should never be used together since they will not have matching performance characteristics. The best practice is to install only fresh, identical batteries. 12. Lead-acid batteries should never be stored in a discharged state. 13. Batteries should never be dismantled or incinerated; instead, they should be recycled. 14. Up-front equipping power plants with the remote monitoring and control system is a very smart investment, since it reduces both the costs and frequency of personnel visits to the cell sites. They usually generate alarm messages but also can provide vital data about power plant condition. Some wireless operators keep replacing cell-site batteries as a part of regular maintenance every 2 years. This is especially the case when there is no control over the operating temperature and batteries have a short life expectancy. Although air conditioners and heat exchangers are usually used for cooling cell sites in the wireless industry, there is an interesting new product designed to maintain small telecommunications battery compartments at optimum temperatures. PCMs are substances that absorb large amounts of thermal energy during the day by changing phase from solid to liquid. At night when temperatures are cooler, the PCM changes from liquid to solid, releasing the stored energy and becoming ready for the next cycle. This product is ideal for places where there is a big difference between day and night temperature (e.g., desert areas). Environmental control is obviously desirable for the electronics as well as the batteries.
8.6.4
Solar Energy
Very often there is confusion about the various methods used to harness the suns abundant and clean energy. Energy from the sun can be categorized in two ways: (1) energy in the form of heat (or thermal energy), and (2) energy in the form of light energy.
404
Transmission Systems Design Handbook for Wireless Networks
Solar thermal technologies use the suns heat energy to heat substances (such as water or air) for applications such as space heating, pool heating, and water heating for homes and businesses. There are a variety of products on the market that utilize thermal energy. Often the products used for this application are called solar thermal collectors and can be mounted on the roof of a building or in some other sunny location. The suns heat can also be used to produce electricity on a large utility scale by converting the suns heat energy into mechanical energy. Photovoltaics (PV) is a technology often confused with solar thermal and is in fact what many people mean when they refer to solar energy. Photovoltaics (photo = light, voltaics = electricity) is a semiconductor-based technology (similar to the microchip) that converts light energy directly into an electric current that can either be used immediately or stored, such as in a battery, for later use. Photovoltaics are solar cells that produce electricity directly from sunlight and they are usually made of silicon. The cells are wafer-thin circles or rectangles, about three to four inches across. Solar cells operate according to what is called the photovoltaic effect. In the photovoltaic effect, bullets of sunlightphotonsstrike the surface of semiconductor material, such as silicon, and liberate electrons from the materials atoms. Certain chemicals added to the materials composition help establish a path for the freed electrons. This creates an electrical current. Through the photovoltaic effect, a typical fourinch silicon solar cell produces about one watt of direct current electricity. PV panels or modules are very versatile and can be mounted in a variety of sizes and applications: for example, on the roof or awning of a building, on roadside emergency phones, or as very large arrays consisting of multiple panels or modules. Currently they are being integrated into building materials (such as PV shingles, which replace conventional roofing shingles). Many remote sites use photovoltaics, which is cost effective and practical now. Photovoltaics is generating power for on- and offshore traffic-control systems, crop irrigations systems, bridge corrosion inhibitors, and radio-relay (microwave radio) stations. It is also providing electricity to remote cabins, villages, medical centers, and other isolated sites where the cost of photovoltaics is less than the expense of extending cables from utility power grids or producing diesel-generated electricity. In remote wireless cell sites, solar energy can be used to replace the electrical-utility-based power supply entirely or just as an additional power that will help during the peak hours and reduce the electrical power consumption. Either way, for every project a very extensive study must be done in the feasibility and cost efficiency of such a system.
Transmission Equipment
405
8.7 GPS Antennas Many sites in wireless networks contain equipment that have timing provided by the GPS. Cell sites could have GPS timing (as is the case in CDMA networks) as well as switch site or NOC for its BITS clock. An optimum location for placing GPS antennas provides an unobstructed view of the sky in a hemisphere around the antenna base. Also, there should be no obstruction above a 10° vertical angle measured from the horizontal plane running through the antenna base and a line of sight from the center of the antenna to the top of an obstruction. The maximum aggregate of all blockages above the 10° mask angle, such as mountains, buildings, and so on, should not exceed 25% of the surface area of a hemisphere around the GPS antenna. The blockage should not be in one continuous quadrant of the hemisphere, and each continuously blocked quadrant should be less than 12.5% of the spheres surface area. Self-support towers or guyed towers will not generally block the GPS signal. A high voltage line will not cause interference with GPS signal reception. For self-supporting and guyed towers it is recommended that the GPS antenna be mounted approximately 12 in from the tower structure. The GPS antenna should be located between approximately 14 and 20 ft from the foot of the tower, so as to be above and clear of the cable and waveguide bridge between the tower and the shelter. On a monopole tower site, the antenna should be located on the shelter at a suitable location. The GPS antenna should not be at the highest point in the installation, but at the same time it should be within the protective cone of any grounded structure such as an antenna tower or ancillary lightning rod. The protective cone is the area between the circumference of a circle with a 150-ft radius, which tangentially touches the ground and the side or top of a grounded structure and the ground or structure. The cable shield must be grounded and bonded to the tower. If the GPS antenna cable run is more than 150 ft (leasehold, roof top sites), multiple grounding bonds are to be employed.
8.8 Quality and Reliability Issues 8.8.1
Quality Assurance
Quality control is the establishment of quality standards for all materials, equipment, and services necessary for the successful execution of the project, followed by systematic measurement of quality actually achieved, in comparison with the standards, and corrective action where called for. Quality
406
Transmission Systems Design Handbook for Wireless Networks
standards should be established jointly and be mutually agreed upon by the client and the equipment or service provider. The client may choose to set standards for some aspects and incorporate them in the statement of requirements, or may choose to seek the advice of the consultants or other engineers. Standards for some aspects may be regarded as options that can be decided during the project, when the client may be better informed to make a tradeoff. Procedures and responsibility for handling quality control should be specifically addressed and consideration should be given to the independence of the quality control function from other project functions. In particular, with respect to construction projects, the engineer should establish the necessary arrangements for the required technical review of construction for conformance with design. The quality control of the project management function itself should also be considered. An overall quality control program would normally include design checks, the quality control of the design management and construction management functions, equipment and materials testing and test certificates, inspection, deficiency lists, and audits. 8.8.2
First Article Inspection
The first article test (FAT) is conducted to verify that the first article samples meet the performance requirements, during and after all specified environmental, durability, and endurance conditions. A first article sample is a production component(s) submitted as being representative of a specific process using production tooling, equipment, methods, techniques, standards, personnel, and controls. First article inspection is a verification that the first article samples manufactured using the normal production process (planning, technical and work instructions, material processing systems, and controls, tools, fixtures, test equipment, and personnel proficiency) will produce a component in compliance with all requirements. Examples of such requirements are dimensional characteristics, material content, process, capability, and performance. Facilities to be utilized in performance of the first article test, if known at the time of the request for quote (RFQ) response, should be identified in the quotation, giving the facility name, location, contact, and a phone number. Also, first article test costs are individual costs for the entire first article test program, including hardware, fixtures, equipment, test procedure development, itemized cost by test parameter, and test report preparation. These costs should be quoted separately from the cost of production hardware, and may be a part of the RFQ process as well. Test plans are written by
Transmission Equipment
407
the supplier or independent test laboratory and may be amended by the customer. A test incident or failure is deemed to have occurred when it is noted that the test unit(s) do not conform to the specified limits or specification requirements. The customer reserves the right of final determination as to the occurrence of a test incident or failure. The customer usually reserves the right to witness not only the first article tests but any other factory acceptance testing on any other quantity of the equipment being manufactured and shipped to that customer. 8.8.3
Factory Acceptance Testing
During the entire duration of the project, the customer has the right to request to witness acceptance testing on their equipment during the manufacturing process. This is very useful during lengthy projects, where there is a doubt that the quality standards could be changed over the projected period of time. The customer will definitely request to witness the acceptance testing in the case when the first shipment of the equipment had some problems that could be directly assigned to the manufacturing process and quality assurance (QA). In most cases when dealing with the reputable equipment supplier, factory acceptance testing will be limited only to the first article inspection. In case of any problems in the field, the customer can request the supplier to include additional tests that may narrow down problems and point out potential solutions. 8.8.4
Equipment Reliability
Reliability predictions are used to compute the predicted failure rate or mean time between failures (MTBF) of the equipment, which is usually expressed in terms of hours. MTBF for an existing product can be found by studying field failure data. For a new product, estimation of MTBF is required even before any field data is available. The most popular models are Bellcore Reliability Prediction Procedure (TR-332) and MIL-HDBK-217 for electronic systems, and NSWC-94/L07 for mechanical reliability prediction. Bellcore TR-332 and MIL-HDBK-217 are the most common models employed for performing reliability prediction analysis calculations. These models are well known and are accepted standards developed over several years that include mathematical equations for determining the failure rate (or MTBF) of various components and systems.
408
Transmission Systems Design Handbook for Wireless Networks
A fault tree analysis (FTA) is a deductive, top-down method of analyzing system design and performance. It involves specifying a top event to analyze (such as a blown fuse or failed trunk), followed by identifying all of the associated elements in the system that could cause that top event to occur. Failure mode and effects analysis (FMEA) is a bottoms-up approach to analyzing system design and performance. Using this standard, each component or functional block in the system is studied as to how it can fail. The effect of each failure mode and its severity on the overall function of the system is taken into account. The likelihood of the occurrence of such failure and its detection is also considered. Maintainability analysis provides calculated information regarding various aspects of maintenance. The goal of performing a maintainability analysis is to determine the amount of time required to perform repairs and maintenance tasks. In other words, if a system does fail, how long will it take to fix it? Maintainability analyses help compute the MTTR of the system. The time required for repairing the system can be used to estimate and optimize its availability. Maintainability analysis can also look at preventive maintenance procedures that help keep the system running in an optimum state. The Bellcore reliability prediction model was originally developed by AT&T Bell Labs. Bell Labs modified the equations from MIL-HDBK-217 to better represent what their equipment was experiencing in the field. The main concepts of MIL-HDBK-217 and Bellcore were very similar, but Bellcore added the ability to take into account burn-in, field, and laboratory testing. This added ability has made the Bellcore standard very popular with commercial organizations. The most recent revision of the Bellcore Reliability Prediction Procedure, TR-332, is Issue 6, dated December 1997. MIL-HDBK-217 supports the ability to perform a parts count or part stress analysis. In Bellcore, these calculations are referred to as calculation methods. There are ten different calculation methods, each designed to take into account different information. This information can include stress data, burn-in data, field data, or laboratory test data. MIL-HDBK-217 is the original standard for reliability. It is used by both commercial companies and the defense industry and is accepted and known worldwide. MIL-HDBK-217 includes the ability to perform a parts count or part stress analysis. A parts count analysis provides a simpler reliability math, and is normally used early in a design when detailed information is not available, or a rough estimate of reliability is all that is required. A parts stress analysis takes into account more detailed information regarding the components, and therefore offers a more accurate estimate of failure rate.
Transmission Equipment
409
When selecting which model to use for the reliability prediction of electronic components, it helps to compare the two most common models and understand their differences. Based on the requirements of the reliability prediction analysis, it may be found that one model is more applicable under certain circumstances than the other. Bellcore is primarily accepted in the United States. Although its popularity is growing internationally, it has not been completely embraced by the international community. The Bellcore model was designed to focus specifically on telecommunications. MIL-HDBK-217, however, is much broader in scope. MIL-HDBK-217 is geared toward both military and commercial equipment and has no specific market focus. The basis of the Bellcore and MIL-HDBK-217 calculations is very similar. However, when comparing the calculations of the two models, it is often found that the Bellcore calculations are more optimistic than MIL-HDBK-217. In addition, the Bellcore calculations generally require fewer part parameters for components. The Bellcore model also includes the additional capability of considering burn-in data, laboratory test data, and field data. This feature is extremely helpful in calculating failure rates that are based on historical data rather than simply stress data. In addition, burn-in data is used to quantify the first-year multiplier, which is an indication of infant mortality. The Bellcore model was designed to calculate failure rates in failures in time, or FITs. This value is expressed as failures per billion hours. This differs from MIL-HDBK-217, which was designed to calculate failure rates in failures per million hours. Bellcore supports additional miscellaneous part types in its model. Bellcore provides models for gyroscopes, batteries, heaters, coolers, and computer systems. These part types are not available in MIL-HDBK-217. MIL-HDBK-217 also has special part types that it supports. MIL-HDBK-217 provides models for printed circuit boards, lasers, SAWS, magnetic bubble memories, and tubes that are not supported by the Bellcore model. Because Bellcore was initially designed for use in the telecommunications industry, the operating environments that Bellcore supported were very limited. Initially, Bellcore only supported three different variations of ground-based environments. Bellcore is rapidly evolving, however, and in its most recent issue, additional operating environments of Airborne, Commercial and Space, Commercial were made available. Bellcore makes the assignment of quality levels very simple, and the current issue supports four standard quality levels. These quality levels are identical for all component types, and are simply based on some generalities regarding the origin and
410
Transmission Systems Design Handbook for Wireless Networks
screening of components. MIL-HDBK-217 offers a different approach to quality levels; those that are used differ from one part type to another. Rather than having a simple classification of general quality levels, the quality levels for components in MIL-HDBK-217 are derived from specific data that is component dependent. Therefore, the quality levels for resistors are different from the quality levels for semiconductors and the quality levels for semiconductors are different from those for integrated circuits. The quality levels for each part type were designed specifically for that classification of component. 8.8.5
Environmental Specifications
It is important to keep in mind that every project is different, and the environmental requirements for the transmission equipment used could vary from country to country and from project to project. Therefore, there is a chance that the equipment will have to be tested and certified to make sure it complies with the local standards and regulations, as well as to make sure that it will survive harsh environmental conditions. In cases in which the MW radio is designed as an indoor-outdoor configuration, the MW radios ODU will be used in an outdoor environment. External temperatures could vary between −40°C and +50°C. In that case, the enclosure should be internally equipped with suitable insulation material in order to minimize temperature variations inside the enclosure due to surface heat absorption, and have suitable temperature and environmental (cooling and heating) control. The ODU enclosure itself should conform to North American NEMA 4 or equivalent European IEC 529 standards (IP55, Protection from Water, Snow, Dust Entry) and, if possible, testing should be done by a third party. The equipment supplier should supply independent lab test results and its detailed environmental testing specification. The ODU will be used for housing electronic equipment and should be shielded from dust, moisture, and any external contaminants entering the ODU. The ODU should also be corrosion resistant. For indoor applications where HVAC is provided, the transmission/microwave equipment should indefinitely satisfy all manufacturers published specifications within a temperature range of 0°C to +45°C (ideally −10°C to +45°C). For both outdoor and indoor applications, the equipment should indefinitely satisfy all manufacturers published specifications within a humidity range of 0% to 90% noncondensing relative humidity. All microwave equipment must satisfy manufacturers published specifications while operating in an altitude pressure density range of −100 to +4,000m (above
Transmission Equipment
411
mean sea level pressure). Microwave radio specifications must be guaranteed over the entire temperature range. 8.8.6
Network Equipment Building Standard
The Network Equipment Building Standard (NEBS) was developed to safeguard networks and networking equipment. The rigorous standards cover all aspects of physical design and reliability. For decades, telephone companies have relied on NEBS equipment to maintain telephone service even in the case of disasters such as earthquakes or fires. NEBS also guarantees that networking products can work well together. NEBS compliance is a crucial feature of carrier-class networking equipment; therefore many telecommunications companies worldwide require that new networking equipment comply to NEBS (or similar) standards. NEBS standards are developed by Bellcore, and the most commonly used standards are fully described in these Bellcore documents: • Physical Design for Reliability (Bellcore Document FR-78). Physical
requirements include equipment dimensions. Products must fit into standard telecommunications company line-up. Also specifies weight, power dissipation, cable distribution, and resistance to environmental hazards, fire, mechanical shock, vibration, and earthquake.
• Component Design for Reliability (Bellcore Document FR-357). Reli-
ability requirements cover component reliability, device quality levels, qualification tests, manufacturer qualifications, lot-to-lot quality, corrective actions, component storage and handling, field reliability analysis, and redundancy modeling.
• Physical Protection (Bellcore Document GR-63). Protection require-
ments cover materials, finishes, specific component requirements (such as connectors, optical components), durability, PWBs, manufacturing methods, modifications, and qualification tests for materials.
• Electromagnetic Compatibility and Electrical Safety (Bellcore Docu-
ment GR-1089). Electrical requirements cover electrostatic discharge (ESD), electromagnetic emissions and immunity, lightning and ac power fault, steady-state power induction, and electrical safety.
412
Transmission Systems Design Handbook for Wireless Networks
Bellcore has changed its name to Telcordia, but the documentation numbers have not changed, so the document can be found under Bellcore or Telcordia nomenclature.
References [1]
Weizmann, M., SDH Radio: The Technology of Today and Tomorrow, Microwave Journal, January 1997.
[2]
Andrew Technical Seminar Series, Terrestrial Microwave, Elliptical Waveguides and Pressurization Products and Systems, 1999.
[3]
Andrew Catalogs #37 and #38.
[4]
Huurdeman, A. A., Guide to Telecommunications Transmission Systems, Norwood, MA: Artech House, 1997.
[5]
Goralski, W., ADSL and DSL Technologies, New York: McGraw-Hill, 1998.
9 Transmission-Network Deployment 9.1 Equipment and Services Ordering Process 9.1.1
Planning and Design
In most countries, telecommunications is now accepted as a basic part of the infrastructure, along with power and transportation, important for the growth of a national economy. Telecommunications is also recognized as the means of accelerating the distribution of the results of economic growth, including to remote and inaccessible areas of a country. Modern telecommunications is expected to usher in a global economy and a single world marketplace, making information in the form of voice, data, or video accessible to persons anywhere in the world. Studies conducted on an international basis show a definite correlation between growth of the economy and availability of telecommunications facilities. Telecommunications is also an eco-friendly means of meeting the communication needs of people, since it cuts down on travel costs and saves natural resources such as fuel and forests. It is in this context that by the beginning of the twenty-first century, telecommunications had been recognized by the governments of almost all countries as a thrust area in their development plans. As mentioned in the introductory parts of this book, copper facilities were the main media for transmission of information over long distance. The next phase, beginning in the 1980s, was the development of fiber-optic facilities, with wireless access being developed over the last few years. Many countries today are actually skipping the copper phase in the development of their 413
414
Transmission Systems Design Handbook for Wireless Networks
telecommunications infrastructure and jumping straight into the wireless phase, sometimes forgetting that fiber-optic transmission systems are also required in order to build wireless networks. In other words, it is difficult and sometimes even impossible to build a wireless network in a country that has no fiber-optic or copper infrastructure. The transmission network must be designed to meet service demands but always with the most economical routing in mind. Two scenarios are most common in wireless-network deploymentleased facilities and microwave networks. For larger networks, usually it is some combination of both, and even the completely leased-lines (facilities) network requires careful transmission-network planning. Survivability and reliability of the network are achieved by means of transmission loopsthe ring configuration or a combination of star and ring configurations. Wireless-network transmission planning and design typically consist of the following steps: 1. Define the customers requirements and task delineation (who is doing what); 2. Start with the RF plan and define the switch location, hub sites, POPs, and other important sites; 3. Calculate the access (backhaul) and core network transmission capacity; 4. Define and include future additions, upgrades, and expansions; 5. Based on capacity, define the most suitable transmission-network topology; 6. Based on the required capacity and the network topology, define the equipment required and identify suppliers; 7. Determine the cost of equipment and services. 9.1.2
RFQs, RFIs, and RFPs
9.1.2.1 Definition and Purpose
The request for information (RFI) and request for pricing (RFP) are somewhat less detailed requests that are usually sent to the equipment supplier or service provider in order to acquire information on their products and services. Response to an RFI could be simply a collection of data sheets, brochures, user manuals, and so on. Response to an RFP could consist of a few pages of standard list pricing, usually without discounts or additional considerations.
Transmission-Network Deployment
415
The RFQ (or tender document) response describes in detail the equipment or services to be supplied. The RFQ is prepared by the customer for the purpose of soliciting hardware, software, or services information for the evaluation and possible procurement by the customer with a specific project in mind. Answers to questions asked in the RFQ will provide the customer with a better understanding, both in terms of finances and system integration and capacity, of the equipment and services that the supplier (vendor) can offer and provide [1]. Topics discussed in the RFQ are usually, but not limited to, commercial conditions of contract, technical conditions, project management, quality assurance and reliability issues, procurement and delivery issues, training and documentation, in-service date, and RFQ response due date. For turnkey contracts, a specific and very detailed scope of work document is also included to define the installation and test services being added to the contract. In some cases, different contractors may provide other manufacturers equipment (OEM) and installation services. Equipment evaluation is usually based not only on the technical specs and price but on other criteria, such as experience of other customers in North America and internationally with the same or similar equipment, warranty, and customer support. Directions in which technology will go in the next decade and compatibility of the equipment with those future trends are also important. Adherence of the equipment with all the existing North American and international telecommunications and quality standards and interoperability with the equipment of other suppliers is usually mandatory. In many transmission networks, speed of deployment will be a critical factor in the process of equipment or supplier evaluation and has to be addressed and discussed in detail. All suppliers are usually provided with the opportunity to discuss RFQ proposals individually; after that, final discussions will be held with up to three top candidates, after which the financial and legal terms are determined. Since the document is confidential in nature, it must be treated accordingly. The suppliers are not permitted to disclose specifications to any person or entity except employees of the supplier and its affiliates who have a need to know and who have been informed of the suppliers obligations. The supplier should use the same degree of care to avoid disclosure of such information as it does with respect to its own confidential information. The supplier should always provide the information on the manufacturer of the equipment offered but not manufactured by the supplier and disclose any OEM agreements with other equipment manufacturers. The pricing model is usually defined by the customer; it must be as close to the planned network as
416
Transmission Systems Design Handbook for Wireless Networks
possible and the supplier should try to adhere to the requirements and definitions provided in the RFQ as closely as possible. The RFQ should be structured to simplify the process for the customer and supplier alike. Responses must specifically address each point set forth in the RFQ and should be clearly answered in the response tables. The amount of information submitted is left to the discretion of the respondent, but it is imperative that pertinent information be submitted and that those individual topics of interest be dealt with completely and concisely. Those suppliers failing to provide complete and accurate responses can be discredited for the quality of the response and appropriately penalized in the response evaluation process. Should the supplier choose not to bid in response to the RFQ, they are usually requested to specify in a cover letter the reasons for the decision. The following issues are usually covered in the RFQ document and must be discussed in detail by all bidders: • Vendors market presence; • Performance; • Future product evolution; • Standards compliance; • Strategic relationship; • Turnkey capabilities (EF&I-Engineering, Furnishing, and Installa-
tion);
• Alternative solutions and unique offerings; • Technical support; • Warranty, repair, and return policies.
It is very important to use the proper terminology when preparing the RFQ. For example, will and will not identify requirements to be followed strictly and from which no deviation is permitted. Should and should not indicate that one of several possibilities is recommended as particularly suitable, without mentioning or excluding others; that a certain course of action is preferred but not necessarily required; or that (in the negative form) a certain possibility or course of action is discouraged but not prohibited. May and need not indicate a course of action permissible within the limits of the document. Can and cannot are used for statements of possibility and capability, whether material, physical, or causal.
Transmission-Network Deployment
417
9.1.2.2 RFQ Process
The main activities in the process of preparing a response to the RFQ are as follows: • The wireless operator prepares the RFQ for equipment or services
and sends it to the equipment or service providers with the expectation of receiving a response within a few weeks time.
• The RFQ is received and analyzed. A decision is taken as to whether
or not to prepare the tender (RFQ response). The type of tender is considered and the necessary resources for preparation of the tender are secured.
• Customer solution and tender preparation and the proposed solu-
tion can, for example, include the commercial tender part, the implementation part, the technical part, and the service part.
• A risk analysis and a profitability analysis are performed, and tender
review meetings are held with all involved in its preparation. At this point, the supplier may decide not to respond to an RFQ.
• Tender finalization: The tender is compiled, reviewed, and approved. • Presentation and follow-up: The tender is submitted as a system
proposal to the customer and followed up on.
• Tender evaluation: After submitting the tender to the wireless opera-
tor, a tender quality evaluation is made.
• The customer (wireless operator) makes a short list of equipment or
service providers and continues negotiations with them.
For bigger projects, usually two or more vendors are picked to share the market and provide backup in case the other has problems delivering on time or at the quality previously agreed upon. The proposal for the transmission network normally includes the following information: • Optimum transmission media and network topology to satisfy all
the requirements given by the customer;
• Dimensioning of the transmission networks; • Bill of quantity (BOQ) and third-party-vendor connection plan; • Pricing;
418
Transmission Systems Design Handbook for Wireless Networks
• Transmission services description [engineering, furnishing, and
installation (EF&I)];
• As-built documentation (optional); • Network management system (optional); • Responsibility matrix, also called task delineation list (optional).
The three optional topics are usually not part of the tender; however, they are important for the contract negotiations and should be included in the contract in order to support the handover of the implementation system to the customer. Transmission-network topology is created based on the RFQ content. In some cases, the RFQ clearly states all the relevant details needed to produce a plan for the transmission network, but sometimes it does not. When this happens, assumptions are made by the network planner. All assumptions need to be agreed upon or otherwise corrected when contracting the system. The transmission-network topology includes both the access (traffic in each city or region) and the transport network (backbone network), as applicable. If connection to third-party vendors is required, a document describing the interface requirements is normally produced. A description of how to install and connect vendors products together with other equipment should be included. The topology of the NMS should be considered at this stage, as it is highly dependent on the network topology. The NMS outlines how the management systems and their respective network elements are to be interconnected. The BOQ is based on the transmission-network topology, which in this phase is still a theoretical model. The level of details in the BOQ can be different depending on whether it is reflecting a budget quotation or a more detailed quotation. Nevertheless, it is a list of all quoted items for the proposed network to be priced in the tender. The BOQ is created using standard configurations. The purpose of creating standard configurations is to minimize the number of options, increase flexibility, and minimize the number of spare parts. Spare parts can be included in the BOQ or listed as stand-alone items. Equipment and installation material spare parts must be considered. Most suppliers have a guideline for spare-parts dimensioning. Pricing can include equipment, spares, services, training, tools, NMS, and so on. The tender can vary from case to case and can be made on a per-item, sales-object, hop-configuration basis or other. Transmission service prices are often divided into two categories. One is related to network design and implementation and the other to support and maintenance after acceptance.
Transmission-Network Deployment
419
The following is a partial list of technical requirements, regardless of what kind of transmission equipment or telecommunications equipment in general, the RFQ is about: • Services (engineering, installation, testing); • Hardware (interfaces and redundancy); • Performance monitoring, alarm information, and NMS; • Access security; • Operational support system; • Physical and environmental specifications; • Power requirements; • Grounding and bonding; • Equipment reliability calculations; • Antennas and cables; • Enclosures; • Installation; • Acceptance testing; • Spares and upgrades; • Shipping, delivery, and labeling; • Technical support; • User documentation and training materials. 9.1.2.3 As-Built Documentation
An as-built document is normally not completed until each site is integrated; however, it should be stressed that an agreement of its content must be included in the contract. It is advisable to prepare a standard installation drawing or show the customer a real site, in order to minimize potential misunderstandings in the future. The customer should approve the standard installation. It is important to note that the as-built documents content as listed below refers only to transmission considerations: • Site situation plan: shows the site location on a map; • Floor plan drawing: indicates the location of the installed transmis-
sion equipment; • Microwave path calculations;
420
Transmission Systems Design Handbook for Wireless Networks
• Cable way drawing: gives indoor copper and fiber-optic cable instal-
lation information, may also include coax cables and waveguides in case of microwave installation;
• Antenna placement information: shows MW antenna arrangement; • Alarm allocation table: indicates alarm cabling; • Power distribution: indicates power distribution for the indoor unit; • Transmission configuration data: gives hop information used for soft-
ware setup;
• Transmission rack layout: shows layout of the indoor equipment in
the transmission rack;
• Transmission traffic layout: gives traffic distribution on the T1/E1
level;
• Transmission trunking diagram: shows the transmission network; • Transmission plant specification: lists equipment to be used on site,
including installation material (e.g., cables);
• Transmission product list: lists main units including equipment serial
number information;
• Transmission acceptance test document: confirms the customers site
acceptance;
• Factory test: gives transmission equipment test results provided by
the factory.
From this overview of the contents of the as-built document, it is clear that all installation aspects are considered at this stage. The as-built document shows how the equipment should be installed on a site. In many cases, the customer is not familiar with the equipment. Therefore, in order to shorten the approval process, it is advisable to prepare an installation demonstration. 9.1.2.4 Responsibility Matrix
The purpose of the responsibility matrix (sometimes also called the task delineation list) is to clearly state the responsibility related to all areas of the project. It is of great importance that all aspects are considered, especially when the customer must fulfill a certain task. During contract negotiations, all aspects of implementation, such as the as-built document proposal, are considered and agreed upon by the vendor and the customer. The transmission network responsibility matrix can contain in excess of 200 activities.
Transmission-Network Deployment
421
9.1.2.5 Contract Negotiations
A partial list of the items to be discussed between parties during the negotiation period is shown here: • Agree on responsibility (task and responsibility matrix); • Meet the customers expectations on the network design by match-
ing business and technologies to the correct cost model;
• Agree upon handover content and medium (deliverables); • Agree upon acceptance test procedure and timescale; • Agree upon how to report site progress and implement change
orders;
• Ensure that there is a legal professional present when preparing the
contract and throughout the project. There will be discussions about prices, timescales, and so on. It is therefore useful to have an experienced lawyer who is aware of contract details;
• Create a standard site with the customer in order to avoid mis-
understandings;
• Discuss site acquisition and civil works process; • Discuss acceptance testing and commissioning.
It is important to note that in large transmission projects, as in any other project, there is compromise between the speed of deployment, reliability of the system, and the price the wireless operator must pay for the network. Spending more money and time in the beginning will guarantee a reliable network that will continue to work even under unfavorable conditions. Implementing sound transmission-engineering techniques, such as ring topology, hardware redundancy when necessary, and so on, will in the future prove to be a good investment. 9.1.3
Negotiating the Statement of Work
Once the great barrier of making the initial sale is crossed, the negotiation starts. One of the great ideals of modern business practice is based around partnership between suppliers and customers. To form a long-term partnership, both parties must reach a mutually beneficial business agreement through negotiations. In telecommunications today, wireless operators are a small group of people that attempt to subcontract as many activities as
422
Transmission Systems Design Handbook for Wireless Networks
possible without permanently hiring too many skilled people. These people are required only during the brief period of build-out time (312 months) and become redundant after that, so it is wise to use experienced contractors and consultants to do the initial engineering and installation work. Some wireless operators go so far in outsourcing that they even subcontract the network operations center (NOC) and operations and maintenance staff and facilities. The customer, in this case the wireless operator, should provide a description of services required; the provider of equipment or services will then supply the following information, describing in detail: • Scope of work (SOW); • Prices, fees, and expenses; • Terms and conditions (T&C).
The wireless operator typically needs to negotiate with engineering and consulting companies outsourcing engineering, supervision, project management, and so on. They also negotiate terms and conditions of the equipment procurement, installation, and testing with equipment suppliers (transmission equipment, such as microwave systems, fiber-optic equipment, DACSs, power supply, and battery backup). Long-term maintenance, optimization of the network, and even network operations could also be part of the contract. 9.1.4
Negotiating with Telecommunications Providers
In the case of leased lines, negotiations are required between the wireless operator and telecommunications providers for the leased T1/E1 lines or any other bandwidth, as required. Leased-lines providers are also service providers (long term) and should be treated as such. They usually offer better terms and conditions to big customers and those contracts are longer, at least three to five years. Quality of service must be specified in the contract, along with the remedy in case the wireless operator has interruption in service, repair time is too long, or the quality of the service, when available, is unacceptable. In such cases, the customer should request penalty charges. Information typically provided by the wireless operator and given to the telecommunications provider(s) is as follows:
Transmission-Network Deployment
423
• Planned RF coverage areas; • General area map of each market location and, if possible, a
preferred-network diagram;
• Unchannelized T1/E1 from cell sites to BSC/MSC; • Potential location(s) of BSC/MSC/NOC.
The telecommunications provider should supply the following information to the wireless operator before negotiations take place: • Description of how serving areas are divided, and existence of local
exchanges within a local access transmission area (LATA); • Existing transmission network;
• Copper, fiber, microwave, and satellite in which cities and for how
much;
• Availability of the existing network based on the statistics and meas-
urements over an extended period of time;
• Diversity and protection in the network (e.g., SONET/SDH self-
healing ring);
• General information about the transmission network and its capac-
ity, POPs, features (e.g., MUXing), and equipment at switching offices, fiber hubs, remote terminals, and so on;
• Colocations and interconnections with other phone companies and
other T1/E1 lines providers;
• Offered products and services; • Possible network design and routing for intracity and intercity sites
and required capacity.
Initially, the wireless operator will only provide information on the search rings (radius of up to 1.0 km, or 0.6 mi) defined by RF engineers. Later, when the primary site candidate is determined, it is possible to check the feasibility of the location from the RF and transmission perspectives. The telecommunications company site walk (site visits) should be attended by the wireless operators RF and transmission engineers and the telecommunications companys outside plant (OSP) engineers and construction engineer; the site walk should be conducted as soon as the primary site candidate has been determined. It is very important to keep the telecommunications
424
Transmission Systems Design Handbook for Wireless Networks
company informed on all changes, and the site locations should be provided as soon as they are finalized in order to shorten the turnaround time for ordering T1/E1 lines. A common problem results from the fact that RF engineers provide cell-site and search-ring locations in degrees latitude and longitude. However, telecommunications company operators must have at least an approximate (and existing) address to provide quotes on the running T1/E1 circuits from that location to the wireless operators switch (BSC) location. The telecommunications provider should supply the following information on T1/E1 pricing structure: • Tariffs (pricing) for private-line T1/E1s; • Interconnection tariffs (i.e., per circuit or per minute); • Offered plans and volume discounts; • Billing methods in general;
The following standard processes and forms must be discussed as well: • Ordering-process flow (to determine stages and intervals for T1/E1 • • • •
and switched [trunk] services); Time to deliver; Performance-monitoring process (7 × 24, T1/T3, or E1/E3 level); Trouble-reporting methods (procedure, average time to restore, charges, and credits); Upgrades and repairs to network (customer awareness).
Regulatory issues, interconnect agreements, NDAs, and the like must also be discussed before and during the negotiation process. The interconnection agreement is one of the items that the wireless operator must have in place in order to have PSTN circuits ordered from the telecommunications operator. The legal departments and regulatory experts within both companies usually handle this. Points of contact and escalation procedure are also a part of the agreement (contract) between the wireless operator and the telecommunications providers and provide the following: • Account and project manager; • Ordering and provisioning contacts for circuits;
Transmission-Network Deployment
425
• Turn-up contact numbers; • Trouble-reporting numbers; • Project coordinator for telecommunications company site walks for
each site.
9.1.5
Negotiating with Equipment and Services Suppliers
The wireless operator, after issuing the RFQ for transmission-equipment procurement and deciding who will provide the transmission-network solution, now has to decide if the equipment supplier will provide the detailed network design as well [2]. A general list of issues to be discussed and agreed upon between the wireless operator and the equipment supplier, regardless of the type of transmission system and equipment involved in the project, is shown in Table 9.1. Table 9.2 shows some of the items and topics to be discussed and agreed upon when defining services required for the project. Depending on the size of the transmission group within the operators organization, a few, all, or none of the listed activities could be subcontracted. Sometimes the equipment supplier is one point of contact and also provides services. For the transmission systems, these services can include engineering, installation, testing and commissioning, optimization, and project management, among others. Price negotiation, the change-order process, the escalation process in case things go wrong (and they always do on a large project), the decisionmaking process, penalties for delays, and legal and commercial aspects of the contract must be defined to the last detail. Although not shown in the Table 9.2, subcontracting and legal departments must be involved in the service delivery as well as the equipment-delivery process.
9.2 Regulatory Issues One of the biggest headaches facing telecommunications-equipment suppliers is very often not technological but regulatory. Equipment is very often sold and installed in countries other than the country of its manufacture, which may have different functionality and safety requirements. In order to sell their products in other countries, suppliers must go through a certification process (homologation). Quick and efficient approval for the hardware and software in each of the markets requires a dedicated knowledge of both telecommunications equipment and the local regulatory environment by the
426
Transmission Systems Design Handbook for Wireless Networks Table 9.1 Transmission-Network Equipment Delivery
Activities
Responsibilities
Transmissionnetwork engineering
Interpretating of contract(s) between wireless operator, equipment supplier(s), and subcontractors Developing technical specifications for supplier contract and SOW Defining required services from supplier Designing transmission-network solution Defining materials needed to establish the system Helping purchasing dept. define the supplier(s) and negotiate the price Monitoring equipment delivery and installation at site with site manager
Contracts and subcontracts
Establishing legal terms and conditions in supplier SOW Negotiating price Developing supplier relationship in procurement of services and materials Completing purchase orders Communicating with supplier for product and services delivery Resolving supplier issues
Delivery logistics issues
Monitoring timeline and raising sensitive issues affecting project schedule Understanding deliverable to customer and ensuring deliverable is completed Gathering export and import terms for the project Facilitating the sales order, export document creation process Working with subcontracts administrator(s), transmission engineers, and project manager in coordinating the suppliers delivery process upon purchase order completion Monitoring delivery of product to destination and reporting to project team
local company intimately familiar with the requirements of the agencies in the countries they cover. Typically, approval support encompasses, but is not limited to, the following range of services: • Detailed assessment of requirements for each piece of equipment by
individual country;
• Arrangement for a proper local applicant for approval where required; • Collection and assimilation of all materials and documentation
required for compliance;
Transmission-Network Deployment
427
Table 9.2 Transmission-Network Services Delivery Activities
Responsibilities
Transmission-network engineering
Understanding the project and scope of work Defining services needed to establish the system (planning, detailed design, project management, installation, testing and commissioning, and as-built documentation) Defining technical specifications for services contract and SOW Defining deliverables and timelines by the contractor(s) Helping subcontracting dept. to define the service supplier(s), issue new RFQs if required, negotiate T&C, and negotiate the price Monitoring services at site with site manager
• Review of test data to ensure agency compliance; • On-site testing support where required; • Close coordination with clients to ensure that all necessary support• • • • • •
ing documentation has been provided and properly prepared; Preparation of approval application or compliance folder to local agencies specific formats; Presubmission and coordination meetings with regulatory agencies as necessary; Modification of application (if necessary) based on agency comments, feedback, and recommendations; Submission of completed application or maintenance of compliance folder; Coordination with agency officials through and until receipt of approvals; Notification of application approval and delivery of approval certificate.
To answer such questions as, What is the dialing and numbering plan in Pakistan or Croatia? What are the specific requirements for a new interface in China? What are the frequency band and channeling plans for the microwave systems in [a specific country of interest]? Local presence across the region and broad technical and regulatory expertise are needed. There are
428
Transmission Systems Design Handbook for Wireless Networks
companies that specialize in finding people who are able to answer questions like these. In some countries, the certification process is simple. In the United States if equipment is approved by the FCC and meets certain minimum requirements, it is considered certified. In other countries, government regulations are tools for excluding foreign manufacturers from their markets. This is done by requiring suppliers to meet stringent requirements and standards and quite often to customize their hardware design. While this is not necessarily technically demanding, it is a very time-consuming process. The CE mark is a requirement for all telecommunications terminal equipment (TTE) products sold into the European Union (EU), effective January 1, 1996. Countries covered are Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, The Netherlands, Portugal, Spain, Sweden, and the United Kingdom. CE marking confirms that a product has been tested and meets the essential requirements of the European Telecommunications Directive to market it throughout the EU. Obtaining the CE mark allows a product to be sold in 15 European countries without any further in-country testing. Also, frequency allocations and channeling plans for wireless services (cellular, PCS, microwave, and satellite) could be different for different countries. For example, some CEPT countries have published their National Table of Frequency Allocations (NTFA) on the Internet. Links to those tables are provided in Table 9.3. CEPT was established in 1959 by 19 countries and expanded to 26 countries during its first 10 years. Original members were the incumbent monopoly-holding postal and telecommunications administrations. CEPTs activities include cooperation on commercial, operational, regulatory, and technical standardization issues. In 1988 CEPT created the ETSI, into which all its telecommunication standardization activities were transferred. In 1992 the postal and telecommunications operators formed their own organizations, PostEurope and ETNO, respectively. In conjunction with the European policy of separating postal and telecommunications operations from policy-making and regulatory functions, CEPT thus became a body of policy makers and regulators. At the same time, central and eastern European countries became eligible for membership in CEPT. CEPT with its 43 members now covers almost the entire geographical area of Europe. As of June 1999, administrations from the following 43 countries are members of CEPT: Albania, Andorra, Austria, Belgium, Bosnia and Herzegovina, Bulgaria, Croatia, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Great Britain, Greece, Hungary, Iceland, Ireland,
Transmission-Network Deployment
429
Table 9.3 Web Links to CEPT National Tables of Frequency Allocations Administration
Web Link to National Frequency Table
Austria
http://www.bmv.gv.at/tk/2ofb/frequenz/fnv.pdf
Belgium
http://www.bipt.be/Pages/French/Telecoms/GestFreq/PlanFreq.htm
Croatia
http://www.telekom.hr/Namjena.htm
Czech Republic
http://www.ero.dk/eroweb/NatTables/ntfa.pdf
Denmark
http://www.tst.dk/uk/frequencies/table.htm
Finland
http://www.thk.fi/englanti/radio/taulu.htm
France
http://www.art-telecom.fr/interactive/frequences/index-d.htm
Hungary
http://www.hif.hu/english/fnft-new/fnfte.htm
Iceland
http://www.pta.is/English/ntfa.htm
Ireland
http://www.odtr.ie/docs/odtr9803.pdf
Italy
http://www.ero.dk/eroweb/NatTables/NatTableItaly.zip
Lithuania
http://www.radio.lt/frequency_table.htm
Netherlands
http://www.rdr.nl/nfr
Norway
http://www.npt.no/english/E_fagomraader/frekvensforvaltning/ frequency_plan.html
Poland
http://www.ero.dk/eroweb/NatTables/NTFA_PL.rtf
Portugal
http://www.icp.pt/publicacoes/freq/indexuk.html
Spain
http://www.sgc.mfom.es/espectro/cnaf.htm
Sweden
http://www.pts.se/radio
Switzerland
http://www.bakom.ch/eng/subpage/category_53.html#topic85
United Kingdom
http://www.radio.gov.uk/publication/ra_info/ra365.htm
Note: Some of these URLs may have changed since the time of this writing.
Italy, Latvia, Liechtenstein, Lithuania, Luxembourg, Malta, Moldova, Monaco, The Netherlands, Norway, Poland, Portugal, Romania, Russian Federation, San Marino, Slovakia, Slovenia, Spain, Sweden, Switzerland, the former Yugoslav republic of Macedonia, Turkey, Ukraine, and the Vatican. CEPT/ECTRA established the European Telecommunications Office (ETO) in 1994 in order to provide expertise for ECTRA members and to contribute to the EUs telecommunications policy on licensing and numbering (http://www.ero.dk). At the end of 2000, as one of several steps in an ongoing reorganization within CEPT, it was decided that the functions carried out by the ETO and
430
Transmission Systems Design Handbook for Wireless Networks
its staff should be transferred to the European Radiocommunications Office (ERO). Since January 1, 2001, the functions of ETO have therefore been carried out by the ERO, under the responsibility of the ERO director. The new European regime should include both harmonization of existing regulations and establishment of common procedures for licensing and numbering. ETO works in both of these areas, on analysis of national situations and studies of issues of topical concern. ETOs activitiescarried out by EROalso provide central and eastern European countries preparing for the liberalization of their telecommunication markets with information on regulatory issues and opportunities for participation in numbering and licensing activities.
9.3 Services 9.3.1
Engineering Services
Selecting appropriately qualified engineers usually results in good engineering designs and can significantly reduce a projects life-cycle costs. Rather than merely meeting minimum standards, the services of appropriately qualified engineers or engineering consultants can enhance a projects value to clients through rigorous consideration of alternatives, analyses of long-term operating and maintenance costs, and innovative design. It is therefore in the clients best interests to use a qualification-based selection method, which demonstrates the competence of the engineering consultant in the performance of the required engineering services. 9.3.2
Project Management
Project management is the application of knowledge, skills, tools, and techniques to project activities in order to meet (or exceed) customer needs and expectations from a project. Meeting or exceeding customer needs and expectations invariably involves balancing competing demands, scope, time, cost, and quality. In the engineering, installation, and construction of a telecommunications network, the project manager oversees the entire project from start to finish and is responsible for maintaining a smooth operation in each step of the deployment of the network. The project-management component of the job emphasizes efficiency by the creation of a comprehensive work schedule that eliminates duplications and maintains continuity, which in turn leads tocost savings. The project manager is accountable for the success of every
Transmission-Network Deployment
431
phase of the project and ensures that even the largest projects are completed on time, within budget, and according to the highest quality standards. The following are the processes that a project manager must follow in order to manage a project effectively: • Integration management: the processes required to ensure that the
various elements of the project are properly coordinated.
• Scope management: the processes required to ensure that the project
includes all of the work required to complete the project successfully.
• Time management: the processes required to ensure timely comple-
tion of the project.
• Cost management: the processes required to ensure that the project is
completed within the approved budget.
• Quality control: the processes required to ensure that the project will
satisfy the needs for which it was undertaken.
• Human resources management: the processes required to make the
most effective use of the people involved with the project.
• Communication management: the processes required to ensure timely
and appropriate generation, collection, dissemination, storage, and ultimate disposition of project information.
• Risk management: the processes concerned with identifying, analyz-
ing, and responding to project risk.
• Procurement management: the processes required to acquire goods
and services from outside the performing organization.
In recent years, there has been a growing awareness that project management is a special skill that can be codified and learned. It is quite different and distinct from the technical design, engineering, and construction skills most readily associated with projects. There are usually aspects of a project that are outside the scope of these technical areas, but that have to be managed as well if the project objectives are to be met. This has resulted in the evolution of project management as a separate and distinct discipline. Of course, highly specialized fields, such as telecommunications networks and wireless networks, require some technical expertise and certain experience from project managers. Some engineering firms choose to offer project management as a separate function within the firm. Still other companies offer project
432
Transmission Systems Design Handbook for Wireless Networks
management without engineering. In addition, some companies have their own in-house project management staff. All of these are legitimate vehicles for achieving the management of a project. As projects become larger and more complex, the effective management of them becomes proportionately more significant. Interfacing the local workforce (engineers, contractors, and suppliers) also requires some specialized skill in dealing with local customs and culture. A good book to have handy when working on projects around the world is Teresa Morrison et al., Dun & Bradstreets Guide to Doing Business Around the World (Prentice Hall, 1977). 9.3.3
Outsourcing Services
Using a single point of contact to design, build, and commission a telecommunications network involves many obvious benefits, such as lowering overall costs, reducing cycle time, regulating quality and safety, and simplifying administration and project management. Turnkey solutions to large-scale projects save money. By using a single supplier with project-management abilities, it has been proven that overall costs for a project can be reduced by 15% to 20%. The theory is that using one company to coordinate all aspects of the job, including resource planning (people, equipment, material management), work schedule, and cost management, will reduce cycle time and duplication and maintain continuity for the project. This will in turn reduce costs for materials, equipment, and human resources, as there will be no allowance for downtime and overlap as a result of bad planning. Performance measurement is not possible when multiple suppliers are used to complete a project; however, performance measurement is crucial to ensuring quality in the finished project. A single source supplier that regularly measures the performance of its resources ensures that the project will be completed on time, within budget, and to the highest quality standards. Quality and safety standards are important in ensuring that a job is completed to the clients expectations without unnecessary delay. Utilizing a company that has established processes and procedures for health and safety, quality, customer service, performance measurement, environmental issues, and training and development will ensure that the work is completed according to the highest standards. There is a significant trend toward outsourcing in the telecommunications industry. Suppliers currently deal with numerous contractors and are finding that they are losing control of the costs and the quality of work. More
Transmission-Network Deployment
433
and more, wireless operators are realizing the benefits of outsourcing to third parties so as to offer services such as network planning, customer-care and billing systems, and construction and operations support systems. As the workload increases, the demand for quality installation and construction services increases as well. Using a large company with a permanent or longterm employee base as a single source supplier will ensure that the skilled manpower will be available to complete the project to the highest quality standards while controlling costs and schedule. The wireless operator (and clients in general) should describe the nature and extent of the project as clearly and precisely as possible, including defining the objective(s) to be met and outlining relevant background information. The client should also state its expectations about how objectives will be accomplished and the anticipated involvement of its staff, the engineering consultant, and other relevant parties. The following criteria may help to define the terms of reference for engineering services: • Objective(s). What should the project accomplish? • Background. What factors led up to the project? • Scope. What will be included in, or excluded from, the project? • Approach. How will the objective(s) be met? • Resources. Who will be responsible for what? • Deliverables. What tangible results are expected? • Timing. When will the project start and finish?
Clients may either retain comprehensive engineering services from a projects start to finish or develop a work plan for contracting out specific phases of the wireless-network project to various parties. They should determine which alternative is appropriate in their situation. A preliminary cost estimate will be only as accurate as the defined scope of a project or problem. In cases where very limited preliminary engineering has been undertaken, this estimate will likely reflect the cost of engineering services contemplated in the scope of the project. In cases where more extensive preliminary engineering has been completed and the scope of the project has been well defined, this estimate will likely reflect the total project cost. It is not easy to provide an accurate estimate of the labor cost (services) for projects in different parts of the world. The engineering team, installation crew, methods, tools, man hours, and test equipment will depend on various items, such as system configuration,
434
Transmission Systems Design Handbook for Wireless Networks
scope of provision requested to the vendor, skill of labor, and technicians and subcontractors available in that country. The availability of construction materials and machines, the weather, site access, transportation, size of the system, and the customers work schedule will affect planning and costing of the services. Even the political situation in the country can affect network build-out. Sometimes it is difficult to provide even typical information, since conditions differ greatly on a country-by-country and project-by-project basis. To avoid surprises, some companies provide heavy cost padding in their service proposals; that way they become less competitive and price themselves out of the race. The decision about outsourcing must be very carefully considered [3]. In the case of network design and planning, the company is giving away strategic responsibility to an external company, so the contract management and continuous control over the outsourcer is absolutely necessary. In highly innovative corporate settings, the risk of outsourcing is very high, due to the potential for revealing proprietary and sensitive information to outsiders. Availability of skilled design and planning personnel in the wireless industry is one of the most critical issues and very frequently it is the only driving factor for outsourcing. Feasibility studies and preliminary studies may include, but are not limited to, the following: • Appraisals and valuations; • Investigations and studies; • Rate structure and tariff studies; • Inspections, explorations, surveys, testing, or other services concern-
ing the collection, analysis, evaluation, and interpretation of data leading to specialized conclusions and recommendations;
• Technology evaluation and comparison with the competition; • Feasibility studies on proposed projects, including studies of clients
needs, analyses of conditions or methods of operation, development of alternative concepts, economic analyses, environmental studies, and site-location studies;
• Development of preliminary design reports, including outline speci-
fications, preliminary cost estimates, and so on;
• Schematic design and design development for building projects; • Expert testimony in some cases.
Transmission-Network Deployment 9.3.4
435
Due Diligence
Before placing an order or even short-listing a specific equipment supplier, due diligence must take place. This usually involves meetings and discussions with the company management team, observing its engineering and design and manufacturing and QA processes, checking references, and possibly even visiting sites and locations where the suppliers equipment has been installed and is in operation. This is an activity in which representatives of the procurement group, engineering group, and QA group are jointly involved. First-article inspection and witnessing of the acceptance testing of one (usually the first to be manufactured) or more pieces of equipment are two of the follow-up activities after the order for the equipment purchase has been issued. In case of leased facilities, there are a number of things that must be verified before signing a long-term, in some cases multi-million-dollar, contract: 1. Existing Network Architecture Map of the network with POP information; Type of interconnect. 2. Reliability MTBF; MTTR; BERwhen measured, for how long, and so on. 3. Capacity Total network capacity; Available capacity. 4. Price and Logistics Price (per T1/E1 or T3/E3); Order time and installation time; Customer service (interview three customers using suppliers services). Transmission-network reliability requirements and quality standards for different operators are not the same; we have to keep that in mind when thinking of using leased facilities and during the negotiations period. The highest standards are traditionally found in electrical utility installations like the following:
436
Transmission Systems Design Handbook for Wireless Networks
1. 2. 3. 4. 9.3.5
Electrical (and military, but not available) utility companies; Telephone companies; Cellular, PCS, and Internet providers; Cable-TV companies.
Network Maintenance
The commissioning of the newly installed network or the network expansion or upgrade involves testing to make sure that the network is up to specifications before it is turned on. Once it is determined that everything is operating according to specifications, it will be launched or, in the case of expansion of the existing network, integrated into the live network. Testing and commissioning of the transmission network is described in more detail elsewhere in this book. Given the size and the multiple components involved in building an integrated telecommunications network, it stands to reason that a certain level of service and maintenance of the network is required to keep it running without interruption. Preventive maintenance is an ongoing process and requires testing and checking the network regularly to ensure that no incomplete areas exist. If a problem is found early, it can often be repaired or replaced without ever causing a disruption of service. Another aspect of maintenance falls under the category of emergency restoration. Whether the situation involves a downed line or the effects of a major natural disaster, damaged equipment antennas or cables must be repaired quickly; otherwise, customers are left without service.
9.4 Project Management in Wireless Networks 9.4.1
Definitions
A project is a nonrecurrent, time-limited, and budgeted undertaking for which a goal has been set. The project is performed and managed by a temporary organization (or team) tailored to its needs. The authority of the project-management function is limited to the lifetime of the project. Even very skilled and experienced engineers may have problems with communicating with and managing others. Project managers quite often find themselves in no-win situationseither the project team resents them for driving team members too hard or the client or end user is not happy because the project is late, too expensive, or not in line with requirements [4]. Successful project
Transmission-Network Deployment
437
management involves the effective control of time, cost, and quality. Every project has unique circumstances that require different levels of planning, tracking, scheduling, and management. The typical approach to project management revolves around three key phasesplanning, construction (build-out), and final acceptance. Within these phases a number of procedures, schedules, and evaluations are implemented to assure on-time and under-budget completion. Project management involves mobilizing a design and construction team to plan, control, and implement all of a projects activities from conception to completion of construction. It also involves meeting client requirements related to the projects function, quality, schedule, and budget. Project-management services include, but are not limited to, the following: • Selecting consultants and contractors; • Conceptual studies and economic feasibility; • Planning, scheduling, monitoring, and controlling; • Estimating, budgeting, and controlling cash; • Engineering and design; • Arranging financing; • Procurement; • Risk management; • Construction management; • Commissioning; • QA.
Additional services will vary according to the clients needs and should be described in the scope of work. These may include, but are not limited to, the following: • Commissioning and start-up assistance; • Preparing maintenance and operating manuals; • Determining deficiencies during the warranty period; • Preparing the final acceptance document at the end of the warranty
period; • Assisting with facility management or operations after the commissioning and start-up;
438
Transmission Systems Design Handbook for Wireless Networks
• Providing as-built drawings. 9.4.2
Project-Management Organizational Issues
The S-curve is a tool for helping with the conceptual understanding of the project. As seen in Figure 9.1, the S-curve models progress as well as other quantities of interest against time. Early in the life of the project, the team is building and learning to work together (Curve 1). Once the team has stabilized, it can begin to work more efficiently and the curve begins to accelerate rapidly. Toward the end of the project, work activities slow as the project is brought to successful completion. This is the ideal scenario in the life of the project. Curve 2 shows a project plan that is too aggressive; too much work is scheduled too early in the project, before the team has even formed and begun to work together. Mistakes will occur, and rework and changes will be required. Curve 3 shows that too little work is accomplished in the beginning, and the pressure increases as the deliver dates approach. This type of curve is to be avoided. Project completion
100%
2/3 Time - 3/4 project completion 2
1
3
1/3 Time -1/4 project completion
0% 0% Figure 9.1 The S-curve model.
Time 100%
Transmission-Network Deployment
439
The importance of communication and clarity cannot be overemphasized. Project management is primarily a communication taskunderstanding what the team is doing as well as understanding the needs of the client and ensuring that the team is meeting those requirements and expectations. Documentation of meetings, agreements, progress reports, issues, and achievements is very important. It is desirable to write things down and date and distribute information widelyestablishing a paper trail is vital. The project manager should know when to use technical jargon or acronyms and when to stay away from them; if the client is not a technical professional, it is better not to confuse him or her, since that can lead to serious misunderstandings. It is practically impossible to manage a project and be an individual contributor at the same time. Delegation and monitoring of the activities are the keys to success. A rule of thumb is that the project manager will have to dedicate 10% to 20% of his or her time to every person under his or her control as management overhead. That means that if the project manager is managing 5 or more people, there will be very little or no time to do any real work. If the team has more than 10 people, chances are that the project manager will need help from other team leaders. Delegation of responsibility should also mean that certain authority, in addition to responsibility, has been given to the person. It is impossible to have one without the other. Managing changes, scheduling, and progress tracking are very important parts of any project. Every project, and especially wireless-network buildout, will require changes in scope at some point. In real life, RF cell-site network design and transmission-network design usually change so often during the initial phase of the project, it is even difficult to track all of the changes on a daily basis. Changes could be very expensive, so it is necessary that the people requesting them understand their impact and are willing to live with the consequences. Although there are a number of project-tracking software tools, MS Project is one of the most widely used. A number of projectmanagement companies have their own software tools to track daily progress of the project and perform other functions. Status meetings should focus on the status of the project, and last no longer than 20 minutes. Management resolutions and other issues should be resolved separately. Team building and celebration of success are important for the morale of all involved in the project. Very often, milestones and deadlines met on time, especially with respect to long-term projects, are good reasons to have lunch together or get together after work. End-of-project celebrations are also important, and sending an invitation to anyone who worked on the project, even briefly, and helped reach the goal is a good idea. The project manager is the one to take heat from sponsors and clients personally. One of
440
Transmission Systems Design Handbook for Wireless Networks
the main rules the project manager should follow is to always praise team members publicly and, when necessary, criticize them privately. 9.4.3
Project Stages
At the beginning of the project, the operator must set out long-term objectives for the long-term wireless-network development, but take into consideration short-term constraints. This is usually called the strategy development stage. Normally the strategic goal can be reached in a number of different ways, all of which should be considered before a final decision is made. With respect to the transmission part of the network design, this usually means deciding whether to own or lease E1/T1 lines. Three options are typically considered: 1. The operator builds (to own) the fiber-optic or microwave transmission network. 2. The operator leases the required E1/T1 and E3/T3 facilities with a plan to replace them eventually with owned fiber-optic or microwave systems. 3. The operator decides on a combination of leased and owned facilities (E1/T1 lines). Making an accurate prediction of the costs of these three models can be extremely difficult and complex. The calculation will depend on a number of factors, such as the number of competitors that have available capacity ready to be leased out, the quantity and quality (and reliability) of E1/T1 or E3/T3 lines, the speed of deployment, strategic alliances with telecommunications operators and competition against them, and so on. The wireless-network planning stage starts with clearly defined customer requirements that all parties involved in the project understand and agree upon. Task and responsibility definition and delineation are required in order to understand who is doing what and in which timeframe. RF design, closely followed by the core network and transmission design, is a starting point of every wireless-network build out. Some of the major activities are listed below: • RF, transmission, and core network design analysis; • Site planning and analysis;
Transmission-Network Deployment
441
• FCC (in the United States) and Industry Canada (in Canada) appli-
cation preparation; • Zoning and litigation support; • Budgetary and timeline projections; • Equipment and services RFQs, RFIs, and RFPs.
The critical path is defined as the longest path, in terms of activity duration, throughout the project duration. In wireless networks, site selection and acquisition is usually the critical path. It drives the completion date of the project and must be identified early in the planning stage. The wireless-network deployment stage, also called network build-out, can include a number of different components. Assuming that most of the design activities are already accomplished, some of the key deployment activities are listed below: • Ordering equipment and services; • On-site and off-site project management and oversight support; • Site selection and acquisition; • Construction supervision; • Civil works; • Tower erection and supervision; • Grounding system installation and testing; • Facility layout and construction; • Minor design changes and modifications; • Equipment installation; • Complete CAD drawings; • Training and education of clients personnel. 9.4.4
Leased-Lines Tracking Process
The tracking process is an important part of managing the wireless-network build-out using leased lines (facilities). There could be hundreds of leased T1/E1 lines whose delivery will depend on other internal and external factorssite readiness (shelter and power available, for example), circuit provider schedule, and weather. A detailed weekly report will show progress in the network build-out, potential bottlenecks, and delay-causing issues [5].
442
Transmission Systems Design Handbook for Wireless Networks
This report will contain site names, site numbers, contacts, and all fields associated with the dates on installation of the E1 and POTS lines (see the list below), and will be a part of the overall network build-out schedule run by the project-management team. This information is required for all the circuits leased from the local telecommunications company: • Circuit typeRBS, PSTN, or 911; • Site numbersite-numbering scheme; • Site namesite-naming scheme; • Site addressa unique address for each site at each cell site; • City, county, and zip codeto coincide with address; • Latitude and longitudeunique coordinates associated with each cell • • • • • • • • • • • • • • •
site; T1/E1 and POTS order datedate entered when circuit is ordered; T1/E1 and POTS order numbernumber assigned by service provider when order is placed and used in maintenance and tracking; POTS phone numberdial-up number; Plant test datedate occurring 2 days before the accept date, when vendor supplies the circuit for testing; T1/E1 and POTS due datedate when circuit is expected to be installed; Acceptance datedate when circuit is accepted from vendor; In-service datedate when site is optimized and is operating; Customer circuit IDtag used by customer to track all circuits across market; Circuit facility assignmentvendor circuit identifier; Carrier ID numbersused if a circuit is in more than one LATA (i.e., an IXC provider); Construction chargesinitial charges to bring T1/E1 circuit to customer; T1/E1 costprice of monthly recurring cost (monthly fees); Contact names and numbersused by vendor to contact customer personnel; Circuit order contactused to track circuit-order contact person; Directionsdirections to the site.
Transmission-Network Deployment
443
A weekly conference call is held between the local network design engineer, the leased-lines coordinator (facility coordinator), and the local service provider, to discuss the implementation progress and status of all circuits ordered. The operations personnel will notify the network designer when each circuit has been accepted for service. 9.4.5
Change Orders
During the fast-paced build-out phase, things are sometimes done without a written order or any written trace of who had done the work, who ordered it, and why. On a big project, this could amount to a significant amount of money that after the fact nobody wants to take responsibility for. To avoid the trap of performing the work without written authorization, a change notification form (see Figure 9.2), also called a change order, is used. 9.4.6
Postinstallation and Optimization Activities
Some of the most important postinstallation activities are listed below: • The punch list points out all minor adjustments (because no matter
• • •
•
how good a plan looks on paper, small changes will always be required). Network testing and commissioning are done per ATP documentation (final acceptance). A final walk-through is conducted with the customer after any punch-list items are addressed, and the system is thoroughly tested. Site documentation and as-built drawings are generated during the project. A copy of this documentation is given to the client upon project completion. The customer may have its own in-house technical staff, which will need this information to upgrade and maintain the new system. User training is provided as part of any project, regardless of size.
It is a good practice to launch the network only after it is clear that it is performing according to plan. This could be assured by starting with a small number of so-called friendly customers (company employees and their families or business partners), who are offered free or deeply discounted service over the first few months after the network has been launched. This can help troubleshoot initial problems that might arise in such a complex system as a
444
Transmission Systems Design Handbook for Wireless Networks
CHANGE NOTIFICATION FORM Project name:
Change #: Market: Site name:
Site No.
Initiated by (name): Company:
Date:
Initiating discipline (check one): RF engineering ¨ Transmission network engineering ¨ Architecture/engineering ¨ Construction ¨ Description of change (include attachments, pictures, and drawings):
Reason for change:
Change initiator:
RF LEAD approval: Approval (name): Signature: Date: Company:
Transmission LEAD approval: Approval (name): Signature: Date: Company:
Figure 9.2 Sample change-notification form.
wireless network and enable tweaking of system components. Many wireless operators want to launch the network as soon as possible and skip this
Transmission-Network Deployment
445
important first step in order to do so. However, bad service in an important market over the first few weeks and months in the life of the network can leave a very negative first impression, and even after the problems are solved, the operators reputation could be negatively impacted. Even when all the correct steps are taken, the wireless network will need improvements, upgrades, and minor changes over time. This could be done by the team that originally designed the network or by some other teams of experts hired to perform optimization services on the wireless network.
9.4.7
Project-Management Tools
There are a number of project-management software tools on the market, and Microsoft Project is one of the most popular ones. Consulting and project management companies sometimes use their own specialized software tools often customized for the unique requirements of a communications network. They facilitate tracking and scheduling of the network build-outs and serve as an information storage tool for the network build-out programs. There are three main areas that this tool must address: 1. The scheduling of tracking and reporting at the client, management, and control levels; 2. The collecting and providing of information related to managing the deployment of a large telecommunications network; 3. The integration of cost with schedule to provide cost control and cost forecasting. The scheduling tool must indicate site-development status using the following items: 1. Milestones: an operating-company or client level for reporting progress; 2. Activities: a project-management or project-control level for analyzing and forecasting site design, acquisition, and construction activities; 3. Tasks: a control-level checklist used by the disciplines to identify and provide the status of the detailed work tasks required for site completion.
446
Transmission Systems Design Handbook for Wireless Networks
Site information is collected and recorded by the following functional areas: • RF engineering; • Switch and core-network engineering; • Transmission engineering; • Real estate site acquisition and zoning; • Architectural; • Building permitting; • Construction; • Power and telecommunications company (utilities) coordination; • Geotechnical; • Land survey; • Procurement.
Reports generation is one of the most important features of any project-management tool. Usually there are a number of different reports to choose from; some of the most commonly used in wireless-network buildouts are the milestones list, the activity Gantt chart, the activity status report, the activity duration report, and the activity completion count by week. A number of customized reports are usually also available.
9.5 Selection of Key Sites 9.5.1
Site Acquisition, Zoning Issues, and Colocation
Everyone wants to use wireless phones, but few people want the towers in their neighborhoods. As more and more of us make wireless phones part of our daily life, many of these coverage problems will be solved by developments in cell-tower equipment and the placement of micros and picocells in areas in which residents are concerned about large cell towers. Wireless providers also need to work together in building monopoles and other tower types in areas lacking existing buildings or towers to hold their equipment. With increasingly more restrictive zoning requirements and a host of new operators rushing to install hundreds of new cell sites in every large city, tools and processes to increase flexibility in site locations without deteriorating coverage, capacity, and service quality are being developed. Site acquisition is
Transmission-Network Deployment
447
a primary bottleneck in most wireless-network build-out projects; it can take up to six months, or even longer, to acquire an appropriate site. When planning or expanding a wireless system, RF design engineers use RF simulation tools to determine optimum coverage of a geographic area based on anticipated RF propagation, capacity, and a number of other factors. Transmission engineers look at the possible connectivity of the cell site to the local telephone companies (leased T1/E1 circuits) and potential use of microwave systems. The real estate group must make a decision based on the potential for the site to be leased, the time required to finalize the required paperwork, zoning restrictions, and so on. Unfortunately, the optimum site is frequently not available, and compromises must be made. Lease negotiations take place after the property for the cell site is selected, and the contractor must negotiate the deal with the landlord, which can be very profitable for the landlord. The owner, in accommodating the antenna(s) and auxiliary equipment, could receive a few hundred dollars per month in rural areas and up to few thousand per month in urban settings. Generally, such leases are written for a 5-year term, with a 5-year renewable option. Some of the main activities in the site acquisition process are shown in Figure 9.3. Of course, this is a very general description and will be different from case to case; document names that are generated during the site acquisition could also be different. It is important to note that this is a multidisciplinary team effort. It requires involvement from the RF group, the transmission group, the real estate group, the legal department, the construction and civil group, and so on. Zoning issues will be different for each jurisdiction, since every city, town, and municipality has its own particular requirements. Negotiating with public officials and civic administrators requires considerable time and leasing experience, so wireless operators quite often hire professionals to do the site acquisition. It is important to note that after weeks and months of site identification, site construction and equipment installation typically take only 2 weeks. The site-acquisition team must be selected with the same due diligence that is warranted when selecting other network design teams (RF, transmission, or switch/core network). It is desirable to find acquisition personnel with construction experience, who will tend to select sites that will meet all specific needs that are required for the wireless-network build out. Before beginning the site-acquisition process it is important to evaluate the market to determine what needs to be accomplished. Some of the issues that must be defined include the following:
448
Transmission Systems Design Handbook for Wireless Networks
Final capacity and coverage specification
Final network design
Final networkdesign report
Approved final network-design report
Search-area map issued to real estate
Real estate process: site searching, identification, and selection
Site candidate information package
No more than these sites selected for feasibility visits
Site feasibility study
Site qualification package and final site selection
Approval of the selected site and request real estate to finalize site acquisition
Detailed engineering visit: NP, MW, civil works
Detailed engineering report (DER)
Real estate secures the site (site lease or purchasing contract, site survey)
DER approval and processing key permits: aviation, zoning, environmental, frequency clearing
Start of site preparation activities
Figure 9.3 Site-acquisition process.
• Size of the market; • Amount of time and money available to acquire and build the sites;
Transmission-Network Deployment
449
• Future expansion of the system; • Whether planning and zoning issues could become a significant fac-
tor in the development of the market; • Type of equipment, enclosures, and antenna mounting structures.
Some things to keep in mind during the site-acquisition process that will help speed up the process are as follows: • It is important to have more than one candidate for the cell site from
the day one. • Monopoles are often an aesthetic and economic alternative to selfsupporting and guyed towers and tend to be more acceptable to planning and zoning committees. They may not be suitable if the loading of the tower is substantial or if requirements for the twist and sway are very stringent (microwave radios). • All legal documents should be as simple as possible; long, compli-
cated documents written in language that is not easily understood may cause the property owner to be wary. Documents should satisfy legal requirements, yet be easily understood by the layman. • Address all local community concerns regarding aesthetics, safety, potential health hazards, environmental impact, and so on, as soon and as precisely as possible. 9.5.2
Switch and NOC Site Selection and Building Requirements
In many cases, the NOC is colocated with the switch and will have one or more NMSs, including an NMS for the transmission system, under the same roof with the rest of the equipment. Sometimes, operators opt to outsource NOC services to a third party, which might be located thousands of miles away and have only a small crew of technicians on site and on call to be dispatched from the remote NOC. The switch site must be chosen very carefully; it has to be close to the POPs of local telecommunications providers (leased T1/E1 lines). If the network is perceived as being even partially microwave, local zoning restrictions and antenna tower-height restrictions could be a limiting factor. As a rule of thumb, if the transmission network is microwave, the switch site should not be close to airports, although this solution is quite often very tempting (availability of inexpensive, large space) due to the tower-height limitations. Also,
450
Transmission Systems Design Handbook for Wireless Networks
airports have a number of different telecommunications systems (including radar systems) that could interfere or be interfered with by the wireless operators RF system or microwave system. Microwave systems will have a major hub at the switch location, so the tower could easily be loaded with a number of 10-ft (3m) parabolic dish antennas and many smaller antennas as well. From that perspective, the downtown area is also not suitable, because very few communities these days allow big antenna towers in the middle of their business or tourist areas. At all times, the local municipal building code will apply as the minimum requirement for all construction, electrical work, HVAC, and plumbing. At a minimum, the following requirements have to be defined before construction of the switch site begins: • Size of the equipment room, battery room, offices, NOC; • Row-and-rack numbering scheme to be used; • dc power and battery backup; • ac power and inverters; • Standard power outlets and inverter-supplied power; • Floor type (usually standard computer-room flooring); • Grounding specifications; • Lightning protection; • Location and description of the telecommunications room; • RF protection; • Heating and air-conditioning and method of delivering fresh air; • Diesel generator requirements; • Fire detection and fire extinguishing systems; • Flooring and floor loads; • Thermal and moisture protection; • Fire retardant walls and access-door requirements; • Overhead cabling and wiring, subfloor wiring description, and appli-
cable codes;
• All cabling entry and exit points to and from the switch and control
areas;
• Lighting; • Prioritizing the equipment installation;
Transmission-Network Deployment
451
• Access to the building (card readers, security levels, or monitoring
facilities);
• Acoustic ceilings, doors, frames, interior partitions, hardware, and
painting.
9.5.3
Cell-Site Selection
Cell-site selection is done by the RF engineers, based on the required coverage of the pertinent geographical area. From a transmission-network perspective, the potential cell-site location must fall into one of two categories: 1. It must have access to T1/E1 lines (leased lines). 2. It must have an LOS with the switch site, hub site, or adjacent cell-site (for microwave link). If neither of the above can be said of a site, construction charges and other logistical complications could delay the site build-out and make it prohibitively expensive. In order to avoid these complications, RF and transmission engineers should work together on the project from the beginning, so that any problems are identified early in the wireless-network design and site-acquisition process. The hub site is the site to which the traffic from the two to three or more cell sites is brought before it is taken to the fiber-optic or microwave backbone system and ultimately to the switch site. This site has a very high importance, and, as such, all the equipment, including the MW system, has to be highly reliable so that the chance of failure is almost negligible. Failure of this site could cause a large part of the network to be out of operation, meaning that a large area would be left without RF coverage. Investing a few dollars more during the initial build-out could save considerable grief in later years. All of the hub sites should have MW systems leading toward the backbone or the switch in a protected (1+1) mode. Protected MW systems are almost double the price of nonprotected (1+0), but will guarantee uninterrupted operation of the important MW hop-carrying traffic from a number of cell sites. Only when there is a sufficient separation between channels or frequencies, or when different frequency bands are used, is a situation like the one shown in Figure 9.4 allowed. If only one channel is used, more than three to four MW systems coming into the same site could be very difficult, if not impossible, to design without an interference between the systems.
452
Transmission Systems Design Handbook for Wireless Networks
Figure 9.4 A microwave hub-site situation to be avoided.
9.5.4
Microwave Repeater Site Selection
When a direct microwave path or LOS cannot be established between two points, it is sometimes possible to establish a path by using a repeater. The function of the repeater is to redirect the microwave beam in order to pass it around or over the obstacle (building or hill). The main requirement here is for the LOS between the repeater and both sides of the microwave link. This could be an active repeater (two microwave radios connected back to back) if distances are long, or a passive repeater if distances are relatively short. There are two types of passive repeaters in use: 1) two parabolic antennas connected back to back through a short piece of transmission line, and more commonly, 2) a flat billboard-type metal reflector that acts as a microwave mirror. The first type is rarely used due to its inefficiency. The effectiveness of a passive repeater is an inverse function of the product of the lengths of the two paths, rather than the sum of their lengths, as one might suppose. Thus, it is highly desirable to keep one of the paths very short (a few hundred meters if possible). The proposed site must have the following:
Transmission-Network Deployment
453
• Access to power (unless a passive MW repeater is used); • LOS with both cell sites; • A place for mounting the antennas.
In a large microwave network, it is hard to imagine that all of the cell sites will have LOS with all other cell sites or the switch office. During the network planning stage, we usually assume that in a wireless network, approximately 5% of the cell sites will need an additional site just to accommodate microwave repeaters. 9.5.5
Colocated Systems and RF Cell-Site Compliance
It is expected that in the United States, antenna capacity will double over the next 3 years for voice traffic alone. Many other developed countries will probably have similar demands for tower space. These demands are driven by improved market penetration, increased airtime among customers using low-cost, one-rate plans, and the need for better network quality [6]. Capacity will also be an important factor in emerging data requirements and thirdgeneration services. Colocation is a logical and creative siting strategy. Colocation can be approached in two ways. The first is for every operator to try to negotiate colocation of its RF and microwave equipment on another operators tower. The second is to outsource the business of antenna installation and rent the tower space from independent companies offering site engineering, acquisition, and installation services and the handling of routine maintenance. Based on the new requirements of the FCC, wireless service providers in the United States will have to prove that they meet safety guidelines for all cell sites constructed, licensed, and activated before October 15, 1997. They must ensure that cell sites comply with safety limits for human exposure to radio-frequency emissions. Some of those requirements for existing and new operators are as follows: • RF emissions of all new cell sites must be assessed and documented
before the facilities are activated.
• Any time a licensee renews its operating license, it must document
compliance of all sites.
• Any time an operator modifies a site in any way, it must prove that
the site remains compliant.
454
Transmission Systems Design Handbook for Wireless Networks
The colocation trend in the industry can actually cause compliance challenges that operators otherwise may not have expected. The reason is that they must submit compliance records for their own equipment and for the equipment owned by colocation tenants at the site. This is particularly important for rooftops, where multiple operators install transmitting facilities. In certain situations, and depending on the site accessibility to the public, if emissions exceed the maximum allowed exposure levels, any company that contributes 5% or more to the emissions in that area is responsible for mitigating the problem. In cases in which the tower meets certain height criteria or the site operates at low power levels, they could be exempt from such routine procedures, but the operator has to prove that the site falls into this category. Such companies must also continue to meet regulations for protecting workers who climb the towers. Three different compliance procedures are acceptable. The first is to conduct a paper study to calculate exposure levels based on the type of equipment and operating conditions at the site. A second method is to use industry-accepted software tools that use computer modeling and simulating techniques to perform the calculations. A third method is to take actual measurements at the site location. RF interference is not only a problem in colocated systems, but in any RF system that can interact with existing systems. Colocation additionally complicates interference analysis and control. Microwave equipment must be included in all the calculations determining interference between operators, as well as safety and RF exposure issues. Table 9.4 shows some of the important information required for the calculation of potential interference between different pieces of microwave equipment from different operators.
9.6 Microwave Deployment 9.6.1
Microwave System Scope of Work
Microwave deployment (build out) is a multidisciplinary activity that involves a number of specialized experts, regardless of whether it is a new microwave system, an upgrade, or expansion of the existing facilities. Full turnkey (EF&I) engineering services for the transmission (microwave in this case) include the following: • System design and feasibility studies; • Program and project management;
Transmission-Network Deployment
455
Table 9.4 Microwave Colocate Operator A site ID
Colocate site ID
Operator A site name
Colocate name
Operator A site address
Colocate contact name
Operator A contact name
Colocate contact phone
Operator A contact phone
Colocate contact fax
Operator A contact fax
Colocate MW engineer
Operator A MW engineer
Colocate MW engineer phone
Operator A MW engineer phone
Colocate MW engineer phone
1
2
3
MW antenna type MW antenna manufacturer MW antenna size
(ft.)
Antenna elevation
(m)
Antenna azimuth
(deg)
Transmission-line type Transmission-line length
(m)
MW-radio type MW-radio manufacturer Transmitter output level
(W)
Maximum EIRP
(W)
Tx frequency
(MHz)
Rx frequency
(MHz)
Bandwidth
(MHz)
Other equipment Power requirements Equipment heat dissipation
(W)
Rack-space requirements
(RUs)
Wall-space requirements
(WxHxD)
Floor-space requirements
(WxHxD)
Please attach following diagrams:
1. C / I curves for the radio 2. T / I curves for the radio 3. Antenna radiation patterns (E and H plane) 4. Site layout with antennas, radios, and Tx lines shown
Total
456
Transmission Systems Design Handbook for Wireless Networks
Table 9.4 (continued) Comments:
Colocate MW engineer signature
________________________________
Date _____________________
Operator A MW engineer signature
________________________________
Date _____________________
• Project planning; • Site and path surveys; • Site civil work; • Site preparation; • Tower and building foundation construction; • Design, procurement, and erecting of antenna structures; • Design, procurement, and installation of equipment shelters; • Design, procurement, and installation of power systems (ac/dc/ •
• • •
solar/diesel generators); Design, procurement, integration, installation, testing, and commissioning of all equipment required to complete the transmission system (including microwave site and path engineering, grounding, lightning, and surge suppression); Project and as-built documentation; Complete training (on-site and off-site), maintenance, technical support, and repair services; Future upgrades and network expansion.
The SOW will be defined between the customer and the supplier and could contain all or some of the mentioned activities. Professionals in various fields of expertise must cover these activities in detail. 9.6.2
Microwave Site Surveys
The process of making a decision about whether a particular cell-site candidate is feasible for the microwave system is shown in Figure 9.5. In a first approximation, topographical and digital maps are sufficient to eliminate certain cell-site candidates if they do not have an LOS with any
Transmission-Network Deployment
457
other cell site due to terrain obstacles. The second step would entail a site visit, and establishment of LOS availability for shorter microwave systems. For longer microwave hops, a detailed path survey will be required. The following text provides general guidelines for the planning, engineering, deployment, and construction management teams when implementing microwave transmission facilities for the wireless network. These guidelines are not to be construed as absolute, since the requirements of individual sites may be specific and unique. It is imperative, however, that every effort, to the extent possible, be made by all teams to follow the same guidelines to ensure a standard configuration. A homogeneous design allows for a commonality of equipment to be implemented worldwide, and ensures high standards and a common practice. By adding a microwave analysis during the site-feasibility evaluation, a considerable amount of engineering time and cost savings can be realized prior to microwave deployment. Planning a microwave system in a wireless network is always a dynamic and continuous process. The transmission (backhaul) system must be able to satisfy present and future capacity demands and provide reliable service. The transmission engineer must be involved in the planning of the radiofrequency (RF) and backhaul network from the beginning, and constant feedback between RF engineers and backhaul network planners is mandatory. Potential backhaul hub sites (POPs, Fiber-Optic (FO), or MW hubs) have to be identified very early in the planning process, since they will be the backbone of the future transmission network. The microwave checklist shown in Table 9.5 is mainly intended for leasehold, rooftop, and tower site evaluations, but it could also be used in any other situation. It includes pertinent site details such as address, site direction, access, restrictions, general information on the site, tower/rooftop/leasehold, available space, type of base station equipment, dc and ac power availability, and zoning restrictions that are valuable to the engineer. This checklist must be issued to the microwave engineering group following the site visit. The information from this microwave survey can be used to determine the feasibility of short microwave hops, generally 3 to 5 km in length, between the site under evaluation and adjacent sites (existing or potential candidates). By using this approach, detail path surveys by outside contractors are minimized. It also opens up the opportunity to install antenna mounts while the site is under construction, thereby utilizing construction resources while they are available. The overall effect is a reduction in cost and the time required to implement microwave. Construction managers, RF engineers, and other site survey team members can be easily trained to perform this quick analysis. By using binoculars,
458
Transmission Systems Design Handbook for Wireless Networks
Transmission network planning (backhaul)
RF design search rings
Long hops
MW hop length less than 3 km?
No
MW plan MW tool analysis
Engineering routing design and preliminary path analysis (MW TOOL, maps/database) No
Short hops
Yes
Frequency coordination
Site visit to verify line-of-sight with adjacent sites (LH/buildings) take digital pictures
No
MW feasible? Yes
Microwave site survey report (checklist)
Path survey
Engineering microwave system design MW tool
MW feasible? Yes
Engineering detailed microwave design MW tool
Site classification for microwave (hub, terminal, etc.)
Frequency coordination Site classification for microwave (hub, terminal, etc.) Microwave system deployment Microwave system deployment Acceptance testing
Site-acquisition team Acceptance testing
Figure 9.5 Microwave feasibility analysis.
Transmission-Network Deployment
459
Table 9.5 Microwave Site Checklist General Information Site ID:
Site type:
Site name:
Tower type and height:
Candidate number:
Max twist and sway of the tower:
Address:
ac power available at the site:
Latitude:
dc power available at the site:
Longitude:
Additional space available:
Contact name:
Additional requirements:
Construction manager:
MW cabinet/telco box required:
CM phone number:
Carrier equipment location:
Date:
Distance from the carrier equipment:
Site Description CDMA
Y N
MW hub
Y N
Cellular
Y N
MW repeater
Y N
MW only
Y N
MW terminal or spur
Y N
Other
Y N
Other
LOS to Adjacent Site Site name/site ID
Azimuth/estimated distance
Type of antenna mount required
Comments
Engineering Comments
Construction manager :
MW engineer :
LOS condition
they can identify remote or adjacent sites. This is generally sufficient to confirm LOS on the short path. Any written observations and comments, such as potential obstacles, buildings, surrounding trees and vegetation, power lines and towers, water towers, lakes, rivers, airports, and communication towers are very important to the microwave designer. Even more beneficial to the engineer are digital photographs that show the possible antenna-mount locations and the actual LOS paths as seen from the site. Panoramic pictures at every 30°, pictures of obstructions and existing facilities (shelter or tower),
460
Transmission Systems Design Handbook for Wireless Networks
electrical connections, and telecommunications installations are recommended and can be very useful. The microwave site checklist should contain at least the following information: • General information providing location and site details that the
engineer must be aware of when designing a microwave system and identification of relevant personnel (e.g., the construction manager [CM], should the engineer have the need to contact the CM to acquire additional information);
• Site type, specifying whether the site is a rooftop shelter, cabinet or
leasehold, or tower;
• If the customers equipment is colocated with other users, informa-
tion about the relevant site and equipment;
• For tower sites, a description of the condition of the tower and
grounding, whether it is an existing site, its capability to accommodate microwave antennas, and the twist and sway values for the tower.
An operator should aim to perform only one site survey to minimize costs. Equipment installation requirements must be confirmedpower, accommodation, and environmental concerns. The ease of service access for maintenance personnel, particularly with respect to tower-mounted equipment, can have a significant impact on costs and repair time. The required loading must be calculated if new tower installations are proposed, and these must take into account the antenna wind and ice loading. Information on the cell-site antenna tower or other antenna-mounting structure is required even before the microwave design has started. In these cases, some assumptions have to be made. A typical loading specification of a tower based on Andrew antennas and typical split configuration of the MW radios follows (ODU-IDU): 1. For urban and dense urban areas (poles and small towers)two 2-ft dishes VHP2-220A; two outdoor radio units (450 mm × 450 mm × 450 mm each), 12 kg each; two half-inch coax cables; 2. For suburban areas (35m and 50m towers)one 4-ft dish (VHP4A-142) and two 2-ft dishes (VHP2-220A); four outdoor radio units (450 mm × 450 mm × 450 mm each), 12 kg each; four half-inch coax cables;
Transmission-Network Deployment
461
3. For rural areas (35m, 50m, and 70m towers)two 4-ft dishes (VHP4A-142) and one 2-ft dish (VHP-220A); four outdoor radio units (450 mm × 450 mm × 450 mm each) 12 kg each; four halfinch coax cables; 4. For highways (35m, 50m, and 70m towers)two 6-ft dishes (HP6-59H) and one 4-ft dish (VHP4A-142); four outdoor radio units (450 mm × 450 mm × 450 mm each), 12 kg each; four halfinch coax cables; 5. For MW high-capacity (backbone) sites (35m, 50m, and 70m towers)four 10-ft dishes (UHP10-59W), two of them facing east (vertically separated by 12m) and two facing west (vertically separated by 12m); four elliptical waveguides. The maximum allowed twist and sway for the antenna-mounting structure for cases 1 to 4 is 0.8°. Maximum allowed twist and sway for the antenna mounting structure for case 5 is 0.4°. Typical cell-site requirements for the nonbackbone MW system (cases 14) are as follows: • One rack space for the MW and transmission equipment (shelters); • At least seven rack units (RUs; 1.75 in) of space for the MW in the
cabinets and telecommunications boxes (outdoor cabinets), and 0°C to 40°C for the indoor radio unit;
• At least four breakers 48V/15A dc; • Four hours of battery backup for all nonbackbone MW systems; • Two openings for cables and waveguides and the feed-through plate.
Switch-site MW/transmission requirements are much more complex and have to be defined separately. 9.6.3
Microwave Path Survey
The selection of a suitable microwave radio site must encompass a number of issues. There are economic and engineering benefits to be gained by maximizing the sharing of infrastructure and sites between the various types of elements in the network, particularly regarding expensive civil infrastructure, such as towers and equipment housings. Good microwave sites, particularly in relation to hub sites, will be relatively high points, providing the maximum LOS availability (Figure 9.6). If a site is unsuitable, alternative sites
462
Transmission Systems Design Handbook for Wireless Networks
Figure 9.6 Large 10-ft MW antenna, viewed from behind.
should be located. This information should be fed back into the network plan, as it can effect both routing and path planning. Attention should be given to future growth requirements in all areas, especially if the site is likely to develop into a future hub. It is always wise to inform land owners of any potential future infrastructure growth to prevent problems at a later date. It is also important to take note of any large bodies of water nearby, since large reflecting surfaces can produce problems due to increased multipath probability. In cases like this, space diversity is the best solution to the problem. As mentioned above, attention should be paid to any local planning restrictions and approvals for projected structures or antenna installations. Such restrictions could be found to eliminate a site at a very late stage of the process, resulting in much wasted effort. A clear transmission path must exist between the two link nodes of any microwave radio link. Furthermore, as the radio wave disperses as it moves away from the source, there must exist additional clearance over any obstructions to prevent attenuation of the transmitted signal. This additional clearance, known as the Fresnel zone, differs for the frequency band of the radio path, where higher frequency translates into a smaller clearance requirement.
Transmission-Network Deployment
463
LOS between two sites can be confirmed via either map-based studies or direct survey. A path profile is established from topographical maps, which, by reference to the contours of the map, can be translated into an elevation profile of the land between the two sites in the path. The Fresnel zone calculation can then be applied, and an indication of any clearance problems gained taking the Earths curvature into consideration. There are various software tools available to assist this exercise if required, but most are reliant upon the availability of topographical data to the appropriate degree of accuracy being available in digitized format. A path survey can be undertaken via site visits and observing that the path is clear of obstruction. It is important to make note of potential future interruptions to the path, such as tree or foliage growth, future building plans, nearby airports and subsequent flight-path traffic, and any other transient traffic considerations. Which of the routes to establish LOS is utilized is a matter of a particular operators engineering practice and it will depend upon factors such as link length, site location, availability of topographical information, and availability of tools. It is not uncommon to use both techniques for certain links. Again, as planning is an iterative process, if LOS cannot be achieved, this information should be processed back through the network plan and alternative path, calculations, and site selection should be performed. The following is a list of recommended path survey equipment: • GPS navigator; • Extra batteries; • Binoculars, preferably with a built-in compass; • High-quality compass (adjusted for the country in question); • Clinometer (for estimating heights of obstacles); • Altimeter (for measuring heightsGPS height measurements are
not reliable);
• Polaroid or digital camera (the ability to print the date, time, and
number on each picture is essential);
• Walkie-talkies or mobile phones; • Maps (detailed city or countryside maps); • Car with local driver for transportation; • Aluminum ladder (optional).
464
Transmission Systems Design Handbook for Wireless Networks
For longer paths and higher-capacity systems (higher importance), it is imperative to perform the physical path survey and not rely on maps or aerial photographs. 9.6.4
Housing the Infrastructure
Wireless communications equipment comes in all shapes and sizes, and in order to protect that equipment properly, shelters are designed and often custom-built to house it. Different materials are used to construct the shelters, including metal frames, concrete, wood frames, and lightweight insulated materials. Each type of material offers benefits for particular environmental conditions and site locations. Wireless operators require shelters that precisely fit their needs; therefore, shelter manufacturers must be flexible enough to be able to design customized shelters. Strict zoning regulations and the fact that most prime locations are already occupied will require shelter manufacturers to make shelters that can be positioned virtually anywhere or use small cabinets for the wireless equipment infrastructure. A larger cabinet for housing the RBS is shown on the left-hand side, and the small cabinet that can be used for housing the microwave IDU and rectifier with a 30-minute battery backup is shown on the right-hand side of Figure 9.7. Shelters erected on the top of a 40-story building may need to withstand strong gusts of wind similar to those built for hurricane-prone areas. Bullet- and vandal-proof shelters and cabinet designs are also quite often required; all shelters and cabinets must pass certain ballistics requirements as
Figure 9.7 Outdoor equipment cabinets.
Transmission-Network Deployment
465
per Bellcore recommendations. To keep the equipment within a telecommunications shelter in proper working order, it is important to control the temperature levels in the shelter. In colder climates, the insulation is increased, and if the shelter is placed beneath a tower in an area where ice is common, the buildings material must be able to withstand ice falling from the tower. Common problems for many locally produced shelters are rainwater due to leaking roofs and damp walls due to condensation. Grounding with a large buried ground loop that includes many outward-pointing arms is probably the best method, but can be difficult to build in densely urbanized areas. It is important to remember that equipment grounding and protection grounding are separate. Battery backup or diesel generators with long-lasting diesel tanks are required in some areas. It is important to build proper diesel fume exhaust and battery ventilation. It is also important to remember that all batteries produce explosive hydrogen gas. Access roads to sites might require a very large financial outlay. The general requirements on access roads are permanent roads a minimum of 3m wide. The transportation of building material should be feasible without using special permits for long vehicles or traffic police. Prefabricated tower and mast sectors consume considerable transportation space, while other constructions can be taken apart and packed in a very economical manner. Obviously, the shelter must accommodate all equipment and power and occasionally also house a guard. Over the last few years, cell sites as part of wireless network are being built and leased at a record pace. Although some of these sites are in urban areas and some in rural, remote areas, there is always a problem of access to the site. It is a costly problem in terms of time involved in managing keys and access to various sites and because of the potential liability involved. Many companies are installing outdoor cabinets next to towers, which have to be secured differently from shelters and, therefore, add another degree of complexity to the access issue. Probably the best way to provide site access to a variety of different contractors, technicians, and other personnel is to use an electronic-access management system. This system controls access at the user level through the use of an intelligent pager-sized key that is PIN-number controlled. Without a valid PIN number, the key cannot be used. The key is programmed to expire at predefined time intervals. This system also tracks access activity and provides management with a record of all access points the keyholder has visited. Technicians visiting sites will gain access to the lockbox with their keys and use the building keys to service the site. This tracking capability makes keyholders more accountable for their site visits and work.
466
9.6.5
Transmission Systems Design Handbook for Wireless Networks
Microwave Antenna Mounting Structures
9.6.5.1 Overview
Many wireless providers optimistically plan for only a 68-month time period for acquiring, permitting, and building their initial set of transmission facilities in order to launch new wireless service in a community. Often, two or more wireless providers compete for the same scarce friendly sites within the target market, sometimes driving prices up for available antenna (RF and microwave) space. With billions of dollars being spent by wireless carriers for the privilege of operating public frequencies, it is no surprise that the industry is attempting to gain quick approval of new tower sites to begin offering service to the public. And it is understandable when major wireless telecommunication companies sue local governmental agencies for passing moratoria on new tower site applications while they take a look at their applicable regulatory codes. Towers can indeed be horrific structures (see Figure 9.8). Many communities, in consultation with their legal counsel, are developing or modifying zoning ordinances to ensure local review consistent with the requirements of the Telecommunications Act of 1996. Not as often focused on, but in the long run even more beneficial, is strong county or regional planning for the siting of wireless towers. This is best done through a collaborative effort involving all parties interested in the issuepublic and private. Apart from the usual telecommunications towers, some of the other site installations used by wireless providers, and therefore microwave systems, include the following: • Hydro towers: These large gray metal towers hold high-voltage lines.
Usually a square rack is placed at the top of the tower.
• Environmental monopoles: These large poles are painted green to
resemble trees. They usually hold high-voltage lines. A rack may be placed above or below the hydro lines.
• Monopoles: These large poles are painted gray, beige, green, or orange
and white and are exclusively for the use of cellular transmission equipment. Round racks may hold cells out from the monopole, or the cells may be directly attached to the monopole.
• Microwave or radar towers: These large towers are filled with micro-
waves and other transmission equipment.
• Chimney racks: These round racks attach to an industrial chimney to
hold cells.
Transmission-Network Deployment
467
Figure 9.8 Picture of a 150m antenna tower, viewed from below.
• Fancy cement poles: These custom-designed poles hold cellular trans-
mission equipment while looking like art.
• Fake trees: These artificial trees hide wireless transmission equipment
in their foliage.
• Real trees: Cells are mounted to the sides of real trees. • Building flat mounts: Cells are hung over the side of a building.
468
Transmission Systems Design Handbook for Wireless Networks
• Building elevator shaft mounts: Cells are mounted on the elevator
shaft. This mounting may create a shadow of coverage around the building.
• Billboards: These large structures are used for advertising and are
very popular in some parts of the world.
• Building pole mounts: Cells are stuck on little poles attached to the
top edge of a building.
• Building tower mounts: A rack or tower is placed on top of the build-
ing to increase the cell height.
• Lamp standards and flagpoles: Cells are placed on top of and blend in
with the supporting object (usually microcells).
• Intentionally hidden sites: These cells are well hidden by camouflage
or other means usually owing to public concern about the unattractiveness of cell sites.
• Ceilings: Cells are suspended from the ceiling (picocells).
Some of these methods will be discussed in more detail later in this chapter. It is important to remember, however, that the requirements for microwave systems and RF systems are not always the same, and that microwave can require the antenna mounting structure to be built according to more stringent specifications. 9.6.5.2 Role of County and Regional Planning
County and regional planning agencies are well situated to assist communities in making sure that new wireless towers are planned to minimize negative impacts. Given that wireless providers plan their networks from a regional (and broader) perspective, it makes sense for the public to plan for the siting of telecommunications facilities at the same scale, instead of each locality seeking to plan for tower siting independently of neighboring communities. The following are some actions that county or regional planning agencies can take to help ensure that the siting of wireless towers meshes with local (and industry) needs: 1. Provide community educational workshops and forums at which planners, industry representatives, and local residents can discuss and begin to plan cooperatively for the development of wireless networks in their area;
Transmission-Network Deployment
469
2. Conduct a countywide inventory of existing structures suitable for use as antenna support platforms, such as communications towers, buildings 70 ft in height or taller, water tanks, and inactive chimneys (as part of the inventory, also identify existing or planned public facilities and lands upon which antennas might be mounted or towers constructed, such as government centers, public-works operation yards, police and fire stations, surplus highway right-ofways); 3. Classify and prioritize preferred land use areas for new towers, which will require cooperation and input not just from local governments within the county or region, but also from the wireless communications providers; 4. Maintain a central database and map of inventoried existing structures, potentially available public facilities and land, and preferred land-use areas; 5. Have wireless service providers submit and annually update a countywide antenna network plan; 6. Develop criteria for tower siting and design, including preferred construction materials, types and colors, setback requirements, height restrictions, accessory equipment location, fencing, access road criteria, colocation capacity certification, FAA lighting requirements, and ground screening; 7. Develop incentives, such as the following, to encourage good tower design and colocation of towers (i.e., having more than one wireless service provider locate their transmitters on a single tower): tax abatements for stealth or camouflaged towers and an expedited review and approval process for towers proposed within preferred land-use areas, using public facilities, or colocating with other providers. 8. Prepare criteria or a checklist, such as the following, for new tower approval (which can be used at the county or regional level or adapted for local use): Review of site-search ring analysis reports documenting the scope of the applicants search for existing structures or property owners in preferred land-use areas and the rationale for selecting the site under consideration; Review of visual impact analysis, including simulations or digitally reproduced depictions of a virtual tower of like size and type viewed from various locations around the proposed site.
470
Transmission Systems Design Handbook for Wireless Networks
9. Provide planning and engineering assistance to communities, including help with a review of tower applications. 9.6.5.3 Civil Construction
Antenna mounting structures can be towers, poles, tripods, walls, and so on. The following things should be kept in mind during the detailed design and deployment of the MW system: • Sufficient space on the tower to install and pan the microwave
antenna;
• Loading of the antenna mounting structure (MW antenna, trans-
mission lines, and outdoor MW unit);
• Maximum allowed twist and sway (antenna deflection) of the
antenna mounting structure (in degrees), which will depend on the frequency and antenna type.
It is very important that civil construction be handled by experienced civil engineers due to the complexity of existing standards, regulations, and requirements. Input for civil construction dimensioning must come after thorough transmission site surveys. The survey should produce basic requirements, such as tower heights and stability, access road existence, and so on. It is important to recognize the difference between towers and masts. A tower is generally a self-supporting construction with built-in load strength, while a mast (also called a guyed tower) mostly requires supporting wires for carrying a heavy load [7, 8]. It is also important to distinguish between the civil construction and the production of material. And it is important to produce documentation on all levels for civil construction and installation in order to avoid future liability issues. The documentation should include calculations clearly showing how construction requirements are met, including maximum tolerances due to load. Site drawings for electricity, alarms, air conditioning, and so on are also required. In the United States, owners of antenna towers taller than 200 ft (60m) above ground level, or that may intersect the flight pathways of a nearby airport, must register the structure with the FCC and have the structures placement studied by the Federal Aviation Administration (FAA). Previously, antenna registration was a duty of the antenna site licensee, but now (since 1995) it is the owner of the real property who is accountable.
Transmission-Network Deployment
471
Tower lights are another important FAA requirement. Unreported tower-light outages can cost a tower owner a lot of money if they result in a federal fine levied by the FCC. Reducing the potential for a serious aviation accident is enough motivation for a reputable tower owner and communications system licensee to maintain tower lighting in good working condition. The design, supply, and installation cost of a self-supporting tower is approximately twice the cost of an equivalent (between 15m and 76m in height) guyed tower; beyond 76m, the difference in cost is even greater. On the other hand, a guyed tower needs considerably more ground area than a self-support tower, which might be costly in some countries. A large ground area requires considerable fencing, which also can become very costly. Guyed towers are prohibited in some countries (Germany, for example), which necessitates the use of self-support towers or monopoles (Figure 9.9); however, self-support towers are in general more expensive than guyed towers, which can often be built very rapidly and are much lighter as less construction material is required. There is hence a smaller wind load area.
a)
Figure 9.9 a) Self-support, and b) monopole towers.
b)
472
Transmission Systems Design Handbook for Wireless Networks
Steel is extensively used for masts and towers due to its superior strength as compared with aluminum, which requires thicker construction, resulting in higher wind load. All steel constructions are to be hot-dip (zinc) galvanized for corrosion protection, at least. The effect of twisting can be accounted for by proper tower design or by using torsion protection guiding wires for supported masts. The governing principles for tower and mast construction are as follows: • Survival: dimension for maximum probable wind load; which is
typically 50 to 55 mps in Sweden and 50 mps in the United States; there is also a high wind option for up to 70 to 75 mps.
• Operation: dimension according to availability objectives, taking
normal wind load and gusts into account; comparable with microwave availability principles (for example, 99.995% availability).
The operational dimensioning is governed by the maximum allowed antenna deflection according to the requirements calculated by transmission engineers. It is important to note that unavailability due to strong wind load does not normally occur simultaneously with flat fading and, therefore, there is no correlation with flat fading. The general outline for operational dimensioning is as follows: • Determine the wind load profile as a function of height according to
the local characteristics, including safety margins; • Determine the maximum allowed deflection considering wind gusts, including safety margins. A simple formula for estimating the maximum allowed deflection in degrees for microwave transmission antennas is a −10 dB = 60 × l /D [°] where l = wavelength (m);
D = antenna diameter (m). Unfortunately, exaggerated stability requirements are very common. Stability requirements stating maximum deflection of 0.35° to 0.5° at 40 to
Transmission-Network Deployment
473
50 mps prevail without any supporting calculations, resulting in extremely high costs for construction. All external equipment must be properly and thoroughly mounted to avoid vibrations or the buildup of resonance. Resonance is generally not a problem and is easily avoided by designing a stiff construction with a resonance frequency above 10 Hz (low frequencies give the largest amplitudes). Steel might become brittle at low temperatures; therefore, a proper choice of metal alloy in the construction is essential in order to account for temperature-dependent effects. The strain in wires for supported towers typically demands adjustment after six months of operation and regularly every few years. Erosion around the foundation can become a problem if the site has been unwisely positioned. Local knowledge is essential for avoiding the problem of erosion. For example, a site in parts of Africa and the Middle East might seem suitable for several years, but can suddenly be transformed into a river. Soil testing is also essential when building heavy constructions. It is advisable to take several soil samples for analysis. A soil test is a 610m-deep sample giving information about soil composition, profile, and density. The soil test gives an indication as to how to avoid long-term misalignment of the construction due to earth sliding or compression. Unauthorized access to towers and masts can be prevented by removing ladders, installing flat plates around the lower parts of the construction, and using different types of climbing locks, in addition to installing heavy fencing with barbed wire. Many types of fall protection can be used, such as guiding tracks with wire and wheel or simply placing a climbing cage around the ladder. Service platforms and rest platforms are compulsory accessories in many developed countries. The painting of the tower or mast is done mainly for corrosion protection. A compulsory aviation-warning coloring scheme is deployed in some countries. Aesthetic requirements must be considered in other countries. The painting can be very costly, entailing an additional 50% to 100% of the invested steel cost if high-quality paint is used. Different types of aviation warning lights are demanded in various countries, including dusk-activated relays. It is widely accepted that lightning protection can be connected to Earth through the tower and mast construction. A maximum resistance of 10Ω to ground according to British standards is also widely recognized, although in the United States requirements are sometimes more stringent. Local construction standards govern the requirements on towers and masts. These standards are often locally adapted with their origins in British or French construction standards. Quality control of manufactured material
474
Transmission Systems Design Handbook for Wireless Networks
and design are key factors for obtaining a high construction standard and longevity. Minimum recommended tower-strength requirements are published in structural standards or are specified by the customer. In Canada, tower strength and design are not regulated and, therefore, Canadian Standard S37Antenna-Supporting Structures: A National Standard of Canadais not a legal requirement. In the United States, the tower design and construction standard is RS-222Structural Standards for Steel Antenna Towers and Antenna Supporting Structures: An American National Standard. A major difference between the American and Canadian standards is that the Canadian standard specifies mandatory ice and wind loads and the American standard does not. The typical time for erecting a mast with prefabricated concrete weights or earth anchors is less than 5 days. The maximum height difference between each leg is ±1m (±3 ft). The entire mast can be erected with all radio equipment installed at ground level in one or two long sections by using a mobile crane or a helicopter. Alternatively, a hoisting device can be installed at the top of the mast for lifting additional sections and radio equipment manually. The general timelines below typically apply for towers under 40m: Digging of tower foundation:
1 week
Mold and concrete reinforcement:
2 weeks
Concrete filling and hardening:
1 week
Pit backfilling and earth packing with watering: 1 week Tower assembly and erection:
1 week
Total:
6 weeks
Concrete reaches its normal strength after 4 to 5 weeks. However, tower assembly on the ground can start before the foundation is completed, and minor loading of concrete is possible after one week. There are a large variety of country-specific regulations and standards for civil construction. Country-specific standards must be followed, with suitable choice of additional standards according to customer-specific requirements. A common error is to apply demands that are too high and increase construction costs. Some important topics to consider in civil construction are quality control, construction of foundations, concrete reinforcement method and anchor design, quality of paint and welding points, and so on.
Transmission-Network Deployment
475
9.6.5.4 Maximum Allowed Antenna Deflection
Table 9.6 shows 10-dB points (maximum allowed antenna deflection in degrees) for some commonly used frequency bands and antenna sizes.
Table 9.6 MW Antenna Deflection (10-dB Points) Antenna Frequency Diameter Deflection (GHz) (m) −10 dB (°) 02
2.4
3.8
02
3.0
3.0
04
2.4
1.9
04
3.0
1.5
06
2.4
1.3
06
3.0
1.0
07
1.2
2.1
07
2.4
1.1
07
3.0
0.9
08
1.2
1.9
08
2.4
0.9
08
3.0
0.8
13
0.6
2.3
13
1.2
1.2
15
0.3
4.0
15
0.6
2.0
15
1.2
1.0
18
0.3
3.3
18
0.6
1.7
18
1.2
0.8
23
0.3
2.6
23
0.6
1.3
23
1.2
0.7
38
0.3
1.6
38
0.6
0.8
476
Transmission Systems Design Handbook for Wireless Networks
It is important to note that the requirements for the twist and sway of the tower are sometimes much more stringent for the microwave than for the RF installations. 9.6.5.5 Monopoles, Self-Support Towers, and Guyed Towers
Ranging in height from 25 to 125 ft, monopoles consist of a single pole, approximately 3 ft in diameter at the base, narrowing to roughly 1.5 ft at the top. They may support any combination of whip, panel, or dish antennas. Monopoles can be used almost anywherein urban or rural areas, near freeways, or in areas where buildings are not of sufficient height to meet LOS transmission requirements. In the cellular mobile phone system, monopoles are used much more commonly than lattice towers. Monopoles in PCS systems are usually shorter than those of the cellular telephone and ESMR systems. Some PCS providers are even proposing an integration of monopoles into existing light poles. Generally there are two types of towers: guyed and self-supporting. Each type has different limitations regarding its capacity, stability, and versatility. Guyed towers are slender, generally three- or four-sided lattice construction, and are uniform in dimension over their length. They are maintained in their vertical position by guy wires, which are attached at various levels on the tower and anchored to the concrete dead men on the ground. Under normal circumstances, a guyed tower requires a radius of 80% of the height of the tower. This physical characteristic can require a large piece of land (Figure 9.10) to guy a single tower, and so they are usually used in rural areas. Guy stays stretch over time, and the tensions within these assemblies should be adjusted from time to time. The scope of recommended tower inspection and maintenance is published in both ANSI and CSA publications. Self-supporting towers stand on their own (Figure 9.11) and require a base spread of approximately 13% of their height. These towers are relatively inexpensive up to about 100 ft (35m) or so, but become very costly as their height increases. They are used mostly in urban areas where minimal land is available. Short self-support towers are sometimes placed on the rooftops of buildings. For the most part, steel (galvanized after fabrication) is used as the material for the mast construction of towers, and galvanized guy or bridge strand is generally selected for the guy stays supporting the guyed tower. The lattice mast of the tower can be made using a variety of steel shapes and connection types. A common mast makeup for a multiple microwave dish and
Transmission-Network Deployment
477
Figure 9.10 Guyed tower surface requirements.
line support tower would be for angle leg, angle bracing, bolted construction, with a climbing ladder that is inside the mast. This configuration offers a relatively low manufacturing cost, significant strength, high stability, ease of antenna attachment, ease of maintenance, and low shipping costs. On the down side, because of the relatively high, flat (angle) projected area of this type of mast, such a tower and foundation system are more expensive than a comparable tower that uses round members (pipe, tube, or solid round) in its makeup. This price difference becomes smaller when the antenna and line load increases.
478
Transmission Systems Design Handbook for Wireless Networks
Figure 9.11 Self-support tower surface requirements.
The most economical steel section to use for tower mast components is hollow tubes, circular in cross section (pipe or tube). These offer the highest capacity-to-weight ratio as well as lower wind resistance than any other shape. Canadian designers and tower manufacturers have, for the most part, excluded the use of thin wall and small diameter hollow sections for primary structural components (legs) in tower construction for a variety of reasons. Even properly vented sections entrap debris and water in the form of ice over time. The entrapped freezing water expands, exerting pressure on the tube, and aids in the entrapment of debris filtering down the tube, which
Transmission-Network Deployment
479
occasionally blocks the drain holes. Once these holes are blocked, the moisture, along with other chemicals present in it, is contained within the tube for a time. The galvanized surface on the inside of the tube section (which may or may not be thorough, since it is impossible to inspect the entire inside surface of a long slender section) is subject to potential corrosive action depending on the acidity of the water mixture within it. An inspection of such a tower would not reveal any structural distress until a crack or hole appears in the wall of the tube. There are many pros and cons to material selection, connection type (welded or bolted), and mast configuration (position of transmission line brackets and ladder; facility to attach antenna mounts) in tower design. Coupled with potential differences in tower accessories (grounding, safety climbing devices, anticlimb devices, waveguide bridges), proposals, given the same specification, often vary considerably in price. A thorough evaluation of a proposal for tower design and the expertise and reputation of the supplier, together with the price, should be used in the selection of a tower structure. 9.6.5.6 Other Antenna Mounting Structures
The advantage of split-configuration microwave radios is that they can be installed in a very limited space. Poles (Figure 9.12), wall mounts, and tripods are just a few examples of antenna mounting structures that could carry the ODU and antenna. The tripod shown in Figure 9.13 is an example of the nonpenetrating tripod used as a temporary solution for rapid deployment of the microwave system. For penetrating tripods, there is usually a requirement for architectural analysis and approval (since drilling of the sensitive roof structure must take place) before any rooftop installation is allowed.
Figure 9.12 Three versions of the MW radio ODU.
480
Transmission Systems Design Handbook for Wireless Networks
5 8
8
104
Figure 9.13 Nonpenetrating tripod.
Figure 9.14 shows a rooftop installation of the tripod with a microwave antenna. It is important to ensure that nobody walks in front of the antenna, since that could interrupt the traffic on the microwave link. Walking in front of the antenna could also be harmful to a persons health. 9.6.5.7 Minimum-Visual-Impact Structures
The fast-paced growth of wireless communications technology in recent years has presented local governments with the challenge of finding places to locate wireless communications facilities in their communities. Unlike ground-wired telecommunications, such as the land-based telephone system, wireless communications technologies, by their operational nature, require numerous antennas mounted at various heights throughout the landscape. To place them at the specific height required by a particular system, these antennas are sometimes mounted on towers, monopoles, tall buildings, or other structures on top of hillsides. One of the greatest concerns faced by
Transmission-Network Deployment
6 diameter antenna
481
Antenna centerline
4 1/2 OD pipe 96
72
18 12 OD-Outside diameter
Figure 9.14 Tripod with MW antenna.
local jurisdictions is the visual impact of wireless communications facilities. A number of companies today develop and sell minimum-visual-impact and multifunction communication structures. Because of their aesthetically pleasing design, these alternative structures can solve unique communication requirements and enhance the permitting process in difficult zoning locations. Each structure naturally blends into the surrounding environment, and its antennas are hidden from view. Examples of these concealed installations include structures that resemble church steeples, with crosses incorporated into the design of the monopole; lighting structures with curved arms and canisters to conceal antennas; signposts that are two- or three-legged structures with optional signage; and the ultimate designthe minimum-visual-impact tree. The design of the tree pole is site-specific to the location, with careful consideration and configuration given to existing trees; pine and palm tree versions are available as well as a minimum-visual-impact structure that resembles a saguaro cactus. Unfortunately, there is no way to conceal a large microwave antenna which also requires a very strong supporting structure. 9.6.5.8 Fall-Prevention Systems
Communications towers have today become as much a part of the landscape as farm silos and electric power lines. Positioned atop hills and high buildings, they are signs of expanded service for the cellular telephone network
482
Transmission Systems Design Handbook for Wireless Networks
and heralds of the information age. But these structures, held aloft by guy wires and cables, present unique dangers for those who construct and maintain them. Everyone would agree that communications-tower safety is not only a life-and-death issue for the men and women who work on the structures, but also for the companies that own them. Despite the common practice of contracting out the construction and maintenance of towers to other companies, the owners of the structures and now company directors and officers share a legal responsibility with the subcontractor to make sure work is carried out safely. This situation is not a new onemost jurisdictions require property owners to provide safe job sites whether or not they directly employ workers. It is their shape that makes communications towers particularly lethalvery high and very narrow, they offer no second chance in the event of a slip. Construction of these towers requires special skills. Without the benefit of cranes to lower pieces into place from above, the towers are constructed with an apparatus known as a gin pole. The gin pole is used to hoist materials from the ground past the worker at the top so that it can be positioned and bolted in place. The work is carried out at the mercy of wind and inclement weather at heights that reach up to few hundred meters. Typically, however, communications towers range between 90 and 150m (300 to 500 ft) in height. Nonetheless, whether it is from 50m or 500m, the results of a fall are equally fatal. Very stringent work procedures and reliable safety equipment are required to protect workers against accidents. A two-component system has become the industry standard. It consists of a work-positioning harness and a fall-arrest component. The workpositioning harness is used primarily when the worker is stationary. The system consists of a body harness and work lanyard which, when properly secured to the steelwork, allows the worker to lean out from the structure as if in a sling, leaving the hands free to carry out the work. The fall-arrest component consists of the usual arrangement of a trolley, shock-absorbing safety lanyard, and body harness. The trolley glides freely along a permanent rail running the length of the tower, until a sharp movement (such as would occur in the case of a fall) locks it in place. The trolley is attached to a body harness, which resembles that used by parachutists. This system protects the worker from falls while he or she is moving on the structure. For example, the Canada Labour Code requires that towers have permanent ladders with a safety cage or landings at regular intervals. Failing this, they must have a permanent rail-grab fall-arrest system. Since landings and cages are impractical for structures as narrow as communications towers,
Transmission-Network Deployment
483
nearly all towers built today have permanent rail systems. This gives workers relative freedom of vertical movement while still protecting them from falls. But before these rail systems became the norm a few years ago, the only protection most workers had was the body-positioning system that is not very useful while climbing. There were systems that involved ropes attached to the tower, but they proved more a hazard than a life-saving device. Many installers admit that until recently, fall protection of this kind was dangerous. It posed a greater risk to have a line attached to the worker, which might catch on some steel and pull him off the structure, than to let him move freely on the steel and attach himself once he reached his work location. 9.6.5.9 Tower Procurement
The procurement of a tower starts with the customer having a particular need, usually defined by a number of antennas and transmission lines to be located at various heights at a specific geographical location. At the outset, the customer should consider including a provision for future antennas and transmission lines in the procurement specification. It is at this time that it is beneficial for the owner to consider and plan for potential changes or additions to antenna loading that could occur in the future. The costs associated with additional tower strength would be small when compared with the high costs of reinforcing an existing tower and foundations. Generally, if significant, unplanned changes in antenna loads are made after the tower is installed, costly reinforcing will likely be required. If the changes or additions are too extensive, then the tower may have to be replaced with a larger one. The best strategy is to create a specification for tower procurement either with existing staff or by retaining a consultant who specializes in this service, and solicit prices from several tower companies. If this approach is used, the specification for tender should include a request for sketches or drawings of the proposed tower and foundations. Towers from several different sources can vary substantially in size and, therefore, price. Under a competitive bid situation, it is the designers task to make each tower as small as possible, given the antenna and transmission line loading, antenna heights, code, deflection, and other customer requirements. The smallest tower, complete with its associated foundations, usually results in the lowest tendered price. The buyer of tower structures should, therefore, work closely with the radio design and network system planning groups (RF and microwave) in the tower tendering phase. A reputable tower design and manufacturing company is a good source of reliable information regarding budgetary pricing, scheduling, and engineering guidance. Businesses such as these have engineering and technical staff that work closely with purchasing,
484
Transmission Systems Design Handbook for Wireless Networks
manufacturing, and erection teams. Consulting firms that specialize in towers are another source of unbiased, objective information on creating a specification for towers and tower-design loading. These firms are also a source of other professional services, such as design, analysis, and inspection, and they could also check designs to ensure specification conformance if there exists any doubt about the tower proposed or installed. For the microwave application, tower loading of the antenna mounting structures includes antennas, wind, ice, waveguides, coax cables, ODUs, and seismic requirements. The detailed loading requirements should be given to the tower designer, and if climatic conditions are unknown, assumptions must be made based on local statistics. For the most part, it is the physical size and shape of an attachment (antenna, transmission lines, and ODUs) that load a tower, and the actual weight of the antenna seldom bears much significance in calculating the required tower size. For example, on a typical 90m guyed or self-support tower, if all of the antennas and transmission lines weighed just one pound, analysis on such a tower would reveal only marginally lower forces in the towers components and similar deflections as for a benchmark real tower. Suppose on a typical 90m guyed or self-support tower all of the antennas and transmission lines weighed 2,000 pounds, but occupied virtually no space. An analysis on this tower would reveal much lower forces in the towers components and a very different deflected shape than that of the benchmark real tower. There are economies to be achieved by attaching the transmission lines to, or close to, the tower legs, thus limiting exposure to wind and ice loads. The financial savings of this method of attachment may be short lived, however, if a line needs to be unbundled in the future, since associated labor costs will be high. Angle adapters are another alternative, but this can add a substantial and often unnecessary cost. Where practical, the preferred method of attachment is transmission line brackets. These galvanized steel brackets have holes to attach standard stainless steel hangers at every 30 to 60 in on the tower mast and eliminate the need for nylon ties, angle adapters, or stainless steel wraplock and permit the most serviceable, neatest attachment with standard hanger kits. How strong should the tower be? Generally, the minimum recommended requirements for tower strength are either published in structural standards or specified by the customer. In Canada, with one provincial exception, tower strength and design are not regulated; therefore, Canadian Standard S37 (Antennas, Towers, and Antenna-Supporting Structures: A
Transmission-Network Deployment
485
National Standard of Canada) is not a legal requirement. In the United States, the standard for tower design and construction is RS-222D (Structural Standards for Steel Antenna Towers and Antenna Supporting Structures: An American National Standard ). These standards are constantly under committee review and are revised and reissued from time to time to reflect the current knowledge of loading, analysis techniques, materials, and workmanship. The content is consistent with other national standards that are referenced throughout these documents. One major difference between the American and Canadian standard is that the Canadian standard specifies mandatory ice and wind loads. The American standard does not require ice loading as mandatory for any state. Some manufacturers produce a commercial, over-the-counter type of tower that is designed according to criteria other than these two national standards. Such a product can provide an owner with a low-cost product for a specific and usually limited application. The incidence of failure tends to run higher with structures that do not conform to either the CSA or ANSI standards, and it is advisable to consider the risk and consequences of failure when selecting this kind of product. The customer has to be careful because these types of towers have been designed to survive a specific and uniform wind velocity with no ice loading and have little or no safety factors beyond that loading.
9.7 Measurement of Radio-Frequency Fields 9.7.1
Health and Safety Issues
Possible health risks related to electromagnetic fields (EMFs) and radiofrequency radiation (RFR) are another major source of local community concern with wireless communications facilities. To date, scientific research on the effects of wireless communications facilities on human health has been inconclusive. Locating communications antennas and towers in or near residential communities often becomes a subject of local controversy, often due to personal health and safety concerns voiced by local residents. When reviewing and considering permits for wireless communications facilities, planners and decision makers alike must be able to respond to local citizen concerns about health risks associated with them. The ANSI and IEEE have established standards for safe human exposure to radio-frequency electromagnetic fields. These standards are considered consensus standards, which are agreed to by committees composed
486
Transmission Systems Design Handbook for Wireless Networks
of academic, industry, and governmental representatives. As a condition of licensure, the FCC requires all cellular, ESMR, and PCS providers to comply with the ANSI standards. Noncompliance may result in revocation of an FCC license. Federal exposure standards for EMF and RFR levels are being developed by the U.S. Environmental Protection Agency, which currently uses the ANSI exposure standards as guidelines. Absent federal standards, the ANSI/IEEE exposure standards are currently the most appropriate health and safety guidelines for wireless communications facilities, and should be incorporated into local review and approval requirements. In 1999 wireless industry trade associations around the world joined to form a new organization, the Wireless Information Network (WIN); this group focuses on health and environmental issues related to the wireless industry. Established at a meeting in London, WIN includes industry representatives from Australia, Austria, Canada, Denmark, France, Germany, Ireland, Italy, New Zealand, Norway, Sweden, the United Kingdom, and the United States, and was initiated jointly by the CTIA and the United Kingdoms Federation of the Electronics Industry (FEI). WIN includes the following industry groups: the Australian Mobile Telecommunications Association, Austrias Forum Mobilkommunikation, the Canadian Wireless Telecommunications Association, Denmarks Association of Telecommunications Hardware Suppliers, Germanys FGF, the New Zealand Telecommunications Organization, Norways Elektronikkbransjen, the United Kingdoms FEI, the United States CTIA, the Mobile Manufacturers Forum (an international manufacturers group), and the GSM Association (an international group of service providers). Three individual companiesFrances Telecom/CNET, Irelands Eircell, and Italys Telecom Italia Mobileare also members, with the understanding that each will aim to represent its countrys wireless industry as a whole.
9.7.2
Measurements and Sources of Emission
The procedures presented may be used for the following: • Measurement of radiating EMF; • Measurement of leakage and reradiated EMF; • Measurement of induced EMF in the body.
Transmission-Network Deployment
487
Sources of emission in the present context refer to the different types of radio-frequency transmitters employed in the different telecommunication services. These transmitters may exhibit very different spectral, spatial, and temporal characteristics due to the nature and the requirements of each type of service. The recommended procedures take into consideration the normal features and the circumstances of each type of service and the characteristics of the transmitters and radiation patterns. Frequency Range 3 kHz to 300 MHz
Services in this frequency range include maritime navigational communications, aeronautical radio navigation and radio communication, analog AM radio broadcasting, shortwave broadcasting, land-mobile communication and fixed services, VHF radio (FM), and television broadcasting and amateur radio communication. Measurement procedures and techniques over this frequency range vary according to the frequency and type of service. In general, for services below 300 MHz, measurements of both the electric (E) fields and the magnetic (H) fields may be required. In addition, in the case of some high-power transmissions (e.g., AM radio service), measurements of induced current and contact current may also be required. Frequency Range 300 MHz to 300 GHz
Services in this frequency range include UHF television and digital radio broadcasting, fixed, land-mobile/PCS, and satellite systems. Over this frequency range, the wavelengths of the electromagnetic fields and the dimensions of the antenna are relatively short, measurement locations are usually situated in the far-field region, and in general, only electric-field measurements are required. In the far-field region, the magnetic field and the electric field are related by a constant. In this case, measuring only the |E|² component can approximate the power density. Source Parameters and Modulation
Radio-frequency electromagnetic sources radiate energy into space through antennas installed on towers and buildings. These sources have widely different characteristics and thus require proper selection of instrumentation in hazard determination. Transmitted electromagnetic waves may have various forms. The most fundamental form is a continuous wave (CW) or unmodulated carrier in which the wave oscillates at a single frequency. Another signal or message may modulate such carriers. When a CW wave is modulated by pulsing, or by varying its amplitude, frequency, or phase, the wave is called a pulse-, amplitude-, frequency-, or phase-modulated wave, respectively.
488
Transmission Systems Design Handbook for Wireless Networks
9.7.3
Near-Field and Far-Field Regions
The space around a radiating antenna can be divided essentially into two regions, the near-field and far-field regions. For an antenna with a maximum overall dimension that is small compared with the wavelength, the near-field region is mostly reactive, and the electric- and magnetic-field components store energy while producing little radiation. This stored energy is transferred periodically between the antenna and the near field. The reactive near-field region extends from the antenna up to a distance R. R = l /2 p where l = the wavelength. There is no general formula for estimation of the field strength in the near field for small antennas. Exact calculations can be made only for welldefined sources such as dipoles and monopoles (antenna types). For antennas large in terms of wavelength, the near-field region consists of the reactive field extending to the certain distance followed by a radiating region. In the radiating near field, the field strength does not necessarily decrease steadily with distance away from the antenna, but may exhibit an oscillatory character. The criterion commonly used to define the distance from the source where the far field begins is R = 2a 2 / l where a = the greatest dimension of the antenna. For a paraboloidal circular cross-section antenna, a realistic estimate for R, which provides close agreement with experimental results, can be obtained using the following relationship: R = 0.5a 2 / l where a = the antenna diameter. In the radiating near field, the electric field strength E and magnetic field strength H are interrelated with each other as
Transmission-Network Deployment
489
h = E /H The power density S is S = E 2 /h = H 2 h where h = the intrinsic impedance.
The value of h may vary with the distance in the near-field region. In the far-field region, the field has a predominantly plane wave character (i.e., the electric-field vector is perpendicular to the magnetic-field vector, and they are both transverse to the direction of propagation). The ratio of the electric- field strength to the magnetic-field strength is constant at any location, and in free space it is equal to E /H = h = 377Ω 9.7.4
Power Levels and Power Density
Radiated power is frequently expressed in decibels above 1 mW (dBm) or 1W (dBW) reference power levels. Depending upon the type of service and source, the range of typical power radiated by transmitting antennas is from under 1W or 0 dBW (e.g., portable transmitters) to over 100 kW or 50 dBW or higher (e.g., radars, VLF transmitters). For safety and efficiency, it is important to have the information on the radiated power prior to taking measurements. For antennas with reflectors, such as parabolic dishes, the maximum power density (within the antenna beam) in the radiating nearfield region can be conservatively estimated as S = 4 P a /A where Pa = power into the antenna; A = physical aperture area. In the far-field region, the power density on the antenna axis can be calculated from the following expression: S = P a G /4 pr 2
490
Transmission Systems Design Handbook for Wireless Networks
where r = distance from the antenna; G = antenna directive gain. The directive gain of an antenna in a given direction is 4p times the ratio of the radiation intensity in that direction to the total power radiated by the antenna. The antenna gain is related to the antenna dimensions by the following equation: G = 4 pA e /l2 where Ae = effective area of the antenna, pA; A = physical surface area on the antenna; p = antenna efficiency; l = wavelength. It should be noted that the effective area of some antennas (e.g., linear arrays) must be derived by other means, since the physical area may not be easily determined. The free-space electric-field strength (rms value) at a distance r, from a source with effective radiating power Pe (the source average output power multiplied by the antenna gain) on the antenna axis, is equal to E =
30P e r
where E is expressed in volts per meter (V/m). 9.7.5
Radiation Patterns and Polarization
Electromagnetic waves are radiated into space by means of antennas. The radiation pattern of an antenna determines the spatial distribution of the radiated energy. A pattern taken in the plane containing the electric-field vector is referred to as an E-plane pattern. A pattern taken in a plane perpendicular to an E-plane is called an H-plane pattern. The directional pattern of an antenna describes how much it concentrates energy in one direction in preference to radiation in other directions. In the near field, the radiation
Transmission-Network Deployment
491
pattern of an antenna changes with distance from the source, whereas in the far field no significant change with distance occurs. The orientation of an electric-field vector in the plane orthogonal to the direction of propagation is called polarization. If the electric-field vector is always oriented in a given direction, the wave is linearly polarized. If the electric-field vector rotates around the direction of propagation, maintaining a constant magnitude, the wave is circularly polarized. If the extremity of the electric-field vector traces an ellipse, the wave is elliptically polarized. The rotation of the electric-field vector occurs in one of two directions, either clockwise or counterclockwise. It is difficult to predict the orientation of the electric field in the near-field region, as the transmitting antenna cannot be considered as a point source in this region. In the far-field region, the antenna becomes a point source, the electric and magnetic components of the field become orthogonal to the direction of propagation, and their polarization characteristics do not vary with distance.
9.7.6
Sources of Electromagnetic Field
At a measurement survey site, there may be a single or several sources of electromagnetic fields. A single source may have strong harmonic content that can produce electromagnetic fields at multiple frequencies. In addition, several types of RF sources, such as AM, FM, TV, land-mobile, and fixed transmitters, may commonly be installed on an antenna farm or multiple-use tower and can produce a complex electromagnetic environment. In these situations, it is difficult to estimate the maximum expected field levels. Both broadband and narrowband instrumentation should be employed to characterize the electromagnetic environment fully. At many transmitting sites, there may be unexpected radiation leakage emanating from electronic equipment (e.g., power amplifiers), a crack in the shielding cabinet or conduit, a joint between transmission cables or sections of waveguide. These leakages can result in localized hot spots with the electromagnetic fields in excess of the exposure limits. The nature of the leakage fields is similar to that of the near field around an antenna. Therefore, any type of polarization may exist in the vicinity of the leak location. There is no reliable method to predict the extent of the leakage radiation, or the type of the field produced (reactive or radiating). In general, the location of the leak is not known and may only be detected by trial and error. Although many types of instruments are available for field measurements, those that have isotropic characteristics are generally better suited to probe radiation leakage.
492
Transmission Systems Design Handbook for Wireless Networks
RF electromagnetic energy from an active radiator induces electric charges or currents on ungrounded or poorly grounded conducting objects such as metal flagpoles, signposts, window frames, fences, and walls of metallic buildings. The amount of the induced current depends on the physical characteristics of the object (size, shape, orientation with respect to the source) and the frequency of the incident field. This current produces its own electric and magnetic fields in close proximity to the object. The produced fields, which are generally reactive, interact with the incident field and may result in so-called hot spots or enhanced E and H fields close to the object surface. Since the conducting objects act as secondary radiators when exposed to ambient RF fields, they are sometimes referred to as passive or parasitic reradiators. The enhanced fields generally diminish to the ambient levels in the surrounding areas within very short distances of the secondary source. Field-strength reduction is generally exponential, with the highest strengths on the surface of the reradiating object. The enhanced fields are highly nonuniform in their spatial distribution on the reradiating object and are generally difficult to predict by theoretical methods. Hot spots are best evaluated by measurements. 9.7.7
Induced and Contact Currents
An RF field induces an alternating electric potential on ungrounded or poorly grounded conducting objects. When a person touches such objects, RF current flows through the persons body to the ground. This type of current is known as contact current. Even though a person may not be touching a metallic object, RF current that is induced in the body by RF fields may also flow through the body to the ground. This type of current is referred to as induced body current. Modest levels of these RF currents may cause perception, while higher values may result in shock or burns. The 1999 version of Canadian Safety Code 6 includes recommended limits for both contact and induced currents in the frequency range from 3 kHz to 110 MHz, with the intention to reduce the potential for shock or burns. Under certain exposure conditions, the contact and induced currents will be evaluated as they may exceed the limits, even though the field-strength limits are not exceeded. These conditions may occur when the electric-field strength is as low as 20% to 25% of the exposure limit. Specific Absorption Rate (SAR) is the rate of RF energy absorption per unit mass in the body. SAR has units of joules per second per kilogram or watts per kilogram (W/kg). This parameter is used as a primary indicator of
Transmission-Network Deployment
493
RF energy absorbed in the body when quantifying the biological effects and thus defining basic exposure limits. At frequencies between 100 kHz and 10 GHz, SAR limits take precedence over field strength and power density limits and must not be exceeded. When carrying out compliance evaluation, the SAR should be determined for cases where exposures take place at a distance of 0.2m or less from the source. For conditions where SAR determination is impractical, field strength or power density measurement is carried out. 9.7.8
Instrumentation
A typical RF field or power density measurement device is composed of a probe, leads, and metering instrumentation. The probe is used to detect the field. It can either be a conventional antenna or another type of sensor. The performance and the application of the measuring instrument as a whole depend to a large extent on the design and characteristics of the probe. The leads carry the detected signal to the metering instrumentation. To reduce the coupling of the leads with the surrounding field in order to minimize any disturbance, the leads take the form of high-resistance wires. The metering instrumentation is primarily designed to process and display received field density. The RF measurement device may be either broadband or narrowband. A broadband device responds uniformly over a wide frequency range and requires no tuning. A narrowband device may also operate over a wide frequency range, but the reception bandwidth is narrow, and the device must be tuned to the frequency of interest. Narrowband and wideband devices have their own advantages and disadvantages depending on the spectral environment and the type of measurements that are projected. Electric- and magnetic-field strength meters are narrowband devices. They consist of an antenna, cable(s) to carry the signal from the antenna, and a signal conditioning and readout instrument. Field-strength meters may use linear antennas, such as monopoles, dipoles, loops, biconical or conical log spiral antennas, horns, or parabolic reflectors. The appropriate field parameters can be determined from a measurement of voltage or power at the selected frequency and at the antenna terminal. The electric- (or magnetic-) field strength can be derived from information on the antenna gain or antenna factor and the loss in the connecting cable. Spectrum analyzers are essentially broadband tunable receivers whose reception bandwidth may be set over a wide range of frequencies. They are used to measure the power at the antenna terminal at the selected
494
Transmission Systems Design Handbook for Wireless Networks
frequencies. If used in combination with a narrowband selective antenna, the overall device becomes in concept similar to a field-strength meter. However, spectrum analyzers can also be connected to relatively short antennas to produce a broad response over a given frequency range. In this case, the analyzer will display the spectrum of ambient signals and, thus, will permit the ascertaining of frequencies involved and their relative contribution to the overall power density. It is difficult to make accurate, broadband E- and H-field probes that cover the long wavelength (1,000m) region, using the conventional means described above. In order to provide a flat frequency response and adequate sensitivity in a dipole probe, the load impedance of the detector and the high-impedance lead in combination must be greater than the antenna (source) impedance. One solution is to provide a high-impedance RF buffer amplifier that is connected directly to a monopole or loop antenna and which acts as the load. This is practical for frequencies between 10 kHz and several hundred MHz. Commercially available magnetic- and electric-field probes, using active electronics, operate at frequencies as low as 60 Hz. A second problem associated with probes without active electronics is that of isolating the signal-carrying leads from the antenna-detector combination. This problem may become severe below about 100 MHz, and particularly below 10 MHz. This is due to the fact that the typical highresistance signal-carrying lead serves as a low-pass filter, and its ability to separate the low-frequency detected signal from the RF field being measured becomes more difficult as the two frequencies approach each other. This results in excessive sensitivity and poor antenna patterns in passive probes. Finally, at frequencies above about 300 MHz, where free-space or uniform irradiation conditions exist, both the sensor and the metal enclosure of the survey instrument can be exposed to similar levels of RF, and scattering from the enclosure to the sensor (probe) can cause significant errors. Active electronic probes eliminate the use of such leads entirely by including the visual display (readout) with the metal box containing the active electronics. A fiber-optic data link can be provided for a remote readout. 9.7.9
Measurement Procedure for Microwave Installations
Only qualified personnel who have a good understanding of electromagnetic radiation and communication systems should perform field measurements. To minimize measurement errors, refer to the survey-meter manufacturers guidelines regarding the following:
Transmission-Network Deployment
495
• Environmental conditions appropriate for survey-meter use (e.g.,
minimum and maximum limits on temperature, humidity, and atmospheric pressure to maintain probe accuracy);
• Precautions to be taken in the handling of probes to minimize lead
pickup and effects of the surveyors body (e.g., probe to be held out at arms length facing radiator or at a right angle to the radiator, while probe cable is moved to see if reading is affected);
• Symptoms of survey-meter overload and precautions to be taken to
avoid overload;
• Uncertainty of measurements taken in the presence of reflecting
objects and multiple radiating sources.
The procedures that assume near-field conditions and metering are 2 2 2 2 capable of indicating both E and H in units of V /m and A /m , respectively. Conversion errors may be encountered using metering in the near field while displaying power density (which assumes far-field conditions). More information about health hazards of the radio-frequency EMF, measurements, and safety criteria can be found in [912].
9.8 Fiber-Optic Cables and Their Installation 9.8.1
Cabling Design Considerations
There is a very small chance that the transmission engineer working on the design and deployment of the wireless network will get involved in the details of the fiber-optic network design and installation. Having said that, interfacing the existing fiber-optic network at the switch office location and hub sites is very likely, and therefore the engineer will require some basic knowledge of practical issues in connection with bringing fiber-optic cables into the building, terminating them, and interfacing (and cross-connecting) them with other types of equipment. Every fiber-optic cable requires proper techniques for a successful installation. Building codes and standards, environmental issues, proper design, routing, equipment for installation, topologies, applications, and reliability concerns are addressed. Considerations of tensile strength, ruggedness, durability, flexibility, size, resistance to the environment, flammability, temp- erature range, and appearance are important in constructing optical fiber cable.
496
Transmission Systems Design Handbook for Wireless Networks
As protection against system downtime, many systems employ redundant devices (redundant electronics) coupled to fibers in the same cable. The redundant system takes over immediately upon failure of the primary system. This protects against active device failure, although it does not help in the rare instance of a complete cable cut. The ultimate protection against system downtime is achieved through redundant routing. Redundant fibers are placed in a second route to take over immediately in the event a cable is damaged. Redundant routing should be considered when zero downtime for the physical cable plant is required. The NEC identifies three intrabuilding regions with regard to cable placement: 1. Plenums; 2. Risers; 3. General-purpose areas. The plenum area is a compartment or chamber that forms part of the airdistribution system and to which one or more air ducts are connected. A room with a primary function of air handling is also considered a plenum space. The riser is an opening or shaft through which cable may pass vertically from floor to floor in a building. General-purpose areas are other indoor areas that are not plenums or risers. Cables are specifically listed for use in each of these. The NEC does allow the use of a cable with a more stringent listing in an application requiring a lesser listing, but not the other way around. When a cable run will be exposed to both indoor and outdoor environments, there are several options to consider. The first option is to run an outside plant cable for the entire run. The second option is to use an indoor-outdoor cable. The third option is to transition splice the outside plant cable to an inside plant cable. The first option is based on NEC restrictions for bringing unlisted cables into a building. As described earlier, unlisted cables must be placed into an acceptable duct for their entire run inside the building. A transition splice will slightly increase the fiber-link loss and may add some cost to the installation due to the additional splice point. These costs have to be weighed against the cost (as well as pulling difficulties) of rigid, metallic conduit or an alternate conduit installation that would be required to install unlisted cable to the desired termination point. The best option is to use a dedicated indoor-outdoor cable. These flame-retardant loose tube cables eliminate the need for both flame-retardant inner ducts and transition splices.
Transmission-Network Deployment 9.8.2
497
Fiber Protection
The optical fiber is a very small waveguide, and in an environment free from stress or external forces, this waveguide will transmit the light launched into it with minimal loss or attenuation. To isolate the fiber from these external forces, two first-level protections of fiber have been developed: loose tube and tight buffer. In the loose tube construction, the fiber is contained in a plastic tube that has an inner diameter considerably larger than the fiber itself. The interior of the plastic tube is usually filled with a gel material. The loose tube isolates the fiber from the exterior mechanical forces acting on a cable. For multifiber cables, a number of these tubes, each containing single or multiple fibers, are combined with strength members to keep the fibers free of stress and to minimize elongation and contraction. By varying the number of fibers inside the tube during the cabling process, the degree of shrinkage due to temperature variation can be controlled, and therefore the degree of attenuation over a temperature range is minimized. The other fiber-protection technique, the tight buffer, uses a direct extrusion of plastic over the basic fiber coating. Tight-buffer constructions are able to withstand much greater crush and impact forces without fiber breakage. The tight-buffer design, however, results in lower isolation for the fiber from the stresses of temperature variation. While relatively more flexible than loose buffer, if the tight buffer is deployed with sharp bends or twists, optical losses are likely to exceed nominal specifications due to microbending. Each construction has inherent advantages. The loose-buffer tube offers lower cable attenuation from microbending in any given fiber, plus a high level of isolation from external forces. Under continuous mechanical stress, the loose tube permits more stable transmission characteristics. The tight-buffer construction permits smaller, lighter-weight designs for similar fiber configuration, and generally yields a more flexible, crush-resistant cable. Once a tight or loose buffer construction is selected, the system designer makes decisions regarding the tradeoffs between microbending loss and flexibility in obtaining optical operation goals. For installation of a cable, mechanical properties, such as tensile strength, impact resistance, flexing, and bending, are important. Environmental requirements concern the resistance to moisture, chemicals, and other types of atmospheric or in-ground conditions. Normal cable loads sustained during installation may ultimately place the fiber in a state of tensile stress. The levels of stress may cause microbending losses that result in an attenuation increase and possible fatigue effects. To transfer these stresses, loads in short-term installation and long-
498
Transmission Systems Design Handbook for Wireless Networks
term application, various internal strength members are added to the optical cable structure. Such strength members provide the tensile load properties similar to electronic cables, and keep the fibers free from stress by minimizing elongation and contraction. In some cases, they also act as temperaturestabilization elements. Optical fiber stretches very little before breaking, so the strength members must have low elongation at the expected tensile loads. Impact resistance, flexing, and bending are other mechanical factors affecting choice of strength members. Strength members that are typically used in fiber-optic cable include aramid yarn, fiberglass epoxy (FGE) rods, and steel wire. Pound for pound, aramid yarn is five times stronger than steel. It and FGE rods are often the choice when all-dielectric construction is required. Steel or FGE should be chosen when extreme cold temperature performance is required, since they can offer better temperature stability. For outdoor applications, fiber-optic cables are configured with a single jacket, two jackets, or armor for aerial, underground duct and direct-burial installation. Fiber-optic cables for indoor applications meet the requirements of the plenum, riser, and vertical tray cable specifications of the NEC/CSA. If future cable pulls in the same duct or conduit are a possibility, the use of inner duct (smaller duct within the duct) to sectionalize the available duct space is recommended. Without this sectionalization, additional cable pulls can entangle an operating cable and could cause an interruption in service. Care should be exercised to ensure that the inner duct is installed as straight as possible, without twists that could increase the cable-pulling tension. When the cable is installed in raceways, cable trays, or secured to other cables, consideration should be given to movement of the existing cables. While optical-fiber cable can be moved while in service without affecting fiber performance, it may warrant protection with conduit in places exposed to physical damage.
9.8.3
Fiber-Optic Cable Types
Two main types of optical fibers are single-mode fiber and multimode fiber. Multimode was used more often in the past, but is being replaced almost completely with single-mode fibers. Single-mode fiber [13] allows only one mode of light to propagate, thus eliminating the main limitation to bandwidth: modal dispersion. The small core of a single-mode fiber, however, makes coupling light into the fiber more difficult, so that lasers must be used. The main limitation to the bandwidth of the single-mode fiber is chromatic dispersion.
Transmission-Network Deployment
499
FO cables are designed to provide optimum performance over their service life. Other factors, such as ease of installation and termination, should also be considered. When selecting an FO cable, a number of factors must be evaluated. Outside plant cables must be capable of withstanding a variety of environmental and mechanical extremes. The cable must offer excellent attenuation performance over a wide range of temperatures. Water-blocking capabilities must be provided to ensure that water cannot migrate into the cable and subsequently freeze. The cable must be sufficiently strong to endure the rigors of installation and must provide protection against ultraviolet (UV) radiation, gnawing rodents, and other mechanical disturbances. Furthermore, the cable should have a high packing density to maximize the use of available installation space. Unlike outside plant cables, inside plant cables generally experience a controlled, stable environment. Therefore, the performance requirements are based on other factors. The cables must meet the requirements of the NEC and local building codes based upon their installed location (general purpose, riser, plenum). The cables should be easy to terminate and must be available in the high fiber counts required by the network architecture. Indoor/outdoor cables are flame-retardant loose-tube cables using materials that enable them to pass the flame-retardant requirements of the indoor environment. Because these cables are flame-retardant, they are suitable for use within the building. At the same time, the rugged loose-tube construction allows the cable to be used in the harsh outdoor environment.
Indoor/outdoor cables.
Tight-buffered cables are designed for use in building backbones, horizontal applications, and patch cords and equipment cables. Tight-buffered cables contain optical fibers that are directly coated with a thermoplastic buffer to a diameter of 900 mm. Tight-buffered cables are desirable for intrabuilding applications because of their ability to meet building fire-code requirements, as well as their increased physical flexibility, smaller bending radius, and easier handling characteristics in low fiber counts. These cables, however, are typically more sensitive to temperature extremes and mechanical disturbances than loose-tube cables, and they are not waterproof. As a result, tight-buffered cables are not recommended for outside plant (interbuilding) applications. Tight-buffered cables.
All-dielectric outside plant cables. All-dielectric outside plant cables are for duct applications and aerial applications where rodents are not a threat. All-
500
Transmission Systems Design Handbook for Wireless Networks
dielectric cables have no conducting materials within the cable. The exclusion of conductive components eliminates the need for grounding or surge suppression at the building entrance. These cables can be run parallel to electrical service cables without danger of induced voltages and in the vicinity of power lines of up to 115 kV. Special cables and sheaths include limited-smoke, zero-halogen loosetube cable designed for use in outdoor cable trays (e.g., industrial sites) and aerial applications, in addition to indoor applications. This cable complies with IEEE-383 flame test requirements. Additionally, the cable is listed as type OFN-LS. The exclusion of halogens in the cable limits the production of corrosive gases in the event of combustion. Standard outside-plant loosetube cables are available with special outer sheaths of nylon or Teflon to provide superior chemical resistance for the cable. Loose-tube cables are also available for special applications, such as submarine environments. Patch cords (jumpers). Patch cords, or jumpers, are used to connect the patch panels or outlets of an optical-fiber cable plant with the actual electronic devices. They are also used for cross-connect applications in patch panels. Patch-cord cables are also available for all building applications: plenum, riser, and general purpose. Pigtails (cables with an optical-fiber connector on one end only) are used for terminating cables. A fiber-optic pigtail is typically a one- or two-fiber cable that has been connectorized on one end with a fiber connector. The other end remains unterminated. This unterminated end is spliced to the cable that requires termination. This splice can be either a fusion or mechanical splice. After splicing, the splice point is protected by a splice tray and placed into terminating hardware.
Aerial fiber-optic cable is exposed to many external forces. Wind and ice loading can put stresses on cables, and temperature changes are more extreme than in any other installation environment. Furthermore, vehicular accidents, rodents, lightning, and even gunshots can damage aerial cables. Some guidelines for aerial FO cables follow: Aerial fiber-optic cable.
• Aerial pole spans of less than 300 ft can be routinely engineered
when a dedicated messenger wire is used.
• In some cases, optical-fiber cable can be overlashed to the existing
aerial plant.
• Cable is also available in self-supporting designs not requiring the
use of a messenger.
Transmission-Network Deployment
501
• The use of tight-buffered cables in an aerial installation is not rec-
ommended. The external factors, specifically ice and wind loading, can place the fibers in tight-buffered cables under constant stress. This can result in fiber failure or significant increases in attenuation over time.
Direct buried cable experiences fewer environmental extremes than aerial cable. This cable is usually laid into a trench or plowed into the soil. Primary dangers to underground cable are dig-ups and rodent damage. Rodent-resistance of the cable is provided by means of a coated steel tape armor. Installation in a duct offers additional protection from rodents, as well as dig-ups. Some guidelines for buried fiber-optic cables follow:
Direct buried cable.
• It is recommended that the cable be buried as deep as possible, at • • • • •
least 30 inches underground, if possible. The cable should be located below the frost line to prevent damage due to frost heave. When planning a direct buried cable route, locate and avoid existing utilities. Plastic duct with an outside diameter greater than 1.5 inches will also provide rodent-resistance where armor cannot be used. A means of locating the cable, especially all-dielectric cable, should be provided. The use of tight-buffered cables in direct buried installations is not recommended. These cables are typically not water-blocked and are susceptible to water penetration and ensuing forces.
Armored cables are recommended for use when the cable will be direct-buried in rocky soil or when rodent protection is required. The coated steel-tape armor enhances the mechanical performance of the cable. Armored cables.
9.8.4
Fiber Count
The selection of the fiber count, or number of fibers used in the cable plant, is an extremely important decision that impacts both the current and future capabilities, as well as the cost, of a communications network. The development and widespread use of fiber in all aspects of the network require the designer to plan not only for the immediate system requirements, but also
502
Transmission Systems Design Handbook for Wireless Networks
for the evolution of future system demands. Depending on the numbers and types of applications in the network and the level of redundancy needed, fiber counts can range from 2 to more than 100 in the backbone or to each wiring closet. Currently, due to the expense of multiplexing equipment, separate, dedicated fibers are typically utilized for each application. So, for example, if two buildings were to be networked with an FDDI backbone, four fibers would be required in the cable connecting the buildingstwo to transmit, two to receive. Further, it is recommended that at least two times the number of fibers needed actually be placed in the backbone to accommodate expansion requirements. Although some systems clearly indicate the number of fibers needed, there are usually no hard and fast rules. Installing the required number of fibers, plus others for backup and for the future, yields a more flexible, expandable cable plant to service networking requirements into the future. Users should plan for growth in the backbone network, especially the interbuilding backbone. Extra fibers can be installed and left unterminated until needed. The cost of these unterminated spare fibers to the overall project is minimal. The major expense of cable installation is the labor associated with the actual cable placement and termination, especially in difficult installations with congested or inaccessible raceways and conduits. The incremental cost of installing extra fibers now will be much less than the cost of installing a new cable later. While a user can purchase a cable with almost any number of fibers, cable manufacturers and cable distributors consider certain fiber counts standard and often stock these cables. Fiber counts in multiples of 6 and 12 are standard for backbone cabling and connecting hardware. For low fiber counts (<24), multiples of 6 are common (i.e., 6, 12, 18, and 24). For higher fiber counts (>24), multiples of 12 are common (i.e., 36, 48, 60, and 72). This should be considered when determining the fiber counts for the system. The number of optical fibers required is dependent upon the following: • Intended end-user application(s) requirements; • Level of electronics integration in the proposed network; • Future growth of the network; • Physical network topology; • Application requirements.
Most of todays fiber-optic system applications fall into three categories:
Transmission-Network Deployment
503
1. Simplex: one fiber for one-way communications; 2. Duplex: two fibers, one to transmit from a location and one to receive at a location; 3. Redundant duplex: four fibers, two to transmit from a location and two to receive at a location using separate duplex routing. Most of todays applications are duplex, utilizing two fibers to establish twoway communicationsone fiber to transmit and the other fiber to receive. It is also possible to transmit and receive simultaneously using the same fiber. This is normally accomplished through the use of optical splitters installed into the passive portion of the fiber system. Utilization of this form of communication requires additional electronics to transmit and receive each signal. 9.8.5
Fiber Splicing
In preparing fibers for splicing, the jacket materials, strength members, and buffer tubes are cut to the appropriate lengths, and the fiber buffer coatings are removed. Once this has been done, the fibers must be cleaved in preparation for splicing. Cleaving is a method of breaking a fiber in such a way as to create a smooth, square end on the fiber. Glass is typically strong until a flaw occurs and creates a region of high stress under pressure. The first step in the cleaving process is to create a slight flaw or scribe in the outer surface of the fiber. Optical fibers can be scribed with a sharp blade made of a hard material, such as a diamond, ruby, sapphire, or tungsten carbide. The scribe is made by lightly touching the cleaned fiber, at a right angle, on the desired cleave point with a scribing tool. Only the lightest pressure is required to produce a scribe if the blade is sharp. It is very important not to use a sawing motion. A crude or slanted scribe will produce shattered or scalloped end surfaces. After the scribe is made, a straight pull will produce the cleanest break. If bending accompanies pulling, a square break is less likely, especially with large fibers. Dispose of broken fiber pieces by placing them on a piece of tape. Wearing safety glasses when working with optical fibers is mandatory. The level of quality required for a given cleave depends on the application. For fusion splicing, mechanical splicing, and some connector systems, cleaves must be nearly perfect. Some connector and splicing systems use cleaving to produce the final end surface on the fiber (no subsequent grinding or polishing). However, during quick continuity checks with a flashlight,
504
Transmission Systems Design Handbook for Wireless Networks
less than perfect cleaves may be acceptable. A hand microscope is useful for quick checks of cleave quality. Cleaving tools are available in inexpensive hand models or more sophisticated mechanized tools. There are two basic types of splices: fusion splices and mechanical splices. Fusion splices are made by positioning cleaned, cleaved fiber ends between two electrodes and applying an electric arc to fuse the ends together. A perfusion arc is applied to the fiber while the ends are still separated to vaporize volatile materials that could cause bubbles. Moving fiber ends together until there is slight pressure between end surfaces completes the final precise alignment. An ideal fusion cycle is short and uses a ramped or gradually increasing arc current. A short, ramped cycle is considered least likely to produce excessive thermal stress in fibers. Cold temperatures require increased time and arc current. Experienced operators consistently produce fusion splices with losses less than 0.2 dB per splice and averaging 0.3 dB on multimode fibers. Sophisticated fusion splicing systems for single-mode fibers produce typical splice losses of 0.05 to 0.1 dB. Mechanical splicing systems position fiber ends closely in retaining and aligning assemblies. Focusing and collimating lenses may be used to control and concentrate light that would otherwise escape. Index matching gels, fluids, and adhesives are used to form a continuous optical path between fibers and reduce reflection losses. In-line connector-to-connector splicing may be used in situations where there is an abundance of optical power. Connectorized cable assemblies are joined through an alignment bushing, which fits snugly over the tip of each connector. Insertion losses for connector-to-connector splices can be as high as 1.0 to 1.5 dB. If these losses are considered excessive, an alternative method should be used. 9.8.6
Connectors
Increasingly, electrical engineers and installers are becoming involved with fiber-optic cable from a specification, design, and installation standpoint. Misconceptions are common about fiber-optic cables themselves, how their handling differs from electrical cable, and how to plan and make high-quality fiber-optic installations. Different types of connectors are used for the following different applications of fiber-optic cables: • ST-compatible connectors: small-size connectors with keyed bayonet
coupling for simple ramp latching or disconnect; dry connection.
Transmission-Network Deployment
505
They are available in multimode and single-mode versions and are fully compatible with existing ST hardware. They are used for dataprocessing telecommunications and LANs, premise installations, instrumentation, and other distribution applications, and have low insertion and return loss. • SMA connectors: small-size connectors with an SMA coupling nut; dry
connection. For use with multimode cables in data-communications applications such as local area and data-processing networks, premise installations, and instrumentation, they have low insertion loss and are fully compatible with all existing SMA hardware.
• Biconic connectors: small-size connectors with a screw thread, cap,
and spring-loaded latching mechanism, having low insertion and return loss. They are compatible with all biconic hardware. Escon connectors are compatible with IBM Escon hardware and are available in single-mode and multimode versions.
• FDDI: a duplex fiber-optic connector system with ceramic ferrule.
Used for data-communications applications, including FDDI backbone, front-end, or back-end networks and IEEE 802.4 token bus, they are dry connection, with positive latching mechanism and low insertion loss.
• FC: a one-piece connector design for easy termination. Compati-
ble with NTT-FC and NTT-D3 hardware, the connectors are dry connection with screw-type strength member retention. Available in multimode and single-mode versions, they are applicable for telecommunications and data-communications networks, premise installations, and instrumentation. They have low insertion and return loss.
• D4 connectors: compatible with NTT-D4 hardware and feature fer-
rule alignment keys for consistent remating and rugged construction for long life and durability. Low insertion and return loss.
• SC connectors: are square in design for high packing density. A
push-pull operation simplifies connections. They are available in single-mode and multimode versions with low insertion and return loss.
506
9.8.7
Transmission Systems Design Handbook for Wireless Networks
Handling Fiber-Optic Cables
From the outside, a fiber-optic cable looks like any electrical multiconductor cable. However, it is lightweight and flexible compared with metal conductor cable. Typical fiber-cable outside diameters (ODs) range from less than oneeighth of an inch up to three-quarters of an inch, depending on fiber numbers and cable construction. The most common fiber-optic cable jacket materials are polyethylene (all types), PVC, and polyurethane. The glass fiber optics in the cable are not fragile. For instance, they wont shatter if you drop the cable on the floor. These glass fibers are usually well protected by buffer tubes inside the cable itself. Once youve found and isolated a fiber, its most notable characteristic is its thinness compared with electrical conductors [14]. Even though the glass in the fiber is actually stronger (having higher tensile strength) per unit area than a metal conductor, there is very little crosssectional area in a fiber available for strength and support. For this reason, most fiber-optic cables have other strength members intended for cable support during pulling, hanging, and so on. The consideration and design of proper cable pathways (i.e., conduit, cable trays, and riser shafts) and termination spaces (i.e., main or intermediate cross connects, horizontal cross connects, and work area telecommunication outlets) is as important as the design of the cable network [15]. Fortunately, a TIA/EIA standard has been approved that addresses this area. It is called the Commercial Building Standard for Telecommunications Pathway and Spaces, TIA/EIA-569. Another useful document is the Telecommunications Distribution Methods Manual from the Building Industry Consulting Service International (BICSI). Placement of fiber-optic cable in conduit is quite common. Conduit offers protection from crushing ground disruption, rodents, and other environmental abuse, plus easier replacement or upgrade in the future. The maximum allowable pulling tension on fiber cable can vary from as low as 50 lbs of force to as much as 800 lbs, depending on the cable construction. The maximum tension for a particular cable should be available from the cable manufacturer and is often found in the cable specs. This maximum recommended pulling tension should be noted on any drawings, installation instructions, and so on. If the field installers do not know the maximum pulling tension, they cannot be expected to respect it. Probably the most common mistake of inexperienced fiber installers is the violation of the minimum bending radius by making tight bends in the cable. Tight bends, kinks, and knots in fiber cable can cause microcrazing or growth of flaws in the fiber, with resulting loss of performance. The
Transmission-Network Deployment
507
minimum bending radius in traditional fiber cable is usually in the range of 20 times cable OD, considerably higher than electrical cable. New fiber technologies, however, are lowering this minimum bend radius. Again, the specific minimum bending radius for a particular cable should be researched in the cable manufacturers specifications. This bending radius must be considered by the engineer when specifying conduit bends and pull box openings or sizing guide pulleys, sheaves, mid-assist capstans, and so on. With respect to tension, friction, and length of pull, fiber-optic cable is often pulled for much longer distances than electrical cable. Continuous fiber pulls of over 4,000 ft are not unusual. These long pulls minimize the number of splices in fiber cable, which is desirable for fiber performance. The light weight of the cable makes these long pulls possible, although proper lubrication and a good conduit installation are also necessities. Special field techniques, such as mid-assist and figure-eighting allow the installation of virtually limitless lengths of fiber cable without splices or breaks. Skilled installation crews have routinely installed uninterrupted lengths of over 20,000 ft in conduit. The theory of pulling tension is the same for fiber-optic cable as for electrical. Pulling tensions can be estimated based on cable weight, conduit system design, and lubricated coefficient of friction. Pulling lubricants with some unique features are required by the special nature of fiber-optic pulling, that is, its long pull lengths and lengthy pull duration. The cables maximum tensile rating must not be exceeded during installation, and the cable manufacturer specifies this value. Tension on the cable should be monitored when a mechanical pulling device is used. Hand pulls do not require monitoring. Circuitous pulls can be accomplished through the use of backfeeding or center-pull techniques. For indoor installations, pull boxes can be used to allow cable access for backfeeding at every third 90° bend. When pulling long lengths of cable through duct or conduit, a fill ratio of less than 50% by cross-sectional area is recommended (Figure 9.15). This is called duct utilization. For one cable, for example, this equates to a 0.71-in OD cable in a 1-in inside-diameter duct. Multiple cables can be pulled at once, as long as the tensile load is applied equally to all cables [16]. 9.8.8
Fiber-Optic Cable Installation Procedures
This section contains information on the placement of fiber-optic cables in various indoor and outdoor environments. In general, fiber-optic cable can be installed with many of the same techniques used with conventional
508
Transmission Systems Design Handbook for Wireless Networks
Spare inner ducts
d2 <50% D2 where: d = Cable diameter where: D = Inner duct diameter
Fiber-optic cable
Inner duct/conduit
d
D
Cable
Figure 9.15 Fiber-optic cable duct utilization.
copper cables. Basic guidelines that can be applied to any type of cable installation are as follows: • Conduct a thorough site survey prior to cable placement; • Develop a cable-pulling plan; • Follow proper procedures; • Do not exceed cable minimum bend radius; • Do not exceed cable maximum recommended load; • Document the installation and testing.
The purpose of a site survey is to recognize circumstances or locations in need of special attention. For example, physical hazards such as high temperatures or operating machinery should be noted and the cable route planned accordingly. If the fiber-optic cable has metallic components, it should be kept clear of power cables. Additionally, building code regulations, such as the NEC, must be considered. If there are questions regarding local building codes or regulations, they should be addressed to the authority having jurisdiction, such as the fire marshal or city building inspector. A cable-pulling plan should communicate the considerations noted during the site survey to the installation team. This includes the logistics of cable-pulling equipment, the location of intermediate access points, splice locations, and the specific responsibilities of each member of the installation
Transmission-Network Deployment
509
team. During cable placement, it is important that the cable not be bent to a smaller radius. After the cable has been installed and the pulling tension removed, the cable may be bent to a radius no smaller than the long-term application bend radius specification. The minimum bend radii values still apply if the cable is bent more than 90°. It is permissible for fiber-optic cable to be wrapped or coiled as long as the minimum bend radius constraints are not violated. While fiber-optic cables are typically stronger than copper cables, it is still important that the cable maximum pulling tension not be exceeded during any phase of cable installation. In general, most cables designed for outdoor use have a strength rating of at least 600 lbs. After cable placement is complete, the residual tension on the cable should be less than this value. For vertical installations, it is recommended that the cable be clamped at frequent intervals to prevent the cable weight from exceeding the maximum recommended long-term load. The clamping intervals should be sufficient to prevent cable movement, as well as to provide weight support. A common practice is to leave extra cable at the beginning and at the end of the cable run. Also, extra cable should be placed at strategic points, such as junction boxes, splice cases, and cable vaults. Extra cable is useful should cable repair or midspan entry be required. Good record keeping is essential. This will help to ensure that the cable plant is installed correctly and that future troubleshooting and upgrading are simplified. All fiber-optic cables have a unique lot number shown on the shipping spool. It is important that this number be recorded. Cable test data, pre- and postinstallation, should be recorded in an orderly and logical fashion.
9.9 Operations and Maintenance 9.9.1
Growth of Multiservice Networks
In recent years, services from wireless and wireline operators have continued to evolve in scope and sophistication. Today, both wireless and wireline are on a fast path of convergence to an IP-based packet backbone network that will redefine communications sophistication by providing multiple voice and data services on a common backbone network. Mobility is a defining aspect of this new public network, and in this case means more than just wireless access. It means that customers will be able to access their services no matter how or where they connect to the networkwhether dialing in from a mobile phone in another country or logging on from their office PC. The
510
Transmission Systems Design Handbook for Wireless Networks
services, applications, infrastructure, and equipment should enable this mobility. In this new public network, consumers and businesses alike will have access to one networkalways on and always connected. It will provide real-time, person-to-person, business-to-business services. It will support Internet services, e-commerce, and news. And it will offer essential mobile portability. As a result, the IP infrastructure is growing exponentially to support IP-based traffic for voice, data, video, virtual private networks (VPNs), and other advanced services. For wireless operators, the objective is to move beyond mere mobile telephony to offer a fuller range of Internet services. Wireless operators need to use a packet network as part of an integrated end-to-end solution, and they need to manage the service transition and protect their core revenue streams. For wireline operators, the challenge is to reduce costs and enable new classes of services. Wireline operators want to transition to a single network that can support voice, data, and Internet services with the same circuit-class reliability that exists for telephony today. And they need to protect existing revenue streams. All of this means a transition from todays circuit-switched networks to the packet-switched networks of the future. When unifying voice and data traffic onto a single backbone, it becomes crucial to evolve not only the network elements, but also the management systems to cope with the needs of a combined circuit- and packetswitched network. The management system needs to maintain its robust and scalable characteristics to cope with the increases in user penetration and higher bandwidth requirements. The management system needs to provide complete management not only of elements, but of services, from an integrated management suite, thereby increasing the efficiency and effectiveness of the operational staff. Managing customer-service levels is a necessity for all carriers and service providers who want to introduce differentiated services into their networks. To achieve this in an efficient way, the wireless-network operator needs a management framework for developing seamless services across multiple platforms and networks. This means end-to-end provisioning capabilities, end-to-end monitoring applications, billing, strong security measures in protocol communications and authentication, service level agreement (SLA) implementations, and more. In a multiservice network, service and network assurance can pose numerous significant challenges, and while todays Internet is largely offered on a best-effort basis, it is evolving to offer different QoS levels as well. Once these premium services are deployed, it is crucial to have the applications needed to monitor and manage customer-service levels and ensure that customers are receiving the level of service they have paid for.
Transmission-Network Deployment
511
Network events need to be correlated against services and customers to see which customers are affected. Also, it may be necessary to monitor various Internet servers to ensure that response times are acceptable (in addition to the network equipment)that is, the Web servers, HTTP servers, and so onand that the service is performing against its SLA. To address these challenges, the ITU has published a basic architecture and accompanying generic standards, which comprise the telecommunications management network (TMN) model. The TMN defines management functions and communications for operation, administration, and maintenance of networks and services in multivendor environments. TMN functions are organized as building blocks and form the following categories: business management, service management, network management, and element management. Business management encompasses such operational functions as problem tracking and inventory management; service management covers functions for serving customers (internal and external), such as subscriber administration and billing. Network management encompasses administration of the network infrastructure, such as line provisioning and PVC management; element management defines the interfaces and protocols for monitoring and configuring hardware devices. 9.9.2
Network Management System
It is important for any network operator to be able to manage the network after it is deployed. This is reflected in the emergence of both proprietary and open network management protocols. These protocols provide a means to collect information and to perform network-management functions relating to the following: • Configuration management; • Performance management; • Fault management; • Security management; • Accounting management.
SNMP and Common Management Information Protocol (CMIP) are examples of the open-architecture management protocols [17]. SNMP is designed to work with TCP/IP and establishes standards for collecting information and for performing security, performance, fault, accounting, and
512
Transmission Systems Design Handbook for Wireless Networks
configuration functions associated with network management. CMIP is more comprehensive in scope and is designed to work with all systems conforming to International Standards Organization (OSI) standards, but requires considerably more overhead to implement than SNMP. It is important for network operators to integrate IP and ATM systems with comprehensive management intelligence that will provide high reliability, scalability, and maintainability for both wireless and wireline deployments. Management applications keep operational costs down by automating configuration tasks and protecting the networks integrity. They support business goals by gathering charging data and mediating it into billing systems, as well as by supporting network design and optimization. And they enable the provisioning and billing of new services in addition to the support of existing services. Without a powerful management environment, a new public network is unable to generate profits to justify the operators investment. The new wireless networks will include many more network elements than traditional networks because a mix of switches, Internet routers, and servers will populate the network. Operators will need to reduce the complexity of management to allow time for integration of new network elements and introduction of new services. Vendors must allow easy customization within a management framework by providing modular solutions. One example is in the network surveillance applications, which vendors may need to make exchangeable for what the customer has already installed. Device-oriented configuration management requires specialized knowledge and applications. The coordination between applications is best achieved higher up in the management chainfor example, at the network management, service provisioning, or customer care layers. The network-management solution also provides the necessary tools for an optimized and cost-efficient operation and maintenance of the transmission network. The network-management solutions constitute different types of management components: local craft terminals, element managers, and transmission-network managers providing integrated transmission management. Planning of the NMS results in a diagram that outlines how the contracted management systems and their respective network elements are to be interconnected. The following outlines a step-by-step approach: 1. Assign network element IDs for each network element. 2. For each subsystem, design a logical network that outlines how the O&M signaling should be routed from the element manager (EM) to the corresponding NES. Special consideration should be taken if
Transmission-Network Deployment
513
the subsystem network is not homogeneous. The strategy for connecting such inhomogeneous subnetworks should be outlined. 3. Design a logical network that outlines how the different EMs are connected to the umbrella NMS, if any. 4. Map all the logical connections onto the physical network. Design a diagram that outlines how the O&M signaling in the NMS network is routed through the physical network. 9.9.2.1 Configuration Management
Configuration management involves collecting information on the current network configuration and managing changes in the network configuration. To configure and manage the network, operators need cohesive applications that are simple, yet effective and secure. As networks grow in size, configuration management becomes more difficult to maintain on a per-NE basis, creating the need for subnetwork managers. Routing protocols are relatively complex to configure, and the ability to modify and verify configuration changes off-line before committing them to the network is important. With the trend toward embedded EMs running on Web servers within the NE themselves, applications are needed to carry out bulk configuration activities over a number of NE simultaneously. IP networks offer new means of managing traffic. Two important new features are QoS and explicit path placement. To manage and provide guaranteed grades and levels of service, carriers must efficiently use bandwidth and optimize network performance through proper engineering and dimensioning. To date, many carriers overprovision bandwidth to ensure service availability and quality. This is very cost-ineffective, and as competition intensifies, carriers will need to leverage their existing bandwidth better. Before new service demands can be supported, the carrier must check to ensure that sufficient network resources exist and that an end-to-end path through the network exists as well. If there are performance problems in the network, carriers should be able to reengineer the network to ensure high-grade service availability. Finally, operators want off-line what-if analysis to determine if the network is optimally dimensioned and sufficiently robust. 9.9.2.2 Fault Management
Fault management is essential to identifying and correcting any faults or defects quickly (as they occur) and accurately to limit their impact on the
514
Transmission Systems Design Handbook for Wireless Networks
availability of the provided services. The management applications should enable easy location of the root of a problem and provide the functionality to dispatch the field force, rectify the problem, and track the progress of open issues. To achieve this, alarms need to be categorized by severity and filtered or correlated to reduce the number of incidents that the operator must address. In addition to managing faults, performance-management applications are required to identify potential bottlenecks proactively before they affect services. It is important to record and analyze usage and performance trends to make informed decisions about network planning and justify investments in new capacity. 9.9.2.3 Performance Management
Performance management provides the ability to select, collect, store, and evaluate a wide range of data on the status, operation, and condition of every major component in the wireless network. Performance management is a continuous process of data collection concerning the GoS, traffic flow, call processing, and utilization of system resources and NEs. By monitoring performance, it is possible to measure overall QoS and detect any change or deterioration of service at an early stage, before major difficulties develop. Performance management involves the long-term assessment of various system operations. The major objectives are to provide the following: • An operational baseline to determine that the wireless system and its
components are functioning at normal levels;
• Information about operational variances and trends in the wireless
system;
• Information about operational changes in the wireless system, con-
centrating on anomalies, deficiencies, or deterioration of service;
• Information about utilization of system services for expanding or
upgrading system services;
• Information for job administration; • Information about variances above or below design limits, or failure
to meet design criteria.
This last objective of performance management comes closest to fault management. It is intended, however, for use in maintaining or restoring the level of service, not for the immediate correction of faults, if they occur in the
Transmission-Network Deployment
515
system. An important part of performance management is the reporting of collected data and evaluations. Performance reporting involves report format definition and scheduling. Reports usually are compiled and produced based on a schedule that fits the granularity periods of the collected data. Reports produced by a particular measurement job must have the same layout and contain the information specified by the requestor. 9.9.2.4 Security Management
Security management is also a vital part of network management. It is responsible for ensuring secure communications and protecting the network operations. Network security requirements are not explicitly considered during the execution of topological network design. Nonetheless, security considerations may have considerable impact on the choice of network devices and services. For example, an often-cited reason for private networks as opposed to public networks is the need for control and security. One level of security is offered by protocol security; thus, it is important to assess the level of vulnerability posed by the presence or lack of good protocol security in the network. For example, SNMP and other network-management protocols that have been designed with security in mind can be used to identify and protect the network against unauthorized use. IP networks, however, are potentially vulnerable to source-address spoofing. Spoofing is a form of masquerading where packets appear to come from a source that they did not. Future versions of IP will have new security provisions, and for IPv6 recommendations were made to provide for IP authentication headers and IP encapsulating security payload. Operational security provides a second level of network security. It involves disabling network services that are not necessary or appropriate for various types of users. Network security can be implemented at the physical level, which safeguards access to the network by securing network components and limiting access to authorized personnel only. Network security can also be implemented at the data level. This involves the use of encryption technology to protect the confidentiality of the data transmission. 9.9.2.5 Accounting Management
Accounting management involves gathering data on resource utilization. It may also involve setting usage quotas and generating billing and usage reports. Finance and billing is the focal point for receiving status reports regarding service-level violations, network plans, designs, and changes and invoices from third parties.
516
9.9.3
Transmission Systems Design Handbook for Wireless Networks
Geographic Partitioning
For a large network, the NMS is organized in such a way that geographical partitioning allows for regional management, while still providing a centralized view of the entire network. Geographically partitioned networks consist of regional subnetworks and a backbone network connecting the subnetworks together. Network division into regions is performed on a geographic basis: each NE (node) belongs to a single region. In geographic partitioning, a hierarchy level is assigned to an NE (node, trunk). The hierarchy levels divide the network into the access and backbone parts. This way, a regional operator can focus on managing one part of the network, while paying less attention to other parts. Each region can be viewed as a separate network that is managed efficiently, no matter how large and complex the overall network structure is. A good example is national wireless networks with regional NOCs (different markets) and one central NOC that maintains visibility of all the other regional NOCs. Partitioning improves the structural organization of the network and, hence, the management procedures. This facilitates network building by creating simple subnetworks, such as partitions, and helps the operators to understand the overall network structure. The nature of the transported traffic within a single region is divided into three categories: 1. Region internal traffic; 2. Traffic terminating at a region; 3. Transit traffic. The NEs (nodes and trunks) are divided into hierarchy levels that indicate the nature of traffic that is carried in the NE. The hierarchy levels of the network elements divide the network into access and backbone parts. 9.9.4
Location-Finding Techniques
There are a number of technology alternatives for locating mobile phones, including network-centric and handset-centric solutions. The former build significant intelligence into the handset to achieve location, while the latter build more intelligence into the mobile network infrastructure. Handset-centric technology solutions include GPS, overlay triangulation technologies based on timing or angle of signal transmission and reception at the handset [enhanced observed time difference (E-OTD), time of arrival (TOA)], and cell of origin (COO) information. The only technology that is widely deployed in wireless
Transmission-Network Deployment
517
networks today is COO information. This scheme is used to meet phase-one 911 emergency-services requirements in the United States, wireless-office location-specific billing applications, and some location-specific informationrequest services. As more network-based location-finding schemes are deployed and GPS capability is integrated into wireless devices, the improved accuracy of location fixing will not only improve current services, but allow the introduction of new services. GPS is the most commonly discussed option. For locating unmodified cellular telephones, the alternatives available are COO, enhanced offset time division, time of arrival, angle of arrival, and signal attenuation. The T1P1 Committee, a subcommittee of ANSI, and ETSI, has recently committed to the standardization of location-finding systems using E-OTD, TOA, and assisted GPS (A-GPS), in addition to COO. In terms of implementation, the following hold true: • A-GPS requires additional equipment or modification to the mobile
station (MS);
• E-OTD requires both network and MS modification; • TOA requires mainly network modification (modern handsets
should work);
• COO requires no modification to the handset.
Because COO requires no modification to the handset or networks, it can be used as the location-finding system for existing subscribers, but is less accurate than the other methods employed. Some would argue, however, that the accuracy of COO in cities is more than adequate for information services, owing to the small cell size. The accuracy of COO is questionable though when the location-finding system is required for assisting with emergency services. In this system, the mobile-network base-station cell area is used as the location of the caller. Positioning accuracy generally depends upon the size of the cell, but a position down to 150m is possible in urban areas with the deployment of picocell sites. Although other schemes offer higher degrees of positioning accuracy than COO, its main advantage is that the speed of response in getting a location fix is fast (typically around 3 seconds). Since the handset or a network upgrade is not required, it can be used to provide location-specific services to existing customer bases.
518
Transmission Systems Design Handbook for Wireless Networks
E-OTD systems operate by placing location receivers or reference beacons, overlaid on the cellular network as a location-measurement unit, at multiple sites geographically dispersed over a wide area. Each of these beacons has an accurate timing source, and a signal from at least three base stations must be received by an E-OTDsoftware-enabled mobile and the location-measurement unit. The time differences of arrival of the signal from each RBS at the handset and the location-measurement unit are then calculated. The differences in time stamps are then combined to produce intersecting hyperbolic lines from which the location is estimated. E-OTD systems may, however, be subject to many of the same urban multipath problems as angle-of-arrival (AOA) systems. In this case, multipath distorts the shape of the signal and the group delay, causing the E-OTD system some difficulty in accurately determining the point in the signal to be measured by all receivers. E-OTD schemes offer greater positioning accuracy than COO between 50m and 125mbut have a slower speed of responsetypically around 5 secondsand require software-modified handsets, which means that they cannot be used to provide location-specific services to existing customer bases. In a similar manner to E-OTD, the difference in TOA of a signal from a mobile device to three RBSs is used to calculate the location. In this scheme, however, there may be no overlay network used as the location-measurement unit. Instead, this functionality is provided by synchronization of the cellular network, using GPS or atomic clocks at each RBS. This capability is found in cdmaOne networks in the United States, which are synchronized, whereas it is not provided as a given in asynchronous GSM networks. Further problems associated with TOA schemes are that the accuracy offered might be little better than that of COO in urban areas, but with a far longer response time than COO or E-OTDtypically around 10 seconds. The cost of synchronizing a GSM network is therefore relatively high for little increase in performance offered over the use of COO, and the investment required for an operator may be prohibitive. However, no handset modification is required and services can therefore be offered to existing customers. AOA operates with no modification of a mobile device. It was first developed for military and government organizations, and was later applied to analog cellular systems. The brevity of the signals and the channel sharing that exists means that AOA schemes are difficult to deploy successfully in digital systems. The most common version of this technique is known as small-aperture direction finding, which requires a complex 412-antenna array at each of several cell-site locations. These antenna arrays can, in
Transmission-Network Deployment
519
principle, work together to determine the angle (relative to the cell site) from which a cellular signal originated. When several cell sites can each determine their respective AOAs, the cellular-telephone location can be estimated from the point of intersection of projected lines drawn out from the cell site at the angle from which the signal originated. Small-aperture AOA systems suffer from distortion of the wavefront of the cellular signal, caused by multipath and other environmental factors. The technology can perform acceptably when tracking a continuous transmission, such as a voice transmission, if longer integration times are used. As the caller moves from cell to cell, the system must follow each voice channel assignment as the call is handed off from channel to channel. This can be difficult if the AOA antenna is not positioned to interpret the in-band voice channel signaling. A significant drawback to AOA systems is the logistical and aesthetic dilemma of adding antenna arrays to cell sites at a time when communities are enacting increasingly restrictive planning regulations. Existing cell-site antennae are not suitable for AOA. Another drawback is that the angular error of the antenna array can translate into a significant error in lateral distance if the cellular telephone is far from the cell sites. The signal-attenuation technique for estimating cellular-telephone locations using existing signals operates on the principle of signal attenuation as the mobile phone moves toward or away from a base station. Most mobile phone antennae are omnidirectional, so power is dissipated rapidly in all directions. If the transmitted power of the mobile was known, and the power was measured at another point, the distance could be estimated using one of several propagation models. However, this technique is generally considered the least reliable method for estimating location, for several reasons. Discovering transmitted power is a significant burden that is complicated by cell-site sectoring, antennae down-tilting, and continuous wireless system tuning. Signals attenuate for reasons other than distance traveled, such as passing through walls, foliage, or glass and metal vehicles. Signal attenuation also experiences seasonal variations due to weather, changes in foliage, and other environmental factors. Powermeasuring circuits generally cannot discriminate among power received from multiple directions, such as a direct path, reflections off of buildings, or even reflections off of trucks. This phenomenon is witnessed by any cellular user who has seen the signal strength bars on a telephone display fluctuate, even when the mobile is not moving.
520
Transmission Systems Design Handbook for Wireless Networks
References [1] Clark, M. P., Networks and Telecommunications: Design and Operation, New York: John Wiley and Sons, 1997. [2] Situation Management Systems, Inc., Managing Negotiation: Selected Readings on Negotiation Skills, 1996. [3] Mann-Robinson, T. C., Network Design: Management and Technical Perspectives, Boca Raton, FL: CRC Press LLC, 1999. [4] Wysocki, R. K., et al., Effective Project Management, New York: John Wiley and Sons, 1995. [5] Norris, M., Understanding Networking Technology, Norwood, MA: Artech House, 1996. [6] Timiri, S., RF Interference Analysis for Collocated Systems, Microwave Journal, January 1997. [7] Harris-Farinon Division, Systems Engineering Applications Guide, Issue 1, 1995. [8] Harris-Farinon Division, Transmission Engineering: An Explanation, Issue 2, 1993. [9] Safety Code 6, Health Canada, Limits of Human Exposure to Radiofrequency Electromagnetic Fields in the Frequency Range from 3 kHz to 300 GHz (also available from http://www.hc-sc-gc.ca). [10] Federal Communications Commission, Office of Engineering and Technology, Evaluating Compliance with FCC Guidelines for Human Exposure to Radiofrequency Electromagnetic Fields, OET Bulletin, Vol. 65, August 1997. [11] IEEE, IEEE Recommended Practice for the Measurement of Potentially Hazardous Electromagnetic FieldsRF and Microwave, IEEE STD C95.3-1991, 1991. [12] Industry Canada, Spectrum Management and Telecommunications Policy, Guidelines for the Measurement of Radiofrequency Fields at Frequencies from 3 kHz to 300 GHz, August 2000. [13] ITU-T G.652, Characteristics of the Single-Mode Optical Fiber Cable, October 2000. [14] Sterling, D. J., Jr., Technicians Guide to Fiber-Optics, Second Edition, Albany, NY: Delmar Publishers 1993. [15] AT&T, Outside Plant Engineering, 1990. [16] Corning, Fiber-Optic Installation Guidelines. [17] Mann-Rubinson, C. T., Network Design: Management and Technical Perspectives, Boca Raton, FL: CRC Press, 1998.
10 Transmission-Network Testing and Commissioning 10.1 Definitions The total test time in transmission systems is divided into two categories: available and unavailable time. The system becomes unavailable if the BER is equal to or worse than 10−3 for more than 10 consecutive seconds. The following error performance parameters are used in describing transmission quality: unavailable seconds (USs), percent availability, degraded minutes (DMs), percent degraded minutes, severely errored seconds (SESs). A list of definitions follows. • ESs. Seconds with one or more errors. −3
• USs. The bit error rate in each of these seconds is worse than 1 × 10
for a period of 10 consecutive seconds. • DMs. The degraded minutes result is the number of minutes during which the bit error rate is worse than 1 × 10−6, but better than 1 × 10−3 (excluding SES). • SESs. The severely errored-seconds result is the number of seconds during which a bit error rate is worse than 1 × 10−3 and occurs within the available time. • ESF. Preferred today (T1) since it provides network performance monitoring without disrupting the service. 521
522
Transmission Systems Design Handbook for Wireless Networks
• Bipolar coding with eight-zero substitution (B8ZS). Uses intentional
BPVs to break up long strings of 0s, allowing their transmission through the T1 link without violating the ones density standard.
• Loopback testing. If CSU loopback is used, far-end CSU in loopback
will affect the results, since CSUs (like many other types of transmission equipment) remove received BPVs before retransmitting the data.
• End-to-end testing. Performed with two test sets so the analysis can be
done simultaneously in both directions, and the direction of errors can be found much faster. It is recommended loopback testing be avoided, and end-to-end testing be performed whenever possible.
• Long-term testing. Long-term testing should be performed over an
8-hour and up to a 24-hour period for leased facilities (longer for the microwave links), using standard pseudorandom patterns or stressful patterns.
• Live-data emulation. A pseudorandom pattern that will be used is
quasirandom signal sequence (QRSS). This pattern provides a good approximation of live traffic with an approximately 50% 1s density. This pattern generates all possible combinations of a 20-bit binary counter except all 0s, and is limited to a maximum of 14 0s.
• Additional tests. QRSS tests are very useful to prove connectivity, but
the system also has to be tested with the following set of tests in order to make sure that there are no hidden issues (e.g., AMI-B8ZS mismatch or faulty equipment) that could cause intermittent problems later on during the operation: • • • •
All onesa fixed test pattern of pulses only; One for one (1:1)alternating 1s and 0s; All zerosused to test circuits for clear channel capability. Stress testingmost commonly used fixed pattern is 3 in 24. The 3 in 24 pattern simultaneously stresses the minimum 1s density and the minimum number of consecutive 0s criteria.
10.2 BERT 10.2.1 T1 Impairments
Any transmission system is vulnerable to impairments in signal quality over distance. Noise can be introduced by electrical transients such as lightning or
Transmission-Network Testing and Commissioning
523
equipment switching on or off, cross talk with other circuits, or simply the thermal noise associated with any electrical circuit operating at a temperature above absolute zero. On an analog system, almost any noise that enters the signal chain is there to stay. Unfortunately, there is no effective method to allow amplifying or receiving equipment to dissect an analog waveform and determine with certainty which portion of it is the original signal and which is noise that has been picked up along the way. Once an analog signal has been converted to digital format, the situation is quite different. A binary digital signal consists of a series of pulses, called bits, which can represent only two states: 1s and 0s. As long as the decoder can tell the difference between a 1 and a 0, the original signal can be reproduced perfectly. As a result, a significant performance difference exists between analog and digital transmission systems under noisy conditions. On an analog system, any increase in noise on the line correlates directly to a decrease in audio quality (poorer signal-to-noise ratio). On a digital system, increased line noise has little effect on decoded audio quality until the noise level gets so high that it prevents the decoder from differentiating between 1s and 0s. A digital transmission facility such as a T1 line carries multiple channels of information in structures called frames. Additional bits (overhead bits) keep track of the frame boundaries and are used to determine which bits go with which channels. When first presented with a signal, T1 transmission equipment searches for the overhead bits. Once it has found them, it is said to be in frame, and begins processing the channel data. The transmission equipment continues to monitor the overhead bits to ensure that the frame alignment remains correct. When there are too many bit errors on the T1 line, the equipment declares itself to be out of frame, causing the circuit to go down. This condition is called loss-of-frame synchronization. National and international standards authorities (ANSI and ITU-T) define certain methods used by most T1 equipment manufacturers, for keeping digital signals in frame. Equipment adhering to these standards can lose frame synchronization within seconds when one out of every one thousand bits is in error. Bit errors sometimes appear in bursts, or brief periods of extremely high error rates, and the goal then is getting back in frame as quickly as possible. 10.2.2 T1/E1 Testing
After installation of a leased T1/E1 or T3/E3 circuit, out-of-service BERT is the most useful tool in verifying equipment operation and end-to-end
524
Transmission Systems Design Handbook for Wireless Networks
transmission quality. Out-of-service testing allows the user to accurately assess the quality of a T1/E1 and/or T3/E3 circuit by transmitting and analyzing test patterns in place of the live data that is normally present. Two methods of out-of-service testing are typically used to analyze T1/E1 networks: end-to-end (two test sets required) and loopback testing (one test set required). Both of these methods require test patterns to stress and analyze a network. Only end-to-end testing should be performed to analyze T3/E3 networks. End-to-end testing is performed with two test sets so that the analysis can be done simultaneously in both directions, and the direction of errors can be found much faster. Loopback testing is recommended over end-to-end testing whenever possible. For many T1 transmission systems, the BERT uses a QRSS pattern, but it is a good idea to use some of the stress test patterns as well. The exact error-free performance requirements vary from one location to another, but −8 if there are no other guidelines to use, start with a BER of 10 or better if the live traffic contains only voice channels. If data channels are present, try to start with a BER of 10−10 or better. Over some facilities, still better performance can be measured, but this takes a tremendous amount of time. For normal telecommunications transmission facilities, acceptance tests are commonly run for periods of a few hours at most. It is most important to record all information regarding the acceptance tests, actual results and a sketch of how the tests were run. Acceptance tests should be kept on file in every central office or equipment room. Whether public or private, T1 circuits and network equipment must be properly tested and maintained to perform to maximum efficiency. Accordingly, all T1 testing falls under one of two prescribed categories: out-ofservice testing and in-service monitoring. 10.2.3 Out-of-Service Testing
Out-of-service testing (Figure 10.1) is so named because live traffic must be removed from the T1 link before testing can begin. In its place, a test instrument transmits a specific data pattern to a receiving test instrument that can recognize the sequence of the pattern being sent. Any deviations from the transmitted pattern are then counted as errors by the receiving instrument. Out-of-service testing can be conducted on a point-to-point basis or by creating a loopback. Point-to-point testing is a general practice, and requires two test instruments (one at either end of the T1 link). By simultaneously generating a test data pattern and analyzing the received data for errors, the test instruments can analyze the performance of the link in both directions.
Transmission-Network Testing and Commissioning
525
Near end
Far end DSX-1
MUX or Channel bank
MUX or Channel bank
DSX-1
Step 1.
loopback channel bank
BERT
Step 2. BERT Step 3.
Bit-error-rate test one way from near to far end Bit-error-rate test one way from far to near end
BERT Bit-error-rate test both ways from near to far and back to near end
BERT
loopback channel bank
BERT Write down results!
Figure 10.1 T1 acceptance testing.
Loopback testing is often used as a quick check of circuit performance or when isolating faulty equipment. In loopback testing, a single test instrument sends a loop-up code to the far-end CSU before data is actually transmitted. The loop-up code causes all transmitted data to be looped back toward the test instrument. By analyzing the received data for errors, the test instrument measures the performance of the link up to and including the far-end CSU. Because loopback testing requires only a single test instrument (and thus only one operator), it is very convenient. However, loopback testing is limited in that it can only analyze the combined performance of both directions of the link. As such, it is extremely difficult to determine whether errors are originating on the transmit or the receive side of the T1 link at any given time. As out-of-service methods, both point-to-point and loopback tests allow detailed measurements of any T1 link. However, since all out-ofservice testing requires that live, revenue-generating traffic be interrupted, it is impractical for long-term testing. Thus, this type of testing is typically
526
Transmission Systems Design Handbook for Wireless Networks
performed when a circuit is initially installed or when errors are discovered when monitoring data. 10.2.4 In-Service Monitoring of Live Data
The in-service method allows live data to be monitored at various access points without disturbing revenue-generating traffic. Since in-service monitoring does not disrupt the transmission of live traffic, it is more suitable for routine maintenance than out-of-service testing. Additionally, in-service monitoring indicates performance under actual operating conditions. But its primary disadvantage is that its measurements may not be as precise as those available in out-of-service testing. Whats more, some network equipment may deter traditional in-service error measuring.
10.3 Transmission-Network Testing Procedure 10.3.1 Testing Leased Facilities
A record of the standard turn-up tests, called the system verification document, defined by the transmission facilities provider (called a carrier or telecommunications company), agreed upon by the customer, and performed after the equipment installation is complete, will be provided by the carriers engineer and witnessed by the customers representative. This document contains the following information: • Site-specific information and contact names; • Circuit number or other identification data; • Emergency hotline to assist the customer in case of service-affecting
faults (examples of emergency technical support include serviceaffecting problems reported by either side or hardware failures that cause outage or degradation);
• Visual inspection of the transmission equipment, cables, labeling,
and miscellaneous equipment;
• Protection switching (if applicable, for E3/T3); • Diversity routing (if applicable, for E3/T3); • Matrix, module, and interrack cable continuity (if applicable); • Coax, abam, and fiber-optic cables and cross-connect continuity; • User interface functionality;
Transmission-Network Testing and Commissioning
527
• Long-term (24 hours) BERT. [Tests should be performed end-to-
end between demarcation points (NIU) with the E1/T1 or E3/T3 provider and customers terminal equipment (cell-site equipment). Long-term testing with printouts should be performed over an 8-hour and up to a 24-hour period for leased facilities, using standard pseudorandom (live data emulationprovides a good approximation of live traffic) patterns and stressful patterns. Test patterns are defined in North American or international standards and facilities provider and customer s internal specifications. Each CSU (or NIU) is looped back (for T1 circuits) to make sure that they respond to both loop-up and loop-down codes and to verify that the circuit is operating properly end-to-end.]; • Signatures. The customer can own transmission (backhaul, transmission) facilities used in the wireless network, and they can be either over copper, fiber, or microwave. 10.3.2 Testing Microwave Systems
A record of the standard turn-up tests, defined by the wireless operator, agreed upon by the microwave supplier, and performed after the MW radio system installation is complete, will be provided by the supplier s technician and witnessed by the wireless operators representative (or its designated representative, subcontractor, or consultant). This system verification document will contain the following information: • Site-specific information and contact names; • MW link engineering details; • Visual mechanical and physical inspection of the MW radio equip-
ment, cables and waveguides, antenna mounting structures, antennas, labeling, and miscellaneous equipment;
• Electrical measurements on the MW radio, which should typically
include the following information (test results): • • • •
Grounding measurements; Antenna/waveguide/coax return loss measurements; Input dc voltage; RF power verification;
528
Transmission Systems Design Handbook for Wireless Networks
Frequency accuracy measurements; Receiver tests, including Automatic Gain Control (AGC) characteristics; • Data service channel and VF order-wire testing; • System alarm and control; • Loopback capabilities. Check pressurization (waveguides only); Short-term BER measurements (1 minute each port); Protection switching operation test; Power supply redundancy verification; Modules and interrack cable continuity; Coax, abam, and fiber-optic cables and cross-connect continuity; User-interface functionality; Spare module testing; Network management system and craft interface verification; Long-term (24 hours) BERT (per hop) as defined in North American or international standards and the telecommunications company and customers internal specifications; Long-term (24 hours) BERT (for the system, if applicable) as defined in North American or international standards and the telecommunications company and customers internal specifications; Emergency technical support hotline, which will assist in case of service-affecting faults (examples of emergency technical support include service-affecting problems reported by either side or hardware failures that cause outage or degradation); User manuals and as-built documentation; Battery backup time for the MW system; Warranty, repair, and return procedure description; Signatures. • •
• • • • • • • • • •
•
•
• • • •
10.3.3 DS1 and DS3 Performance Objectives
The DS1 performance objective [1] for short-haul (less than 400 km) North American dedicated digital service over a 24-hour time period is as follows: BER = 2.6 × 10 −10
(35 ES − 99.96 PEFS/day )
Transmission-Network Testing and Commissioning
529
The DS3 performance objective for short-haul (less than 400 km) North American dedicated digital service over a 24-hour time period is as follows: BER = 18 . × 10 −11
(70 ES − 99.92 PEFS/day )
This BER assumes random errors (one error per ES). If errors are bursty (more than one error per ES), a higher BER can correspond to the same percent error-free seconds (PEFS). 10.3.4 DS1 Test Procedure and ATP Form 10.3.4.1 DS1 Test Procedure
This test can be performed end-to-end (the preferred way of testing) or as a loopback test. Loopback can be achieved remotely, using loop-up code, or simply by physically looping the circuit at the far end. The test is performed as follows: • The test set is in B8ZS line coding, and uses ESF mode of operation. • Run all ones framed to check the circuit continuity. • Run 1:1 framed to check the circuit continuity. Send 10 errors to
prove continuity.
• Run 3 in 24 framed test set in B8ZS mode. Any problems with
FRAME or PATTERN SYNCH indicates that a B8ZS-AMI mismatch exists somewhere in the circuit.
• Run all 0s framed to check for the AMI-AMI mismatch. A problem
will appear as FRAME and BIT ERRORS.
• Run QRSS for 2 hours (simulates live traffic). Check the results and
report any problems. • In case the 2-hour report contains excessive errors, conduct a 24hour QRSS test. 10.3.4.2 Transmission ATP Form
Results of testing (Table 10.1) the transmission facilities must be recorded, for a number of reasons. The first and the most obvious reason is that recording is a part of the overall transmission-network acceptance procedure. The other reason is that such a record will be the benchmark for all future troubleshooting of the transmission network. During the maintenance tests of
530
Transmission Systems Design Handbook for Wireless Networks Table 10.1 Transmission Acceptance Test Results Form
Work order number: ___________________________
Page: ________________
Circuit number: ______________________________________________________________________________ Carrier: ____________________________________
Carrier circuit number: ____________________________
Test site: __________________________________
Address: ______________________________________
Far end site: _________________________________
Address: ______________________________________
Type of Test:
❑Loopback
Test Pattern Time Required Problems
❑End to End All 1s framed
1:1 framed + 10 errors
3 in 24 framed
QRSS
All 0s framed
QRSS
1 min
1 min
1 min
1 min
1 min
2 hrs
❑ No
❑ No
❑ No
❑ No
❑ No
❑ No
❑ Yes
❑ Yes
❑ Yes
❑ Yes
❑ Yes
❑ Yes
In case of excessive errors during the 2-hour test, perform the 24-hour test. Notes:
Test set should be set for ESF, B8ZS. Printout of the long-term testing is required. Group/Dept.:
_________________________________
Tested by:
_________________________________
Date:
_________________________________
Approved by:
_________________________________
either leased lines or owned microwave systems, the results will always be compared against the acceptance testing results. Any change in these results should be either noted or further investigated.
10.4 Fiber-Optic Cable Testing There are a number of different tests and measurements that could be done in the laboratory to verify the quality of fiber-optic cables and terminal equipment. However, only a small number of these tests are performed in the field; the most important ones are measurements of optical power and of optical losses. When measuring optical power, the optical power sensor converts light propagating through optical fibers (or free space) from a laser diode or light-emitting diode (LED) source.
Transmission-Network Testing and Commissioning
531
Optical losses can be measured using two different techniques: the two-point technique or backscattering technique. In the two-point technique, the optical losses are measured by comparing the input optical power and the output power from the fiber being tested. Extremely accurate measurements can be made using this technique, but it requires two sets of test equipment and two operators, one at each end of the cable. In the backscattering technique, an optical time domain reflectometer (OTDR) is used to measure the losses by collecting the light that is backscattered from the fiber, allowing measurements to be made from only one end. OTDRs are also used for locating faults that occur in optical-fiber cables. Any anomalies in the fiber-optic cable produces changes in the weak, reflected light (backscattering), also called Rayleigh scattering. The level of backscattering depends upon the propagation time and drops off exponentially. 10.4.1 OTDR Test Procedure for Single-Mode Fiber-Optic Cables 10.4.1.1 Fiber-Optic Link Loss Test
Fiber-link loss testing is performed after final splicing and termination of all the fibers in the cable. An OTDR is used to test each fiber; this test is made through the optical connectors in the fiber distribution panel at one end of the link and thus provides a final measurement of the total losses for each fiber in the cable. A soft-copy record of each measurement is kept for final acceptance documentation. OTDR should be capable of the following: • Single-mode optical testing at both 1,310-nm and 1,550-nm
wavelengths;
• Noise-free measurements of the near end of the fiber with no dead
zone restrictions;
• Floppy disk storage of all types of waveform traces, including the
MS/DOS- or MS Windows-based software required to display and print the traces internally or by using a PC;
• Minimum 30-dB dynamic measurement range; • User-written notes and other details that show up on the display and
on the printout for trace and fiber identification.
Values of the losses in fiber-optic systems are typical values that could vary from project to project depending on the equipment used.
532
Transmission Systems Design Handbook for Wireless Networks
• Fiber losssingle-mode fiber loss per kilometer will not exceed
0.4 dB.
• Fiber splice losssplice loss of any individual splice will not exceed
0.1 dB.
• Fiber connector lossconnector loss through any individual connec-
tor will not exceed 0.5 dB.
• Fiber link losslink loss will be within ±0.1 dB of the design link
loss per km of link distance.
10.4.1.2 Test Procedure
The test procedure used for testing fiber-optic cable is as follows: 1. Connect the OTDR to the fiber under test through a fiber-optic jumper cable to the fiber distribution panel connector. Use this same jumper cable for testing all fibers in the cable under test. 2. Set the OTDR for an output wavelength (1,310 nm or 1,550 nm) equal to the wavelength to be used by the link fiber-optic equipment. 3. Adjust the OTDR for a proper trace, and record the trace on the OTDR floppy disk. Make a printout of each trace for each fiber. Be sure that the trace shows the loss for the entire length of the fiber; that is, past the open end of the fiber. 4. Repeat the first three steps of this test procedure for each fiber in the cable under test and, if required, again in the opposite direction for each fiber in the cable under test. 5. Using the test results for opposite directions, calculate the average loss value for each fiber. 10.4.2 Characterizing and Testing Fibers for DWDM Applications 10.4.2.1 DWDM Specifics
Theoretically predicted and confirmed by field experience is that the characteristics of the fiber itself can have significant impact on the performance of DWDM networks and that the particular characteristics, which are most important, are not necessarily those of greatest concern in conventional single-wavelength links. Chromatic dispersion, the variation of the index of refraction of the fiber with wavelength, can be a critical determinant of system performance in
Transmission-Network Testing and Commissioning
533
DWDM systems, especially those that use a judiciously selected amount of dispersion to minimize certain undesirable nonlinear effects in the fiber itself. Its value is determined during fiber manufacturing, however, and few situations have arisen in which it is necessary to verify this value in the field. As DWDM systems are operated ever closer to their limits, however, a need is likely to emerge to verify that chromatic dispersion is adequately controlled at every point in the optical path. The eventual development of field instrumentation to measure chromatic dispersion is likely, especially if the management of chromatic dispersion on installed fiber turns out to be more complex than expected. Polarization mode dispersion (PMD), in which various polarization states of the optical signal propagate at different velocities, is especially difficult to deal with. Its effects prevent many present-day optical systems from using high-bandwidth transmission equipment meeting 10 Gbps (OC-192) specifications. Since current state-of-the-art DWDM technology offers eight such OC-192 channels, where the fiber can support the rate, PMD can be a serious limitation to system performance and to prospects for upgrading that performance. PMD affects the transmission quality by spreading signal pulses and, therefore, raising the BER of the system. It arises in the first place because of asymmetries in the fiber itself, so the primary remedy must be applied at the manufacturing level. But the damage does not necessarily end there. During installation, the fiber can be crushed, kinked, or otherwise overstressed. Environmental and climatic changes can also affect its circular geometry and thus worsen its PMD characteristics. Postinstallation testing may be needed to ensure that a network does not overly suffer from PMD and that the installed facilities can be upgraded to support tomorrows higher bit rates (Figure 10.2). Second-order PMD, the variation of polarization mode dispersion with wavelength, is considered to have a negligible effect on network performance, however, it acts as a completely random contribution to the networks chromatic dispersion, possibly negating deliberate steps taken in network design to provide the exact amount of this dispersion to reduce nonlinear disturbances in signal propagation. Although this parameter bears watching, its long-term importance cannot yet be predicted. 10.4.2.2 Field Testing DWDM Systems
As previously indicated, the implementation of DWDM transmission systems in the field on a large scale will have a major impact on each level of installation and system verification. New skills will have to be developed to
534
Transmission Systems Design Handbook for Wireless Networks
Sensitivity penalty (dB) 5 4 10
3
5 2 2.5 Gbps 1 0 0
10
20
30 40 50 60 Differential group delay (ps)
70
80
Figure 10.2 PMD effects on system performance.
face these new challenges, and existing test instrumentation will require adaptation. Nevertheless, one instrument emerges as the apparent reference DWDM system characterization toolthe optical spectrum analyzer (OSA). The OSA is suited to almost all the field testing needed in DWDM systems: measurements of signal levels, signal-to-noise ratio, and cross talk, as well as channel spacing and stability. The graphic presentation of modern OSA instruments, clearly showing how the parameter of interest varies with wavelength, gives an excellent overview of many of the phenomena crucial to the proper operation of DWDM networks and valuable clues for the subsequent investigation of any problems that the measurement might reveal. Nevertheless, in many contexts it offers too much information and often not the specific information the field maintainer or troubleshooter needs. Operating and readout procedures and tools must be greatly simplified as compared with those appropriate to laboratory OSAs if the instrument is to be cost-effective in the field (see Figure 10.3). However, to complement OSA testing in the field, center wavelengths must be accurately measured. This parameter can be important, especially if the system under study is part of a larger one whose standards must be respected. Other instrumentation offering more accurate wavelength calibrationa wavelength meter, most likelyis also needed for such operations as the measurement of DFB characteristics.
Transmission-Network Testing and Commissioning
l2
Receivers l1
Erbium-doped amplifiers MUX
DEMUX
l2
l3
l3
l4
l4
OSA OSA OSA wavelength wavelength meter meter ORL tester
OSA
OSA
OSA power meter wavelength meter
Appropriate test instrument
Transmitters l1
535
Figure 10.3 DWDM test instrumentation.
10.4.2.3 New Requirements for Traditional Fiber-Optic Test Instruments
In addition to instrumentation specifically designed for the maintenance of DWDM systems (i.e., the new OSAs and wavelength meters), conventional field installation and test equipment must also be considered because of the strong influence that some of the properties of fiber-optic links have on DWDM transmission. Although many of the basic attributes of these links are independent of the transmission mode used (TDM or WDM) and can thus be measured using conventional instruments, a few parameters are critical to proper DWDM operation, and special care must be taken in selecting field instruments to measure them. Because of the use of several channels at different, precisely defined wavelengths, dedicated WDM power meters must be calibrated at specified wavelengths in the 1,5301,565-nm band, in order to measure the power in individual channels at the output of demultiplexers. Optical loss test sets (OLTSs) will also be used at the wavelengths used for optical supervisory channels (OSCs)1,480 nm, 1,510 nm, and 1,625 nm, depending on the system design. The longest OSC wavelength, 1,625 nm, requires particular attention, since this wavelength lies outside the range in which the fiber or cable manufacturer guarantees the performance of its product. A clear tendency is emerging in the OTDR world to offer capabilities in the fourth window spectral region, at 1,625 nm. In addition to the ability to
536
Transmission Systems Design Handbook for Wireless Networks
test and troubleshoot the important 1,625-nm optical supervisory channel, using this wavelength presents other important advantages. In particular, in many circumstances live fibers may be tested at the 1,625-nm wavelength, while normal DWDM transmission continues uninterrupted in the EDFA spectral region. Because optical losses due to fiber bending are more pronounced at 1,625 nm than at the shorter DWDM operational wavelengths, OTDR testing at the long wavelength can reveal critical points in the installed fiberfor example, places where the performance of the fiber is acceptable at the time of installation but could degrade over time (Figure 10.4). In a conventional (non-WDM) network, the optical return loss (ORL) can be determined with a single measurement using a back-reflection meter at the operating wavelength. In DWDM systems there are two possibilities: an aggregate measure covering the entire wavelength band in use or a detailed one, giving results for each channel wavelength. Although the first is obviously quicker to perform and may provide enough information to satisfy a gono-go acceptance test, ORL can vary considerably from channel to channel. This ORL variation with wavelength may be caused by defective Bragg gratings or, more often, results from bad connectors at the output port of a multiplexer or demultiplexer. Excessive back reflection can cause instability 16 14
Loss on 100 turns [dB]
12 10 8 6 4 2 0 −2 1,300 1,350 1,400 1,450 1,500 1,550 1,600 1,650 1,700 Wavelength (nm)
Figure 10.4 Bending loss versus wavelength.
Transmission-Network Testing and Commissioning
537
in DFB source lasers, thereby affecting the overall system performance. As a result, an ability to perform the more complex wavelength-dependent measurement will often be needed. An aggregate measurement is made with a broadband source and an independent power meter in the same way the measurement is carried out in a single-wavelength optical link. The measurement result is a single valuethe total ORL power at the test point, over the entire transmission spectrum. The value of the ORL as a function of wavelength is often a more useful parameter intrinsically, and its determination may be essential if the simpler aggregate test should fail on a particular link. It is measured using a high-power broadband source, usually an erbium-amplified spontaneous emission (ASE) source. High power is needed to provide enough power in each measurement band, which may be as little as 0.1 nm wide, to give an adequate signal-to-noise ratio at the detector for the lowest ORL of interest. The detector is an optical spectrum analyzer of adequate resolution and sensitivity. The result, of course, is an individual ORL readingoften just the information needed to guide a troubleshooting sessionfor each DWDM channel.
10.5 Packet-Network Testing 10.5.1 Testing the Voice Traffic
A packetized approach to transmitting voice faces a number of technical challenges that spring from the real-time or interactive nature of the voice traffic. Some of the challenges that need to be addressed are described here. Echo is a phenomenon where a transmitted voice signal gets reflected back due to unavoidable impedance mismatch and four-wire/two-wire conversion between the telephone handset and the communication network. Echo can, depending on the severity, disrupt the normal flow of conversation, and its severity depends on the round-trip time delay; if a round-trip time delay is more than 30 ms, the echo becomes significant, making normal conversation difficult. Another problem concerns end-to-end delay. Voice traffic is most sensitive to delay and is mildly sensitive to variations in delay (jitter). It is critical that end-to-end delay is minimized to hold interactive communications. Delay can interfere with the dynamics of voice communication in the absence of noticeable echo, whereas in the presence of noticeable echo, increasing delay
538
Transmission Systems Design Handbook for Wireless Networks
makes echo effects worse. When delay reaches above 30 ms, echo canceller circuits are required to control the echo. Packetization delay (or cell-construction delay) is the time taken to fill in a complete packet/cell before it is transmitted. Normal G.711 PCM encoded voice samples arrive at the rate of 64 Kbps, which means that it can take approximately 6 ms to fill the entire 48-byte payload of an ATM cell. This problem can be addressed either with partially filled cells or by multiplexing several voice calls into a single ATM virtual circuit channel (VCC). Sometimes, due to delay in transit, some cells arrive late. If this happens, the ATM SAR function provided by the adaptation layer might have to underrun, with no voice data to process, which results in gaps in conversation. To prevent this, the receiving SAR function would accumulate a buffer of information before starting the reconstruction. In order to ensure that no underruns occur, the buffer size should exceed the maximum predicted delay. The size of the buffer translates into delay, as each cell must progress through the buffer on arrival at the emulated circuits line rate. This implies that the CDV must be controlled within the ATM network. Voice, by nature, is variable. In fact, a typical conversation has a speech activity factor of about 42% due to pauses between sentences and words where there is no speech in either direction. Also, voice communication is half-duplex, which means that one person is silent while the other speaks. One can take advantage of these two characteristics to save bandwidth by halting the transmission of cells during these silent periods. This is known as silence suppression. G.726 adaptive differential pulse code modulation (ADPCM) and G.729 adaptive code excited linear prediction (ACELP) are the two major compression algorithms that are used. The benefit of compression is efficient use of bandwidth. Most voice packets are transmitted today using G.711 encoding, which does no compression and therefore adds further delay. 10.5.2 ATM Network Testing
Operators expect error-free delivery of data, voice, and video services in their networks. ATM service quality depends not only on error-free transmission facilities but on error-free performance of ATM network elements. ATM network elements, unlike those in SONET networks, support very little protection switching capabilities and circuit redundancy [2]. Since ATM networks are designed for speed, the underlying assumption is that the physical layer will perform flawlessly and therefore there is no need for retransmission. Instead,
Transmission-Network Testing and Commissioning
539
the ATM network places the responsibility for retransmission on the application using ATM. The ATM switch test is conducted during out-of-service periods to verify proper provisioning of the most critical element in any ATM network. Symptoms of an improperly provisioned ATM switch include misrouted, dropped, or incorrectly prioritized cells. These problems can prevent cells from arriving at their destinations or corrupt another users data. Consequently, one or more users of the network may experience lost data, echoing during voice conversations, or a jittery picture during a video transmission. In addition, failure of the ATM switch to regenerate correctable cells or drop noncorrectable ATM cells can result in unnecessary retransmission or incorrect routing. There are three main reasons why ATM technology is currently very difficult to test, and one of them is that ATM is an emerging technology, and therefore products are being developed based on specifications that may have been drafted based only on theoretical projections. As actual working models are produced, the specifications must be modified to match the realities of contemporary production methods. It is expected to take a number of years before ATM reaches a stable state comparable to other products of todays technology. The second reason is complexity of the standards and specifications, and although the concept of ATM appears relatively simple, the high degree of complexity is introduced from the numerous control protocols required to set up and maintain the connections. Another factor is support for multiple priority levels for different traffic types based on service classes (i.e., UBR or ABR). There are basically three types of testing: 1. Conformance testing; 2. Interoperability testing; 3. Performance testing. Conformance testing is used to validate a specific product according to a standard or specification. The equipment is always tested under out-of-service conditions. The ATM test engineer must simulate an operating environment by generating traffic flows or emulating specific protocols. This should include simulating valid traffic flows and network errors to ensure that the device functions correctly under a range of operating conditions. Interoperability testing must be performed to ensure that products from multiple vendors will function properly together in a network. With ATM specifications being constantly updated to meet the increasing demands of users, interoperability
540
Transmission Systems Design Handbook for Wireless Networks
testing is the most critical phase of ATM system verification. Performance testing provides essential measurement criteria for evaluating ATM switch operation. Conformance verification alone does not guarantee that the same equipment will operate satisfactorily under all anticipated network conditions. The ATM edge-to-network test isolates the source of ATM problems between the ATM edge device and the customers or service providers equipment. This is also an out-of-service test and will identify problems before ATM cells have an opportunity to cause alarms or congestion in the network. Carrier signal problems found at this location include configuration and synchronization. User traffic problems uncovered here include bandwidth and priority configuration. The end-to-end ATM network connectivity test verifies the end-toend setup and configuration of all ATM elements responsible for transmitting cells. These elements include ATM switches, SONET/SDH network elements, and SONET/SDH cross-connect devices. This out-of-service test simulates the users addresses and bandwidths, verifies the proper configuration of all of the devices, and verifies that the destination properly receives the ATM cells transmitted. A possible approach for out-of-service monitoring is to establish a virtual path connection (VPC or VCC) at the appropriate measurement point, introduce a test cell stream and timing at that point, and then observe the test cell stream at the remote measurement point. ITU-T O.191 describes equipment and procedures exactly suitable for these purposes. The in-service ATM test is a valuable tool for maintaining proper ATM network operation without affecting the traffic. Since ATM network utilization is hard to predict, ATM network performance can be compromised even after end-to-end simulation pass testing. Very often, the only way to know how the network is performing is to monitor it in-service. This test provides early detection of service degradation and indicates delay variations, congestion, and overall utilization of the circuits. Details of OAM functions supporting performance measurements are provided in ITU-T I.610. Inservice performance monitoring will likely be performed only on a selected number of VPCs or VCCs on an on-demand basis. All of these tests could be accomplished using external test equipment or an advanced networkmanagement system that provides this as well as a more detailed view of the health of the ATM network. The primary issues for ATM-layer testing are to test the equipment function and measure system performance when transporting ATM cells between any two points across a network connection. Each ATM cell can pass through one or several ATM switches during end-to-end transmission. Cells can be processed through buffers (queues) in ATM switches, introducing variable
Transmission-Network Testing and Commissioning
541
delays to the ATM cell traffic throughout the entire ATM network. The cell delay through switch buffers is dependent on the aggregate traffic loading on the network. Switch buffers can also become overloaded and cause a loss of cells. Each ATM switch may implement a different queuing mechanism. The queuing mechanism may be simple or extremely complex depending on the types of ATM service categories supported by the switch (i.e., UBR or ABR). This design may cause cells belonging to different ATM service categories to be processed through different queues in the ATM switch. Even within a single ATM service category (i.e., UBR), different VCCs may be processed through separate queues. These advanced cell-scheduling mechanisms introduce a high degree of complexity to the ATM switches and require extensive testing. ATM switches may also perform shaping of the output cell traffic. This introduces another form of cell delay called jitter. The switch matrix design may include several layers (one for each ATM service category) and this delay may result in cell loss or cell misinsertion (in rare cases). One of the major ATM-layer functions of an ATM switch is to perform VCI or VPI translation from the input to the output. This function is achieved using an address lookup table that is programmable through signaling setup messages or via network management. Cell misrouting and cell misinsertion can occur if the address translation function does not operate correctly. QoS tests are performed under out-of-service conditions. Testing of the ATM layer requires an ATM analyzer capable of generating ATM traffic and then monitoring the same traffic to make specific measurements. The ATM analyzer traffic generator must be capable of transmitting cell streams at full line rate to stress test ATM switches and devices. It should also allow selection of channels from a range of available VCs on the link under test and adjustment of the cell transmission rate and background traffic conditions. These different channels are then mixed together to emulate real-world traffic to ensure that the switch can simultaneously process traffic for different ATM service categories. Use of the ITU-defined O.191 test cell provides the most accurate method to calculate ATM-layer performance parameters. The primary advantage of using the O.191 test cell is that measurements are interpreted in a consistent manner and a standardized format enables monitors to distinguish between cell errors and cell misinsertions. The following ATM layer statistics are significant for measuring ATM QoS: • CTDminimum round-trip time for transmitted (foreground) test
cells;
542
Transmission Systems Design Handbook for Wireless Networks
• Maximum CTDmaximum round-trip time for transmitted (fore• •
•
•
•
•
•
ground) test cells; Mean CTDaverage round-trip time for transmitted (foreground) test cells; Peak-to-peak CDVdifference between maximum and minimum (foreground) test cell delay; generally a good first estimate of the cell delay variation tolerance (CDVT) for the network; CDV distributionhistogram indicating the number of foreground test cells detected having a CTD within a specified range, monitored over specified sampling time intervals during the test period; Cell loss and CLRmeasurement of the difference between the number of foreground test cells transmitted and the number of test cells received; Cell sequence integrityidentifies any out-of-sequence cells by comparing the sequence of received foreground test cells with the transmitted test cells; this is a severe error that should cause immediate notification to the user; Cell error and cell-error ratio (CER)the number of foreground test cells received with single or multiple payload bit errors divided by the total number of transmitted test cells; Cell misinsertion and cell-misinsertion rate (CMR)the number of cells detected on one VC having payload information belonging to another VC; this test provides a good indication that network equipment is overloaded or is reconfiguring and is misrouting or mismulticasting cells.
It is important to note that CMR is calculated as a rate (not a ratio), since it is independent of ATM traffic load. Another valuable feature for ATM-layer testing is the ability to introduce controlled impairments into a live ATM networkdirectly affecting the network QoS levels. This allows the ATM terminal vendors or ATM terminal purchasers to verify equipment operation under worst-case conditions (with cell losses, cell delays, and bit errors). O&M cells are used for alarm surveillance, performance monitoring, and troubleshooting. The O&M cells consist of five flow levels (F1F5) where the highest (F4F5) are dedicated to the ATM layer. The O&M cell traffic is transmitted intermixed with the user cells on each channel. An ATM test analyzer should be capable of sending and receiving O&M cells as
Transmission-Network Testing and Commissioning
543
well as generating the lower-layer O&M signals to properly test the complete O&M functionality. The allocation rules for several of the objectives are based on the G.826 rules for allocating physical-layer performance. Physical-layer impairments contribute strongly to the ATM-layer performance parameters SECBR, CER, and CLR. Performance impairments for each ATM layer parameter grow with increasing distance and complexity. In this context, the term complexity refers to impairments that increase with additional switching and queuing stages or increase as more international and jurisdictional boundaries are crossed. The term distance refers to impairments that are not directly tied to switching or queuing stages and are less directly controllable with the ATM network design. ITU-T I.357 defines availability and unavailability for semipermanent B-ISDN connections [3]. Cell transfer performance measurement methods can be used in determining entry into the unavailable state and in determining when a transition back into the available state has occurred. Properly used, these methods can also be used to develop availability estimates for comparison with availability-related objectives. However, no measurements of I.356 parameters made during periods of unavailable time are ever used in comparison with either long-term cell transfer performance objectives or the I.356 QoS class definitions. In particular, when making either in-service or out-of-service measurements for performance analysis, extreme care must be taken to recognize transitions into and out of periods of unavailability. Mechanisms must be established to exclude all performance-measurement results collected during unavailable periods from any determinations about support of QoS classes and from any estimations of long-term CER, CLR, CMR, SECBR, CTD, frame transmission delay, and corrupted frame ratio performance [4]. A cell block is a sequence of N cells transmitted consecutively on a given connection. A severely errored cell block outcome occurs when more than M errored cell, lost cell, or misinserted cell outcomes are observed in a received cell block. The severely errored cell block outcome and parameter provide a means of quantifying bursts of cell transfer failures and preventing those bursts from influencing the observed values for CER, CLR, CMR, and associated availability parameters. 10.5.3 Quality Tests in VoIP Networks
IP networks were originally designed to carry data that is not time sensitive and therefore cannot guarantee the performance needed for real-time audio transmission and reception. IP-based networks use packet switching that routes packets over many paths. Data can travel over a packet-switched network in
544
Transmission Systems Design Handbook for Wireless Networks
packets that arrive at varying times. A receiver that runs IP can reassemble the packets into the correct order. For voice transmissions, however, it is impossible to tolerate hearing packets of digitized voice in anything but the correct ordera garbled or inconsistent voice would be unintelligible. It is possible to improve the QoS by using special VoIP hardware and software that compensates for many of the problems of sending voice over packet-switched networksdelay, jitter, and lost packets. This equipment will send packets of digitized voice data over a network, and then it decodes the packets, extracts the digitized voice, and converts it back to an analog signal. The following are VoIP network components that must be tested prior to deployment: • Gateway (GW) and media gateway (MGW); • Gatekeeper (GK) and media gateway controller (MGC); • Signaling gateway; • Interactive voice response (IVR) and voice mails; • Billing and prepaid system; • NMS.
After the VoIP network has been proven for functionality, a series of stress tests should be conducted. Ideally, these tests should be performed in a lab environment so as to minimize deployment, troubleshooting, operational, and maintenance costs [5]. A terminal can be an Ethernet phone or a PC containing a VoIP interface card, a headset, and a microphone. VoIP networks can also use analog phones connected to routers, and most VoIP networks require a gateway to connect the VoIP terminals to other networks. Terminals on the same side of a router can communicate directly with each other; calls between people in the same department might not pass through a gateway. While gateways provide the interface to other networks, gatekeepers manage several terminals within a zone on a LAN. Gatekeepers route calls from gateways to terminals, a process called address resolution. Gatekeepers also manage the calls based on the available bandwidth of the LAN. A multipoint unit (MCU) lets a VoIP network carry multiple conversations in one calla conference call. Testing a gateway gets to the heart of the convergence VoIP networkthe connection between the packet side and the circuit side. One has to test the functionality of the gateway and its capability to operate under
Transmission-Network Testing and Commissioning
545
stress. Signaling performance is measured as the GoS and media performance as the QoS. The tests include the generation of a large volume of calls from the circuit side, and analysis of the signaling and media performance of these calls on the packet side. A second stage includes the generation of a large volume of calls from the packet side and analysis of the performance of these calls on the circuit side. Finally, it is recommended that the complete system be tested using an end-to-end test scheme. Each of the components described here requires software that implements communications protocols defined by several ITU recommendations. The ITU has developed Recommendation H.323 as a standard for VoIP protocols. H.323 incorporates other protocol standards, which perform specific tasks on top of the IP protocol. These protocols define audio and video codecs that digitize and possibly compress voice and image signals prior to transport. Not all VoIP equipment follows Recommendation H.323, since the document is just thata recommendation. Most VoIP equipment, however, is based on H.323. It is also expected that most of the future (3G) wireless networks will use some kind of IP over Xtype transport mechanism. Audio-quality tests are essentially QoS tests, measuring the quality of audio across the network under test. To do that, a call generator, test-sound files, and audio-analyzer software in a PC that controls the test equipment are required. A call generator makes calls and performs QoS tests and also tests for the call-setup and call-breakdown messages. Typically, a call generator establishes a call and sends an audio file (such as a WAV file) over the network. The test system digitizes the received audio and analyzes it. Test systems may use any of several methods to analyze speech. All of the applications, including VoIP, voice over frame relay, voice over ATM, or simply testing of the voice quality of PSTN, use the same audio-quality tests. Using these methods, the test equipment attempts to simulate how the human ear perceives incoming audio. A method called perceptual speech quality measurement (PSQM) grades the audio quality on a scale of 0 to 6.5, with 0 being perfect quality. QoS test systems identify QoS problems, but they do not identify where or why a problem occurs. For that, a protocol analyzer is required. Protocol analyzers decode the IP packets, and they also decode the H.323 set of protocols in the packets. It is possible to connect a protocol analyzer at any point in the network if there is a proper line interface, such as Ethernet. Protocol analyzers capture data and look at information such as the order of the received packets and their arrival times. With that information, protocol analyzers can calculate statistics on the delays between sent times and arrival times on a packet-by-packet basis. The statistics will tell if VoIP packets
546
Transmission Systems Design Handbook for Wireless Networks
arrive in acceptable times to produce acceptable voice quality. It is possible to connect a protocol analyzer across a network component, such as a gateway, and measure the time between when a packet arrives and when it departs from the gateway. The analyzer time-stamps IP packets and measures the delay, called latency, across the gateway under test. Typically, IP audio will pass through two gateways, one at each end of the IP network. A gateway is but one contributor to network latency. Each component adds latency to the network. If the latency across a network is too long, it will be impossible to hold a conversation. Typically, people can tolerate delays of no more than about 200 ms to 250 ms before the conversation gets annoying. Delays of 400 ms to 500 ms make conversations impractical. By measuring the latency of each network component, it is possible to build a latency budget that shows where each component adds to the total latency of the network. That information can help to decide if the size of a gateway or gatekeepers jitter buffer is sufficient to deliver acceptable voice quality over a network. In addition, it is important to measure jitter and packet loss across a network. Jitter describes the variations in latency of a VoIP transmission. In data networks, jitter refers to packet jitter, not bit jitter. Too much packet jitter causes voice to sound garbled. Network components compensate for jitter with buffers. Jitter buffers store incoming packets and send them in a more constant stream; the buffers smooth the delivery of packets to produce a more even flow of voice data. The size of a jitter buffer affects both jitter and latency. If audio transmissions have enough jitter to annoy users, then increasing the capacity of a jitter buffer can reduce jitter to acceptable levels. Too large a buffer, however, may cause latency to increase to where it is annoying to users. A typical jitter buffer delay is 20 ms, but often reaches 80 ms. There is no optimal size of jitter buffer, because the buffer size will vary from network to network. Excessive packet loss also affects the QoS of a VoIP system. Typically, more than 5% lost packets will annoy users. Protocol analyzers will display the number and percentage of lost packets. It is important to remember that some VoIP networks may use the Internet as a transport medium. On the Internet, packet latency and jitter are more unpredictable than on a private network. In that case, it is a good practice to add an Internet simulator to the test setup. An Internet simulator can inject impairments into packet transmissions. For example, it might limit the bandwidth of the network or force bursts of lost packets. Limiting the bandwidth can introduce latency into the packet transmissions. Bursts of lost packets result in some voice being lost. When testing VoIP equipment and networks, a recommendation is to simulate more than one call at a time. It is
Transmission-Network Testing and Commissioning
547
important to test VoIP equipment with as many simulated calls as possible to see how the equipment operates under a heavy call load. Billing systems are probably the most mission-critical part of the voice over IP network. If they fail, the service providers bottom line can be adversely affected. It is crucial to ensure call detail record (CDR) integrity when the network is operationalwhich means 24 hours a day, 7 days a week, 365 days a year. CDR integrity consists of the correct transmission and measurement of the following parameters: • Calling line identification (CLID); • Call duration; • ID of called side; • PIN
When the network is used for both voice and data traffic, the billing system should also be able to measure bandwidth used by the customer, as well as the QoS provided. The network management system will typically have connections to the gateway and the gatekeeper of the voice over IP network. It will aggregate and report on network alarms such as overutilization of the assigned bandwidth, bottlenecks, and network degradation situations. This is usually done in two ways: 1. Proactive and preventive: a status report is generated every preconfigured period of time. 2. Breakdown maintenance: alarms are sent when a specific failure has occurred. The testing should include alarm verification when specific failures occur. This can be accomplished by emulating the types of errors that might occur in the real world.
References [1]
Gruber, J., and G.Williams, Transmission Performance of Evolving Telecommunications Networks, Norwood, MA: Artech House, 1992.
[2]
TTC, T-BERD, ATM TestingApplication Note.
548
Transmission Systems Design Handbook for Wireless Networks
[3]
ITU-T I.357, B-ISDN Semipermanent Connection Availability, November 2000.
[4]
ITU-T I.356, B-ISDN ATM Layer Cell Transfer Performance, March 2000.
[5]
Agam, O., Voice over IP Testing: A Practical Guide, RADCOM white paper, June 2001, www.radcom-inc.com.
Appendix: Units Conversion A.1 About Units of Measurement Americans probably use a greater variety of units of measurement than anyone else in the world. Caught in a slow-moving transition from customary to metric units, in the United States there is a fascinating and sometimes frustrating mixture of units of measurement that refer to the same things. Americans measure the length of a running race in meters, but the length of the long-jump event in feet and inches. They speak of an engines capacity in terms of horsepower and its displacement in terms of liters. In the same dispatch, a hurricanes wind speed will be described in knots and its central pressure in millibars. Furthermore, English customary units do not form a consistent system; reflecting their diverse roots in Celtic, Roman, Saxon, and Norse cultures, they are often confusing and contradictory. There are two systems for land measurement (one based on the yard and the other on the rod) and a third system for distances at sea. There are two systems (avoirdupois and troy) for small weights and two more (based on long and short tons) for large weights. Americans use two systems for volumes (one for dry commodities and one for liquids) and the British use a third (British Imperial Measure). Meanwhile, only a few Americans know that the legal definitions of the English customary units are actually based on metric units. The U. S. and British governments have agreed that a yard equals exactly 0.9144m and an 549
550
Transmission Systems Design Handbook for Wireless Networks
avoirdupois pound equals exactly 0.453592370 kilograms. In this way, all the units of measurement Americans use every day are based on the standards of the metric system. Since 1875, in fact, the United States has subscribed to the International System of Weights and Measures, the official version of the metric system.
A.2 The International System of Units All systems of weights and measures, metric and nonmetric, are linked through a network of international agreements supporting the International System of Units. The International System is called the SI, the first two initials of its French name, Système International dUnités [From Recommendation ITU-R V.430-3: Use of the International System of Units (SI)]. The key agreement is the Treaty of the Meter (Convention du Mètre), signed in Paris on May 20, 1875. Forty-eight nations have now signed this treaty, including all the major industrialized countries. The United States is a charter member of this metric club, having signed the original document back in 1875. The SI is maintained by a small agency in Paris, the International Bureau of Weights and Measures (BIPM, for Bureau International des Poids et Mesures). It is updated every few years by an international conference, the General Conference on Weights and Measures (CGPM, for Conférence Générale des Poids et Mesures), attended by representatives of all of the industrial countries and international scientific and engineering organizations. The 21st CGPM met in October 1999; the next meeting will be in 2003. At the heart of the SI is a short list of base units defined in an absolute way without referring to any other units. The base units are consistent with the part of the metric system called the MKS system. In all there are seven SI base units: 1. 2. 3. 4. 5. 6. 7.
The meter for distance; The kilogram for mass; The second for time; The ampere for electric current; The kelvin for temperature; The mole for amount of substance; The candela for intensity of light.
Appendix
551
Other SI units, called SI derived units, are defined algebraically in terms of these fundamental units. Currently there are 22 SI derived units, including: • The radian and steradian for plane and solid angles, respectively; • The newton for force and the pascal for pressure; • The joule for energy and the watt for power; • The degree Celsius for everyday measurement of temperature; • Units for measurement of electricity: the coulomb (charge), volt
(potential), farad (capacitance), ohm (resistance), and siemens (conductance);
• Units for measurement of magnetism: the weber (flux), tesla (flux
density), and henry (inductance);
• The lumen for flux of light and the lux for illuminance; • The becquerel for radioactivity; • The gray and sievert for radiation dose; • The katal, a unit of catalytic activity used in biochemistry.
In addition to the 29 base and derived units, the SI permits the use of certain additional units, including: • The traditional mathematical units for measuring angles (degree, arc-
minute, and arcsecond );
• The traditional units of civil time (minute, hour, day, and year); • Two metric units commonly used in ordinary life: the liter for vol-
ume and the tonne (metric ton) for large masses;
• The logarithmic units Bel and Neper (and their multiples, such as
the decibel );
• Three nonmetric scientific units whose values represent important
physical constants: the astronomical unit, the atomic mass unit, and the electronvolt.
552
Transmission Systems Design Handbook for Wireless Networks
A.3 Common Units 1m = 1.09361 yd = 3.28084 ft = 0.001 km = 6.21371 × 10−4 mi = 39.3701 in; 1 m2 = 1,550 in2 = 10.7639 ft2 = 1.19599 yd2; 1 m3 = 61023.7 in3 = 1.30795 yd3 = 35.3147 ft3; 1 km/hr = 0.277778 m/s = 0.621371 mi/hr = 3.28084 ft/sec; 1 lb = 0.453592 kg; 1 N = 0.1019716 kp = 0.224809 lbf; 1 N/m2 = 10−5 bar = 1.45038 × 10−4 = 0.0208854 lbf/ft2 = 0.101972 kp/m2 = 9.86923 × 10−6 atm; 1 kWh = 3.6 × 106 J = 859.845 kcal = 3,412.14 btu; 1 hp = 76.0402 kpm/sec = 745.700 W; 1W = 0.238846 cal/sec = 3.41214 btu/hr; 1 kcal = 4,186.8 J = 0.745700 kWh.
Glossary AAA authentication, authorization, and accounting AALn ATM adaptation layer type n AAL2 ATM adaptation layer 2. AAL2 has been adopted by ITU-T and ATM forum to reduce packing delay. The idea is to multiplex voice packets from several sources into one ATM cell so that the time to fill a cell can be reduced significantly. ABAM a designation for 22-gauge, 110-ohm, plastic-insulated, twistedpair Western Electric cable normally used in central offices ACS access control server A/D analog-to-digital conversion ADM add/drop multiplexer ADPCM adaptive differential pulse code modulation ADSL asymmetric digital subscriber line AFR absolute frequency reference 553
554
Transmission Systems Design Handbook for Wireless Networks
AGC automatic gain control AIN Advanced Intelligent Network Air interface the radio communications between a mobile handset and the base station (called RBS) AIS alarm indication signal. A signal transmitted to maintain continuity of transmission. The AIS usually is sent to notify the far end that a transmission fault exists on the line. ALBO automatic line build-out AMI alternate mark inversion. A line code in which the signal carrying the binary value alternates between positive and negative polarities AMPS Advanced Mobile Phone System. An analog wireless standard widely used throughout North and South America, as well as the Asia Pacific region and Eastern Europe, that operates in the 800-MHz frequency band. AMR adaptive multirate ANSI American National Standards Institute. Coordinates the development of U.S. voluntary national standards in both the private and public sectors APS automatic protection switch. Ability of the network element to detect a failed unit/line and switch to the spare one; 1+1 pairs a protection unit/line with each working unit/line; 1+n pairs a protection unit/line for every n working units/lines ARIB Association of Radio Industries and Business (Japan) ASE amplified spontaneous emission ASAE adaptive IF slope amplitude equalization ASPR automatic span powering repeater ASTM American Society for Testing and Materials
Glossary
555
Asynchronous a type of transmission in which each character is transmitted independently and without reference to a standard clock Attenuation reduction in signal magnitude or signal loss, usually expressed in decibels ATDE adaptive time domain equalization ATM Asynchronous Transfer Mode ATM Forum organization originally founded by a group of vendors and telecommunication companies; this formal standards body is comprised of various committees responsible for making ATM-related recommendations and producing implementation specifications. ATPC Automatic Transmitter Power Control AT&T American Telephone and Telegraph Company AWG American Wire Gauge Backhaul portion of the wireless network that carries the wireless calls from cell-site radios back to the base station controller (BSC) and the mobile switching center and then on to the appropriate service termination points such as the public switched telephone network and the Internet Bandwidth the information capacity of a communications resource usually measured in bits per second for digital transmission and hertz for analog transmission. Also see narrowband, wideband, and broadband BHCA busy-hour call attempts B8ZS bipolar eight-zero substitution; a line-coding scheme Bellcore Bell Communications Research (called Telcordia now) BER bit-error rate BERT bit-error-rate test
556
Transmission Systems Design Handbook for Wireless Networks
BH busy hour BLER block-error rate. Block of data in which one or more bits are in error. BLER = (errored blocks received)/(total blocks sent) BOC Bell operating company. BOM bill of material (same as BOQ) BONDING Bandwidth on Demand Interoperability Group BOQ bill of quantity (same as BOM) Bluetooth a radio technology developed by Ericsson and other companies built around a new chip that makes it possible to transmit signals over short distances between phones, computers, and other devices without use of wires. Find more information at http://www.bluetooth.com Broadband a classification of the information capacity or bandwidth of a communication channel. Broadband is generally taken to mean a bandwidth higher than 2 Mbps bps bits per second BPV bipolar violation; the detection of any isolated error BSI British Standards Institute BSC base station controller BSS base station subsystem; includes BSC and RBS BWIF Broadband Wireless Internet Forum Canadian Electrical Code (CEC) Canadian version of the U.S. National Electrical Code (NEC) Carrier a telecommunications provider that owns switching equipment
Glossary
557
Calling card a telecommunication credit card with an authorization code for using a long-distance carrier when customers are away from their home or office CAS channel-associated signaling CCC clear channel capability; usually requires B8ZS line coding on all elements CCIR International Radio Consultative Committee (now ITU-R) CCIS common channel interoffice signaling CCITT Comite Consultatif International de Telegraphique et Telephonique (obsolete term). International Telephone and Telegraph Consultative Committee; now ITU CCS common channel signaling Churn a term that denotes loss of customers (turnover) Country code two- or three-digit codes used for international calls outside of the North American Numbering Plan area codes. Dial 011 + country code + city code + local phone number (e.g., 011+91+22+123−4567; 91 = India, 22 = Bombay). CDMA code-division multiple access. The code-division technology was originally developed for military use over 30 years ago. CDMA is a multiple-access technique that uses code sequences as traffic channels within common radio channels. Used for cdmaOne (IS-95) and CDMA2000 air interface. cdmaOne (IS-95) 2G digital air interface technology pioneered by the U.S. firm Qualcomm and further developed in South Korea CDMA2000 A 3G digital air interface technology CE circuit emulation CELP code excited linear prediction
558
Transmission Systems Design Handbook for Wireless Networks
CEPT Conference des Administrations Europeenes des Postes et Telecommunications (Conference of European Postal and Telecommunications Administrations) CFM composite fade margin CIR committed information rate cell. The basic geographical unit of the wireless communications system. Service coverage of a given area is based on an interlocking network of cells, each with a radio base station (transmitter/receiver) at its center. The size of each cell is determined by the terrain (coverage) and forecasted number of users (capacity). Circuit switching basic switching process whereby a circuit between two users is opened on demand and maintained for their exclusive use for the duration of the transmission, as opposed to a dedicated circuit, which is held open regardless of whether data is being sent or not. CLLI code common language location identifier. The CLLI code is used to locate wire centers and switches. CLR cell-loss ratio CN core network CODEC coder-decoder; converts analog voice to digital and vice versa Concatenate the linking together of various data structures (for example, two bandwidths joined to form a single bandwidth) CORBA Common Object Request Broker Architecture CPE customer premises equipment CRC-n cyclic redundancy checkn bits CSU channel service unit; the interface from CPE to the public T1 line CT-2 cordless telephone 2
Glossary
559
CO central office. The building that contains the switch for the local telephone company Cross connect (DSX panel, MDF) distribution-system equipment used to terminate and administer communication circuits. In a wire cross connect, jumper wires or patch cords are used to make circuit connections. In an optical cross connect, fiber patch-cords are used. The cross connect is located in an equipment room, riser closet, or satellite closet. CSES Consecutive severely errored seconds CVSD Continuous Variable-Slope Delta Modulation CWTS China Wireless Telecommunication Standard Group D4 fourth-generation digital channel bank DACS digital access and cross-connect system D-AMPS (IS-136) digital AMPS. The digital wireless standard widely used throughout the Americas, Asia Pacific, and other areas. D-AMPS services can be introduced in the 800-MHz and 1,900-MHz frequency bands. D-AMPS uses a TDMA-based air interface. Dark fiber unused fiber through which no light is transmitted; fiber only, no service or equipment DAT digital audio tape dB decibel dBdsx decibels with respect to the standard level at the DSX-1 cross connect dBm decibels (relative to 1 mw) dc direct current DCT-U digital cordless telephoneUnited States
560
Transmission Systems Design Handbook for Wireless Networks
DE discard eligible DECT Digital European Cordless Telecommunications. A standard issued by ETSI for local-area digital cordless communications. There is a European and American version of DECT systems. DFB distributed feedback DFM dispersive fade margin DLC digital loop carrier DLCI data link connection identifier DOCSIS data-over-cable-service interface specification DoD Department of Defense DoT Department of Telecommunications DS0 digital signal, level 0; 64 Kbps DS1 digital signal, level 1; 1.544 Mbps, the North American standard DS3 digital signal, level 3; 44.736 Mbps, the North American standard DS-41 direct-sequence air interface on an ANSI-41 core network DS/CDMA direct-sequence CDMA DSI digital speech interpolation. Type of silence suppression during the transmission of voice DSSS direct-sequence spread spectrum DSLAC dual-subscriber line audio-processing circuit DSP digital signal processing DSU digital service unit
Glossary
561
DSX-1 digital service cross connect, level 1; part of the DS1 specification DTE data terminal equipment DTMF dual-tone multifrequency; touch-tone dialing DWDM dense-wavelength-division multiplexing EC European Commission ECTA European Commission Trading Authority ECTF Enterprise Computer Telephony Forum EDFA erbium-doped fiber amplifier EDGE Enhanced Data Rates for GSM Evolution EFS error-free seconds EIA Electronic Industries Association. EIA specifies electrical transmission standards, including those used in networking EDI electronic data interchange 802.11b IEEE standard ratified in 1999. Defines wireless LANs EMC electromagnetic compatibility EPROM erasable programmable read-only memory ERA European Regulatory Authority ERO European Radiocommunications Office ESF extended super frame. A DS1 framing format of 24 frames Erlang Agner Krarup Erlang (18781929) was a Danish mathematician who invented the formulas commonly used to forecast telecommunications traffic.
562
Transmission Systems Design Handbook for Wireless Networks
EVRC enhanced variable-rate code E-TDMA Hughes-enhanced TDMA ETSI European Telecommunications Standards Institute. A body formed by the European Commission in 1988, which included vendors and operators. ETSIs purpose is to define standards that will enable the European market for telecommunications to function as a single market. Exchange area the geographic area in which telephone prices and services are the same. An exchange area might have one or more central offices and wire centers. A subscriber in the exchange area could get service from any of the central offices within the exchange area. FA foreign agent FAA Federal Aviation Administration FAS frame-alignment signal FB framing bit FCC Federal Communications Commission. Regulates interstate communications: licenses, rates, tariffs, standards, limitations, and so on. Appointed by the U.S. president. In Canada, the same function is conducted by Industry Canada. FDD frequency-division duplexing FDDI fiber-distributed data interface 5ESS five electronic switching system FDL facility data link. An embedded overhead channel within the ESF format FDM frequency-division multiplexing FDMA frequency-division multiple access
Glossary
563
FEBE far-end block error Federal tax also known as federal excise tax. This tax appears on both local and long-distance phone bills. It is charged as a set percentage regardless of which telephone service provider you use. A little-known fact is that it started as a temporary luxury tax in 1898 on telephone service to pay for the Spanish-American War. Now the proceeds go to the U.S. Treasury as general revenue. For more details on this tax, you can contact the Internal Revenue Service excise tax branch. Federal subscriber line charge also called federal access charge, customer line charge, interstate access charge, interstate single-line charge, FCCapproved customer line charge, subscriber line charge, and SLC. This federally ordered charge billed by the local telephone company pays part of the cost to the local telephone company of supplying a phone line into a home or business. It is designed to help local phone companies recover the cost of providing local loops, which refers to outside telephone wires, underground conduit, telephone poles, and other equipment and facilities connecting end customers to the telephone network. This is not a tax. It is a charge that is part of the price you pay to your local telephone company. Neither the FCC nor any other government agency receives the federal subscriber line charge. The FCC places a maximum cap on this charge. FER forward error correction FERF far-end receive failure FH frequency hopping FHSS frequency-hopping spread spectrum Fixed wireless network also called fixed cellular network or wireless local loop (WLL). The apparent contradiction in terms signifies a cellular network that is set up to support fixed rather than mobile subscribers. The fixed wireless network is increasingly being used as a fast and economic way to roll out modern telephone services, since it avoids the need for fixed wires. 4ESS four electronic switching system FRA fixed radio access
564
Transmission Systems Design Handbook for Wireless Networks
Frame Relay protocol for packet-switched data communications FRAD frame-relay access device FTTC fiber to the curb FTTH fiber to the home FTTN fiber to the neighborhood FWM four-wave mixing GIS Geographic Information System. A computer system for storing, analyzing, and displaying geographic data. GGSN gateway GPRS support node. The GGSN is one of the primary GSM/GPRS core network nodes. GGSN provides interfaces to the external PDNs (public data networks), including connections to ISPs (Internet service providers) and other PLMNs. The GGSN includes functions like routers and servers and thus acts as a router for IP addresses of all subscribers served by the GPRS network. It also outputs charging data for the billing center. The GGSN includes functionality for associating subscribers to the appropriate SGSN and may be connected to a number of SGSNs. GPRS General Packet Radio Service. A GSM data-transmission technique that does not set up a continuous channel from a portable terminal for the transmission and reception of data, but transmits and receives data in packets. It makes very efficient use of available radio spectrum, and users pay only for the volume of data sent and received. Grooming consolidating or segregating traffic for efficiency Global System for Mobile Communications Originally defined as a panEuropean standard for a digital cellular telephone network, to support crossborder roaming, GSM is now one of the worlds main digital wireless standards. Uses TDMA air interfaces and can be implemented in 900-MHz, 1,800-MHz, or 1,900-MHz frequency bands.
Glossary
565
GSM 1800 Also known as DCS 1800 or PCN, GSM 1800 is a digital network working on a frequency of 1,800 MHz. It is used in Europe, AsiaPacific, and Australia. GSM 1900 Also known as PCS 1900, GSM 1900 is a digital network working on a frequency of 1,900 MHz. It is used in the United States and Canada and is scheduled for parts of Latin America and Africa. GSM 900 GSM 900, or simply GSM, is the worlds most widely used digital network and now operating in over 100 countries around the world, particularly in Europe and Asia Pacific. GND Ground (0V) GoS grade of service H.323 ITU standard that defines interoperability standards for voice and multimedia applications over IP. It determines end-point negotiation and information format. It does not, however, address encoding, prioritization, or security. And like other standards, some of the H.323 definitions are subject to interpretation. Therefore, H.323 does not guarantee interoperability between the equipment of different vendors. Some IP vendors are collaborating on an interoperability profile based on ITU H.323 and on the upcoming H.225.0 Annex G standard. Its goal is to achieve interoperability between gateways and gatekeepers of different vendors, enabling deployment of different IP platforms at either end of the network. HA home agent Handoff A frequency channel will be changed to a new frequency channel as the user moves from one cell to another without the users intervention. HATS head and torso simulator HDR high data-rate system developed by Qualcomm for CDMA 1.9-GHz carriers. HF high frequency HiperLAN high performance local area network
566
Transmission Systems Design Handbook for Wireless Networks
HFC hybrid fiber/coax HLR home location register HSCSD high-speed circuit-switched data Hz hertz, cycles per second Iu interconnection point between the RNS and core network. It is also considered a reference point. Iub
interface between RNC and the Node B
Iur a logical interface between two RNCs. While logically representing a point-to-point link between RNCs, the physical realization might not be a point-to-point link. IC integrated circuit ICP IMA Control Protocol IC IXC (IEC is preferred). A company providing long-distance phone service between LECs and LATAs ICMP Internet Control Message Protocol. Provides a number of diagnostic functions in IP networks IDEN Integrated Digital Enhanced Network. Digital TDMA system developed by Motorola (also called ESMR) IEC interexchange carrier IETF Internet Engineering Task Force IEEE Institute of Electrical and Electronics Engineers. Professional organization that defines networking and other standards IFM interference fade margin
Glossary
567
IMA inverse multiplexing for ATM IMT-2000 the term used by the International Telecommunications Union for the specification for the projected 3G wireless services. Formerly referred to as FPLMTS, Future Public Land-Mobile Telephone Systems IMTC International Multimedia Teleconferencing Consortium IMTS Improved Mobile Telephone Service IMUX inverse multiplexer IN intelligent network. A capability in the public telecommunications network environment that allows new services. Also implies a well-developed network infrastructure Interexchange communication between two different LATAs InterLATA communication between local access transport areas. The 1982 MFJ requires LECs to use an IEC for InterLATA services. Interstate communication between states. Interstate communications are regulated by the FCC. IntraLATA communication within a local access transport area. The 1982 MFJ allows LECs to handle these calls without an IEC. Intrastate communication within a single state. Intrastate communications are regulated by each states PUC. Internet the name given to the worldwide collection of networks and gateways using the TCP/IP protocol; the Internet functions as a single virtual network. IP Internet Protocol (See also TCP/IP) IPT IP telephony
568
Transmission Systems Design Handbook for Wireless Networks
IS-54TDMA the first North American TDMA cellular system interim standard specification. After a number of revisions, it is known today as IS-136. IS-95 the specification for the CDMA wireless system IS-2000 CDMA2000 1X ISDN Integrated Services Digital Network. A digital public telecommunications network, in which multiple services (voice, data, images, and video) can be provided via standard terminal interfaces. Offers two times 64 Kbps over the landline network ISM industrial, scientific, and medical microwave bands (2.4 GHz and 5.8 GHz) that do not require licensing (in the United States and Canada, at least) ISO International Organization for Standardization. ISO is responsible for a wide range of standards, including many that are relevant to networking. Their best-known contribution is the development of the OSI reference model and the OSI protocol suite. ISTO Industry Standards and Technology Organization ISUP ISDN user part ITU International Telecommunications Union. Based in Geneva, the ITU is an organization of the UN that oversees telecommunications standards around the world. IXC interexchange carrier. A long-distance telephone company (not an LEC) IWF interworking function JISC Japanese Industrial Standards Committee JPEG Joint Photographic Expert Group. Cooperative effort between ISO and ITU (starting 1986). The original objective of JPEG was to develop a standard for graphics in videoconferencing.
Glossary
569
kA kiloamperes kHz kilohertz, thousand cycles per second LAN local area network LATA Local access transport areas (200 in the United States). A geographic service area defined in the AT&T Modified Final Judgement. The RBOC (Baby Bells) and GTE are restricted to operations within, but not between, LATAs. Long-distance service within a LATA is provided by the LEC. Service between LATAs is provided by an IEC. LATAs are represented by a 3-character code, and there are 164 of them across the country. IntraLATA calls are those that begin and terminate within a local telephone companys territory. InterLATA calls pass outside that jurisdiction to an IXC. Latency the amount of time it takes a packet to travel from source to destination. Together, latency and bandwidth define the speed and capacity of a network. Leased line a permanent telephone connection between two points set up by a telecommunications common carrier. Typically, leased lines are used by businesses to connect geographically distant offices. Unlike normal dial-up connections, a leased line is always active. The fee for the connection is a fixed monthly rate. The primary factors affecting the monthly fee are distance between end points and the bandwidth of the circuit. LEC local exchange carrier. The local telephone company (CO, central office). The local or regional telephone company that owns and operates lines to customer locations and Class 5 Central Office Switches. LECs have connections to other COs, Tandem (Class 4 Toll) offices and may connect directly to IECs like WorldCom, AT&T, MCI, Sprint LEO low Earth-orbit satellite LES loop emulation service LIF Location Interoperability Forum
570
Transmission Systems Design Handbook for Wireless Networks
LEC billing arrangement whereby the local exchange carrier invoices the customer for some or all telecommunications services LMDS local multipoint distribution services in the 28-GHz band Local loop general term for the line from a telephone customers premises to the telephone company central office (CO) Long-distance carrier a company providing long-distance phone service between LECs and LATAs LNP local number portability. This fee started to appear on many local telephone bills in February 1999. It allows local telephone companies to recover costs associated with supporting the technical capability to allow a consumer or business to retain their existing telephone number when switching to another local provider. Local companies are allowed, but not required, to pass on these costs; however, most do. They are only allowed to charge this fee for five years from the first date they start to charge the fee, and are not allowed to start charging the fee until they can provide the ability to the end user of retaining their phone number in the event of switching local telephone companies. Local telephone companies are required to make this number portability service available within six months of being requested to do so by another local telephone company wishing to service the area. This is not a tax. It is a charge that is part of the price you pay to your local telephone company. Neither the FCC nor any other government agency receives the local number portability fee. Local telephone companies are not allowed to charge this fee for customers on the Lifeline Assistance Program. LRN local routing number. Ten-digit number assigned to an individual telephone number to enable local number portability LTE line terminal equipment LTS laser transmission system MAS multiframe-alignment signal MBS Mobile Broadband Systems M24 multiplexer that converts one DS1 line to 24 voice channels for a CO
Glossary
571
M44 multiplexer that converts one T1 of 44 ADPCM channels into two PCM T1s mA milliamperes MGW media gateway MHz megahertz MHSB Monitored Hot Standby Minimum bending radius the amount of bend a fiber or copper cable can withstand before experiencing problems in performance MMDS multichannel multipoint distribution services at 2.1 and 2.7 GHz MPE maxium permissible exposure (to potentially harmful radiation) MPEG Moving Picture Experts Group; part of ISO MPEG-1 an image compression scheme finalized in 1992 and covering coded representation of multimedia, hypermedia, audio, video, and pictures MPEG-2 known within ITU as H.262, the newer, broader standard, which covers high-speed coding efforts between 2 and 15 Mbps for broadcasting, video recording, and other new applications MPI multiple path interference MPIR main path interface at the receiver MPIS main path interface at the transmitter MPLS Multiprotocol Label Switching. A specified framework for routing traffic across multiple IP networks MS mobile station. All types of terminals, installed in vehicles or handheld MSC mobile switching center. The MSC receives calls from the access network and routes them to the called party that is within the same MSC or to a
572
Transmission Systems Design Handbook for Wireless Networks
subscriber located at another MSC. In this case it makes an interrogation to the HLR (home location register), which keeps track of the mobile subscriber locations. The MSC also includes a register VLR (visited location register) of the subscribers located within the MSC service area. MSS mobile switching system. Includes MSC, VLR, HLR, and so on MTSO Mobile Telephone Switching Office. Same as MSC MTBF mean time between failures MTBO mean time between outages MUX multiplexer. Combining two or more signals into a single bit stream that can be individually recovered MWM multiwavelength meter NAMPS Narrowband Advanced Mobile Phones System. Proposed new standard for cellular radio that combines voice processing with digital signaling to triple the capacity and improve overall performance of existing AMPS systems NANP North American Numbering Plan. The NANP is the numbering plan for the public switched telephone network in the United States and its territories, Canada, Bermuda, and many Caribbean nations, including Anguilla, Antigua and Barbuda, Bahamas, Barbados, British Virgin Islands, Cayman Islands, Dominica, Dominican Republic, Grenada, Jamaica, Montserrat, St. Kitts and Nevis, St. Lucia, St. Vincent and the Grenadines, Trinidad and Tobago, and Turks and Caicos Narrowband a classification of the information capacity or bandwidth of a communication channel. Narrowband is generally taken to mean a bandwidth of 64 Kbps or below. NCRP National Council for Radiation Protection and Measurements NEXT near-end cross talk NF noise figure
Glossary
573
NIST National Institute of Standards NIU network interface unit. Test unit installed at the demarcation point NMT Nordic Mobile Telephone. NMT 450 is a low-capacity system and NMT 900 is a high-capacity system. NNI network-network interface NPA numbering plan areas. North American area codes. Three digits: 2 to 9, 0 or 1, 0 to 9 NPL net path loss NTIA National Telecommunications and Information Administration N-WEST National Wireless Electronic Systems Testbed O&M operations and maintenance OA optical amplifier OA&MP operations, administration, and maintenance provisioning OC optical carrier OC-1 optical carrier level 1 (51.84 Mbps) OC-3 optical carrier level 3 (155 Mbps) OC-12 optical carrier level 12 (622 Mbps) OC-48 optical carrier level 48 (2.4 Gbps) OD optical demultiplexer ODSI Optical Domain Service Interconnect. A new optical internetworking initiative OEM other equipment manufacturers
574
Transmission Systems Design Handbook for Wireless Networks
OEO optical-electrical-optical converter OM optical multiplexer ONA open network architecture. A concept intended to permit all vendors of basic or enhanced services, including the phone company itself, to procure basic network functions and interfaces on an unbundled equal-access basis 1x from CDMA2000 1x (IS-2000), derived from 1xRTT, signifies 1 × 1.25 MHz 1xEV-DO 1x Evolution for Data Only 1xEV-DV 1x Evolution for Data and Voice OOF out of frame OOS out of synchronization, or else out of service OPGW overhead optical ground wire OS operations system OSS operations support system OSC optical supervisory channel OSI Open System Interconnect. A seven-layer architecture model for communications systems developed by ISO and used as a reference model for most network architectures today OSNR optical signal-to-noise ratio OTDR optical time domain reflectometer OWAD optical wavelength add/drop OXC optical cross connect
Glossary
575
Packetized voice refers to any means by which voice traffic is split into packets and then transferred to its end destination. This category would include both IP and Internet telephony. Packets are sent separately over available network resources. The packets have headers denoting their place in the message and when the destination is reached, the message is reassembled. The process is quick but not yet as quick as traditional telephony resulting in a delay known as latency. Packet-switched network A packet-switched network breaks up information into digital packets, which are addressed and individually routed and then reassembled in the correct sequence at the destination. These networks allow the medium to be shared, and so are more efficient than circuitswitched networks. Packet switching Packet switching is a method of handling high-volume traffic that allows for efficient sharing of network resources as packets from different sources can all be sent over the same channel in the bit stream. PACS Personal Access Communication Systems. PAM pulse amplitude modulation PBX private branch exchange. A private telephone switching system PCM pulse-code modulation PCN Personal Communication Network PCS personal communications service. A generic term for a mass-market mobile personal communications service, independent of the technology used to provide it PDC Personal Digital Cellular. The digital wireless standard (800 and 1,500 MHz) used in Japan. Uses TDMA air interface PDSN Packet Data Serving Node PF picofarads, a measure of capacitance PHS Personal Handyphone System
576
Transmission Systems Design Handbook for Wireless Networks
PIN personal identification number PL private line. A leased line, not switched PLC Power Line Carrier Plenum an air-handling space such as that found above drop-ceiling tiles or in raised floors. Also, a fire-code rating for an indoor cable PMD polarization mode dispersion PMP point-to-multipoint POI point of interconnect PMU performance monitoring unit PoP point of presence POTS plain old telephone service PPP point-to-point protocol PPS precise positioning system PRS primary reference source. The master clocking source in a network PSTN public switched telephone network. The traditional, wired telephone network PTT post, telephone, and telegraph company. A governmental agency in many countries PUC public utilities commission. A state regulatory body PVC permanent virtual circuit or permanent virtual connection QA quality assurance
Glossary
577
QoS quality of service. A term used to characterize network availability, quality, and reliability. It is also used to designate different classes of ATM service. QRSS quasi-random signal sequence Radio cell the area served by a radio base station in a cellular or cordless communications system. This is where the term cellular came from. Cell size ranges from a few tens of meters to several kilometers RAN radio access network RANOS radio access network operations system RBOC regional Bell operating company RBER residual bit-error rate RBS radio base station Repeater a device used to regenerate an optical or electrical signal (of any frequency) to allow an increase in the system length RFEM radio-frequency electromagnetic field RFI request for information RFP request for proposal RFQ request for quote. Sent to equipment suppliers or service providers to solicit price for a specific item, which could be hardware or services RFT request for tender RLL radio in the loop RJ-11 standard four-wire connectors for phone lines RJ-45 standard eight-wire connectors for IEEE 802.3 1BaseT networks
578
Transmission Systems Design Handbook for Wireless Networks
RNC radio network controller RNS radio network subsystem Roaming ability of a mobile phone user to travel from one cell to another with complete communications continuity. Supported by a cellular network of radio base stations. Also the term given for internetwork operability; that is, moving from one network provider to another (internationally) RLL/WLL radio in the local loop/wireless local loop. A radio access technology that links subscribers into a fixed public telecommunications network. The radio link replaces the traditional wired local loop. rt-VBR real-time VBR RSL receive signal level RX optical receiver RZ return to zero SBS stimulated Brillouin scattering SCADA Supervisory Control and Data Acquisition SCC Standards Council of Canada SCP service control point SCCP signaling connection control part SD space diversity SDH synchronous digital hierarchy SEAL simple and efficient adaptation layer SES severely errored secondseconds with BER of 10−3 or worse SF superframe format. A DS1 framing format of 12 frames
Glossary
579
SGSN serving GPRS support node. The SGSN node is a primary component in the GSM/GPRS core network. SGSN forwards IP data packets to/from connected mobile terminals in the area served by the SGSN. SGSN includes the VLR for packet-switched connections. SGSN provides routing and transfer of data packets to/from the SGSN service area, ciphering and authentication, session management, mobility management, logical link management toward mobile terminals, GPRS-SMS Short Message transfer, and output of charging data to the billing center. The SGSN has an integrated router for interconnection to IP routers. The SGSN can be connected to only one MSC. SLA service-level agreement. The contract between a provider and a customer that sets amount of bandwidth and quality of service, among other things SLAC subscriber line audio-processing circuit SLC subscriber loop carrier SLIC subscriber line interface circuit Smart jack network interface unit with cross-connecting and monitoring capabilities SMDS switched multimegabit digital service SMPTE Society of Motion Picture and Television Engineers SMR Specialized Mobile Radio SMS Short Messaging System SMV select mode vocoder SNMP Simple Network Management Protocol. A set of protocols for managing complex networks. SNMP works by sending messages, called protocol data units (PDUs), to different parts of a network. SNMP-compliant devices, called agents, store data about themselves in management information bases (MIBs) and return this data to the SNMP requesters.
580
Transmission Systems Design Handbook for Wireless Networks
SNR signal-to-noise ratio SONET synchronous optical network SPS Standard Positioning System SPID service provider identifier SRDM subrate data multiplexing SSB single-side band SS7 signaling System 7 SSP service switching point STM synchronous transfer mode. An ITU-defined communications method that transmits a group of time-division multiplexed (TDM) streams synchronized to a common reference clock. It reserves bandwidth according to a rigid hierarchy regardless of actual channel usage. STMN synchronous transfer module level N STP service transfer point. Stratum level of clock source used to categorize accuracy SVC switched virtual circuit Synchronous type of transmission in which the transmission and reception of all data is synchronized by a common clock and the data is usually transmitted in blocks rather than individual characters. Can also mean that the data stream has the same capacity in both directions Synchronous mode standard for data transmission. Data is transferred without start and stop bits together with a clock signal to synchronize the receiver. This mode gives higher data throughput than asynchronous mode, but can be less secure.
Glossary
581
TACS Total Access Communications System. Analog cellular system in 900-MHz band introduced in the United Kingdom in 1985. The modified version for Japan is called JTACS. Tariff a rate, charge, or condition approved by a regulatory agency for a regulated utility TASI Time Assignment Speech Interpolation T1 a 1.544-Mbps transmission standard T1P1 wireless/mobile services and systems technical subcommittee T-BERD a trade name for T1 bit error rate tester made by TTC TCAP Transaction Capabilities Control Part TCP/IP the data protocols used for the Internet TDD time-division duplexing TDM time-division multiplexing TDMA time-division multiple access. A digital transmission technique used for GSM, D-AMPS (IS-136), and PDC air interfaces. D-AMPS in North America is often called simply TDMA. TE terminal equipment Telcordia Technologies Formerly Bellcore, an SAIC company, Telcordia is the worlds largest provider of operations support systems, network software, and consulting and engineering services to the telecommunications industry TFM thermal fade margin 30B+D thirty bearer plus data
582
Transmission Systems Design Handbook for Wireless Networks
3G the next, or third, generation of wireless technology. These networks are specified to operate at a minimum of 2 Mbps when stationary and 384 Kbps when used at pedestrian user speeds. Third-generation standards are coordinated through the ITUs IMT-2000, the European-based UMTS, and the 3GPPa group formed by GSM-supporting standards bodies. 3GPP Third Generation Partnership Project. It was set up to expedite the development of open, globally accepted technical specifications for 3G services. It was responsible for creation of UTMS (W-CDMA) systems. 3GPP2 Corresponding group for transition from cdmaOne to 3G CDMA2000 3x From CDMA2000 3x (IS-2000-A), derived from 3xRTT, signifies 3 × 1.25 MHz TIA Telecommunications Industry Association TM Terminal multiplexer TMN Total or Telecommunications Management Network (ITU-TS M.3010) TND Transmission (Transport) Network Design TOH transport overhead TSI timeslot interchange TTA Telecommunications Technology Association (Korea) TTC Telecommunications Technology Committee (Japan. TU tributary unit TUP telephone user part 20B+D twenty bearer plus data 2B+D two bearer plus data
Glossary
583
2B1Q two binary one quaternary 2W 2-wire TX transmission TIA Telecommunications Industry Association. U.S. telecommunications industry standards body that handles the evolution of D-AMPS and IS-95. UHF ultrahigh frequency UL Underwriters Laboratory. A nonprofit laboratory that examines and tests devices, materials, and systems for safety and has began to establish safety standards; uplink UMTS Universal Mobile Telecommunications System. The European 3G system developed under the auspices of ETSI. It is a highly advanced system optimized for GSM operators. This means that UMTS will, in the long term, support all applications currently served by second-generation cellular systems such as GSM and PDC, cordless systems like DECT, and satellite systems. It will converge contents from the telecommunications industry like video telephony, IT industry like Internet applications, as well as from the broadcasting industry like video on demand. UMTS will have to support this wide array of services with data rates ranging from 8 Kbps to 2,048 Mbps (later even higher) regardless of location, network, or terminal (adaptive terminals). The costs of such terminals could be kept low by being compensated from a mass market generating huge volumes of data traffic. The SMG decided to make UMTS backward compatible to GSM in the beginning, but also to upgrade GSM beyond its initial capabilities. This way, too many compatibility issues would not compromise UMTS development. In order to achieve that goal, UMTS is being developed in a modular way. This also reduces the risk for operators and allows the creation of a consumer basis for mobile data, which will eventually drive the full deployment of the system as its demand for improved service increases. UAS unavailable seconds. Number of seconds elapsed after 10 consecutive SES events are received (ESF framing only)
584
Transmission Systems Design Handbook for Wireless Networks
Unbundling the separate pricing of hardware, software, and related services so that vendors can buy only those services they need. Unbundling is a key to achieving the goal of open network architecture. UNI user-network interface U-NII Unlicensed National Information Infrastructure; 5-GHz microwave band that does not require licensing (in the United States, at least) UPS uninterrupted power supply UTP unshielded twisted pair V.35 a CCITT standard for the trunk interface between a network access device and a packet network. It defines signaling for data rates greater than 19.2 Kbps. V.52 terminal emulation standard. CCITT standard (1976) for various loopback tests that can be incorporated into modems. They are used for testing the telephone circuit and isolating transmission problems. Operating modes include local and remote digital loopback and local and remote analog loopback. V.54 a CCITT standard for loop test devices in modems. It defines local and remote loopbacks. VC virtual channel. A term to describe unidirectional flow of ATM cells between connecting (switching or end user) points that share a common identifier number (VCI) Virtual connection a connection established between end users (source and destination), where packets are forwarded along the same path and bandwidth is not permanently allocated until it is used Virtual circuit a connection set up across the network between a source and a destination where a fixed route is chosen for the entire session and bandwidth is dynamically allocated VCI virtual circuit identifier
Glossary
585
VP virtual path VPC virtual path connection VF voice frequency VG voice grade; the common analog telephone line VHF very high frequency VMR violation monitor removal VMS Voice Mail System VIFDM vector orthogonal frequency-division multiplexing Voice over the Internet calls that travel across the Internet for the main part of their connection between customers. These calls will travel across the PSTN to reach the local point of presence of an ISP before being transported across the Internet, and the PSTN is used for terminating the call. VoIP voice over IP. Refers to calls made over a predominantly native IP (where IP is not run across a layer 2 protocol such as ATM or frame relay) network other than the Internet. This would include any calls made over a company Ethernet network but not calls made across a company ATM WAN that runs IP. Voice over packet refers to calls made over a predominantly packetswitched network. This would include calls made over ATM and frame relay networks as well as networks that run native IP. Due to the lack of traffic management functionality of IP and the associated quality of service issues, IP is run on top of layer 2 protocols in WANs and public networks. VRLA valve-regulated lead-acid batteries. WAN wide-area network
586
Transmission Systems Design Handbook for Wireless Networks
WAP Wireless Application Protocol. A free, unlicensed protocol for wireless communications that makes it possible to create advanced telecommunications services and to access Internet pages from a mobile phone WDM wavelength-division multiplexing WECA Wireless Ethernet Compatibility Alliance Wideband a classification of the information capacity or bandwidth of a communication channel. Wideband is generally taken to mean a bandwidth between 64 Kbps and 2 Mbps WCDMA wideband CDMA. The air-interface technology selected by the major Japanese mobile communications operators, and in January 1998 by ETSI, for wideband wireless access to support 3G services (UMTS). This technology is optimized to allow very high-speed multimedia services such as full-motion video, Internet access, and videoconferencing Wireless communications a system that uses radio transmitters and receivers in place of wire lines which, when connected to the evolving public switched network, provides comprehensive telephone service to customers WLL wireless local loop WTB Wireless Telecommunications Bureau WSP wireless service provider xDSL (generic) digital subscriber line XPM cross-phase modulation ZRBSI zero byte time slot interchange zero code suppression the insertion of a 1 bit to prevent the transmission of eight or more consecutive 0s. Used to guarantee minimum pulse density.
About the Author Born in 1959 in Zagreb, Croatia, Hrvoj (Harvey) Lehpamer received his bachelors degree in 1982 and his masters degree in 1985 from the Department of Radio-communications and Professional Electronics at the School of Electrical Engineering of the University of Zagreb, Croatia. Mr. Lehpamer has 20 years of experience in the planning, design, and deployment of wireless and wireline networks, including microwave, fiberoptic, and other transmission (transport) systems in Europe, Canada, the United States, and other countries. He also has experience in teaching, electronic circuit design, and manufacturing and testing environments. He has worked for such companies as Ericsson Wireless Communications, Qualcomm, Clearnet, Ontario Hydro, Lucas Aerospace, and ElectroprojectConsulting Engineers. He is a licensed engineer and a professional engineer of the Province of Ontario, Canada. Mr. Lehpamer can be reached at
[email protected].
587
Index objectives, 17074 performance, 17879 QoS, 15861 quality, 17879 ring topology, 17273 traffic modeling, 16568 3G wireless-network architecture, 13148 CDMA network, 14245 core, 140 development, 13136 in GSM networks, 14142 horizontally layered, 13640 traffic classes, 14548 UMTS, 141 3G wireless networks, 3, 1519 2G coexistence, 16870 3GPP, 18 CDMA, 14245 characteristics, 1516 concept, 134 data rates, 1819 defined, 1517 environments, 1819 main interfaces, 143 packet call in, 145 radio access, 1617
1G wireless networks, 15 2G wireless networks, 15 3G coexistence, 16870 architecture, 12931 multiplexing, 152 See also Wireless networks 3/1 multiplexing, 7374 3G core network (CN), 140 backbone network connection, 150 defined, 140 evolution from GPRS, 141 3G traffic classes, 14548 background, 148 conversational class, 14647 defined, 145 interactive, 14748 QoS, 146 streaming class, 147 3G transmission networks, 14879 2G coexistence, 16870 AAL2, 15458 architecture, 17079 ATM in, 14854 ATM physical layer, 16165 availability, 17879 cluster topology, 17478
589
590
Transmission Systems Design Handbook for Wireless Networks
3G wireless networks (continued) RNC dimensioning, 32743 standardization process, 1718 voice call in, 144 See also Wireless networks AAL1 circuit emulation, 155 defined, 84 functions, 85 See also ATM adaptation layer (AAL) AAL2, 84, 85, 144 defined, 84 efficiency, 157 functions, 85 importance of, 15458 ITU-T Recommendation I.363.2, 156 minicell header, 150 packing delay and, 153 structure, 85 switching applications, 157 VBR service, 157 See also ATM adaptation layer (AAL) AAL3, 84, 86 AAL4, 84, 86 AAL5 defined, 84 overhead, 86 utilization problem, 85 See also ATM adaptation layer (AAL) Acceptance test procedure (ATP), 116 Access control server (ACS), 99 Accounting management, 515 Acoustic echo/handset specifications, 38384 Acoustic echo return loss (AERL), 384 ac power, 399 Active MW repeaters, 222 Adaptive code excited linear prediction (ACELP), 538 Adaptive differential pulse code modulation (ADPCM), 318, 538 Add-drop multiplexers (ADMs), 69 Administrative unit groups (AUGs), 287 Advanced intelligent network (AIN), 5658 defined, 56
functionality, 57, 58 service control points (SCPs), 56 Advanced Mobile Phone System (AMPS), 6 Aerial fiber-optic cables, 500501 Alternate mark inversion (AMI), 1023, 104 line coding, 120 transmission link, 103 Analog cellular systems, 79 CDPD, 89 frequency reuse, 78 Angle diversity (AD), 211 Angle-of-arrival (AOA) systems, 51819 Antennas gain, 490 GPS, 405 microwave, 35861 microwave, mounting structures, 46685 Application layer, 13638 Application program interfaces (API), 138 Approved sites list, 291 Armored cables, 501 As-built documentation, 41920 Asymmetrical DSL (ADSL), 376, 378 Asynchronous TDM (ATDM), 71 Asynchronous transfer mode (ATM), 74, 7991, 299 availability ratio, 91 availability in, 8991 basics, 7982 bearer network, 133 call control parameters, 8788 cell formats, 81 cells, 81, 151 defined, 79 end points, 80 information transfer parameters, 8889 interfaces overview, 82 inverse-multiplexing for, 8687, 166 IP over, 153 low-BER link design, 164 in microwave radio networks, 16365 MTBO, 91 network elements, 538 networks, 80
Index NMS operation, 165 physical layer, 16165 protection-switching architecture, 163 protocol layers, 83 QoS, 8789 replacing TDM with, 14854 as safe migration path, 132 in SONET/SDH networks, 16263 switch test, 539 UMTS and, 269 in UTRAN networks, 329 virtual circuit channels (VCCs), 538 See also ATM adaptation layer (AAL); ATM testing ATM adaptation layer (AAL), 81, 8386 AAL1, 84, 85, 155 AAL2, 84, 85, 144, 150 AAL3 and AAL4, 84, 86 AAL5, 84, 85, 86, 145 defined, 81 at entry/exit points, 84 function, 83 services, 84 types, 8486 ATM testing, 53843 conformance, 53940 difficulty, 539 edge-to-network test, 540 end-to-end network connectivity test, 540 in-service test, 540 issues, 54041 QoS tests, 541 types of, 539 See also Asynchronous transfer mode (ATM) Attenuation cable, 187 fiber-optic transmission, 18385 rain, 21719 Automatic line build-out (ALBO), 103 Automatic transmitter power control (ATPC), 219 Availability in ATM networks, 8991 microwave link design, 21516 SONET, 195
AWM cable, 385 Background class, 148 Backhaul, 6162, 143 cell-site, 230 defined, 61 technological developments, 6162 Backhoe fade, 61 Bandwidth DWDM, 19798 efficiency, 154 FDMA and, 9 fiber-optic transmission, 18385 formula, 333 GPRS and, 12 as key resource, 161 Bandwidth-on-Demand Interoperability Group (BONDING), 77 Batteries, 4013 flooded, 401 valve-regulated, 401 VRLA, 400, 4013 See also Power Biconic connectors, 505 Billboards, 468 Bill of quantity (BOQ), 417, 418 Binary phase-shift keying (BPSK), 37 Bipolar eight zero substitution (B8ZS), 102, 104, 522 Bit error rate test (BERT), 116 Bit error ratio (BER), 88 defined, 114 T1, 1056 BITS clock system, 31213 defined, 312 illustrated, 313 BIX Cross-Connect System, 387 Bluetooth, 3944 air interface, 41 baseband controllers, 42 baseband protocol, 41 connection support, 41 CVSD modulation, 41 defined, 39 encryption, 43 error-correction schemes, 42 forward error correction (FEC), 40
591
592
Transmission Systems Design Handbook for Wireless Networks
Bluetooth (continued) frequency bands, 40 functions, 41 headset, 44 link support, 42 robustness, 40 uses, 4344 Broadband Wireless Access (BWA), 3135 BSC-MSC connectivity, 17071, 293 Building Industry Consulting Service International (BICSI), 506 Building Integrated Timing Supply (BITS), 303 Business plan, 2 Busy-hour traffic (BHT), 51 Cabling, 38492 digital cross connects, 38688 DS1 signal termination, 38892 fiber-optic, 495509 NEC categorization, 38486 NIU and, 392 T1-class, 391 See also Transmission equipment CBR equivalent bandwidth, 166 QoS, 156 CDMA2000, 16, 17, 1922, 143 1x, 19, 23 1xEV-DO, 2224 3x, 1920 defined, 19 PCN, 20 topology, 21 CdmaOne, 14 Cell-delay variation (CDV), 88, 89 Cell-insertion ratio (CIR), 88 Cell-loss ratio (CLR), 88 Cell-of-origin (COO) systems, 517 Cell-packing delay, 159 Cell sharing, 156 Cell-site requirements, 461 RF, compliance, 45354 selection, 45152 timing, 31617 Cell-transfer delay (CTD), 89
Cellular digital packet data (CDPD), 89 bit rates, 8 defined, 8 Cellular radio, 5 Cement poles, 467 CEPT, 62 digital hierarchy, 6566 E1 digital hierarchy, 68 establishment, 428 members, 42829 National Tables of Frequency Allocations, 428, 429 pulse-code modulation hierarchy, 67 reorganization, 42930 Change orders, 443, 444 Channel-associated signaling (CAS), 158 Channel service unit/data service unit. See CSU/DSU Chemical wells, 397 Chimney racks, 466 Chromatic dispersion, 53233 Circuit emulation (CE) defined, 168 structured, 169 unstructured, 16970 Circuit-switched networks, 29798 Circuit-type services, 166 Civil construction, 47074 concrete strength, 474 deflection formula, 472 erosion and, 473 handling, 470 mounting and, 473 painting and, 473 standards, 473 steel, 472 strength requirements, 474 timelines, 474 Clock-system arrangements, 310 Cluster topology, 17478 defined, 174 design factors, 17475 intercity connection, 176 RBS, 175 redundancy configuration, 177 remote BSCs and, 17778 Coaxial cable, 59, 60
Index Code-division multiple access (CDMA), 6 3G network, 14245 defined, 13 flavors, 14 IS-95 standard, 1314 narrowband, 14 as spread-spectrum technology, 14 synchronization and, 31617 technology benefits, 14 wideband, 16, 17 Code-excited linear prediction (CELP) coders, 319 Coding and compression delay, 159 Common Channel Interoffice Signaling (CCIS), 54 Common channel signaling (CCS), 158 Common Management Information Protocol (CMIP), 511, 512 Communication management, 431 Companding, 127 Competitive access providers (CAPs), 265 Composite fade margin (CFM), 208, 209 Configuration management, 513 Connectivity layer, 13840 defined, 138 illustrated, 139 solutions, 140 Connectors, 5045 biconic, 505 D4, 505 FC, 505 FDDI, 505 SC, 505 SMA, 505 ST-compatible, 5045 See also Fiber-optic cables Continuous variable-slope delta (CVSD) modulation, 41 Contract negotiations, 421 Conversational class, 14647 Coordinate systems, 24047 datums, 24041 facts, 247 geometric Earth models, 24142 GPS, 24446 reference ellipsoids and, 24243 See also Microwave systems
593
Core transport, 170 Cost management, 431 Cranes model, 207, 230 Cross connects, 201 CSU/DSU, 37576 defined, 375 loopback, 375 T1 framing, 376 See also Wireline equipment Customer premises equipment (CPE), 11920 Customer requirements analysis, 26264 Customer service representative (CSR), 27475 D4 connectors, 505 Daisy chaining, 31718 Dark fiber/dark copper, 344 Data error rates, 11415 Datums, 24041 conversions, 241 defined, 240 types, 241 dc power, 400 Dedicated leased service, 29496 defined, 294 features, 294 See also Leased lines Demarcation point, 386 Dense-wavelength-division multiplexing (DWDM), 196201 capacity requirements, 19798 defined, 196 deployment, 198 field testing systems, 53334 flexibility, 198200 implementation, 53334 layer, 150 multiplexers, 36667 optical layers, 200 overview, 19697 protection, 200201 specifics, 53233 target customers, 197 testing fibers for, 53237 test instrumentation, 535 Differential GPS (DGPS), 24546
594
Transmission Systems Design Handbook for Wireless Networks
Digital-access cross connects (DACS), 299, 37074 application, 371 application with single BSC, 372 blocking, 371 continuous performance monitoring, 373 defined, 370 mini, 374 nonblocking, 371 services, 37173 types of, 371 See also Wireline equipment Digital cross connects (DSX), 38688 DSX-1, 38889 DSX-3, 387 panels, 387 signal level example, 390 See also Cabling Digital network hierarchy, 305 Digital speech interpolation (DSI), 156 Digital subscriber line (DSL), 133, 29697, 37679 ADSL, 376, 378 bridge tap, 297 defined, 296 HDSL, 296, 376, 377 IDSL, 296 loading, 296 Digital subscriber loop access multiplexer (DSLAM), 379 Digital transmission, 5966 CEPT hierarchy, 6566 copper lines, 59, 60 defined, 59 DSX-1 digital interfaces, 6263 fiber-optic, 6061 media (physical layer), 5961 North American hierarchy, 6365 satellite, 60 schemes, 59 wireless, 60 Direct-buried cables, 501 Direct-sequence spread spectrum (DSSS), 232, 233 defined, 232 modulation processes, 233
See also Spread-spectrum systems Dispersive fade margin (DFM), 208, 209 Dispersive fading, 207 Distribution poles, 34546 Diversity, 21012 angle (AD), 211 frequency (FD), 211 hybrid (HD), 212 space (SD), 21011 Driven rods, 39697 DS1 connectivity, 304 performance objective, 528 signal precompensation, 38890 signal termination, 38892 test procedure, 529 DS3, 29596 channelized, 29596 nonchannelized, 296 performance objective, 529 DSX-a digital interfaces, 6263 defined, 62 illustrated, 63 Due diligence, 43536 E1 system, 6566, 11927 customer premises equipment (CPE), 11920 defined, 295 digital hierarchy, 68 framing formats, 12224 framing structure illustration, 123 framing synchronization, 122 global framing formats, 127 HDB3, 12122 introduction, 119 link illustration, 121 pulse density, 121 signal characteristics, 120 spare bits, 12427 testing, 52324 transmission facilities, 12021 tutorial, 11927 Echo, 537 Echo cancellers, 37984 applications, 379 in long-distance networks, 380
Index in packet-switched networks, 383 specifications, 38384 in VoX networks, 38283 in wireless networks, 38082 in wireline networks, 380 See also Wireline equipment Electric transmission towers, 345 Electrolytic rods, 39798 Electromagnetic compatibility (EMC), 181 Electromagnetic fields (EMFs), 485 from active radiator, 492 sources, 49192 Electronic probes, 494 Encryption, Bluetooth, 43 End-to-end delay, 53738 End-to-end testing, 522 Engineering services, 430 Enhanced observed time difference (E-OTD) systems, 516, 518 Enhanced variable rate codec (EVRC), 321 Environmental monopoles, 466 Environmental stress screening (ESS), 363 Equipment/services order processing, 41325 as-built documentation, 41920 contract negotiations, 421 equipment/service suppliers negotiation, 425 planning and design, 41314 request for information (RFI), 414 request for pricing (RFP), 414 request for quote (RFQ), 41519 responsibility matrix, 420 statement of work negotiation, 42122 telecommunication providers negotiation, 42225 See also Services; Transmission equipment Equivalent isotropic radiated power (EIRP), 36 Erlang B, 51, 52 Erlang C, 5253 European Conference of Postal and Telecommunications Administration. See CEPT European Radiocommunications Office (ERO), 430
595
Factory acceptance testing, 407 Fade margins, 20710 composite (CFM), 208, 209 dispersive (DFM), 208, 209 interference (IFM), 2089 rain and, 218 thermal, 207 Failure mode and effects analysis (FMEA), 408 Fake trees, 467 Far-field region, 48889, 491 Fast facility protection (FFP), 178 Fault management, 51314 Fault tree analysis (FTA), 408 FCC, 6 defined, 278 license issuing, 279 Wireless Telecommunications Bureau (WTB), 27879 FC connectors, 505 FDDI connectors, 505 Federal Communications Commission. See FCC Fiber-optic cables, 495509 aerial, 500501 all-dielectric outside plant, 499500 armored, 501 conduit, 506 connectors, 5045 design considerations, 49596 direct-buried, 501 duct utilization, 508 fiber count, 5013 fiber protection, 49798 fiber splicing, 5034 handling, 5067 indoor/outdoor, 499 installation, 49596 installation procedures, 5079 intrabuilding regions, 496 maximum tensile rating, 507 minimum bending radius, 5067, 509 multimode fiber, 498 outside diameters, 506 patch cords, 500 pulling plan, 5089 record keeping, 509
596
Transmission Systems Design Handbook for Wireless Networks
Fiber-optic cables (continued) single-mode fiber, 498 tight-buffered, 499 types of, 498501 Fiber-optic cable testing, 53037 DWDM applications, 53237 instrument requirements, 53537 OTDR test procedure, 53132 See also Transmission-network testing Fiber-optic equipment, 36370 ITU-T recommendations, 36768 multiplexers, 36566 OPGW, 36870 SONET/SDH, 36368 See also Transmission equipment Fiber-optic systems features, 182 fiber cross section, 182 fiber size, 183 Fiber-optic transmission, 6061, 181202 attenuation, 18385 bandwidth, 18385 basics, 18183 design principles, 18389 distance calculations, 18589 losses, 188 power budgets, 18589 Final equipment list, 291 First article test (FAT), 4067 First-in-first-out (FIFO) buffers, 77 Fixed microwave systems. See Microwave systems Fixed satellite systems, 2627 Forward error correction (FEC), 40 Fractional T1 (F-T1), 118 Frame relay, 298, 299 Free-space loss, 2034 Frequency-diversity (FD) systems, 211 Frequency-division multiple access (FDMA), 6 bandwidth and, 9 defined, 9 Frequency-division multiplexing (FDM), 59, 71 Frequency-hopping spread spectrum (FHSS), 23233 defined, 232
dwell time, 233 modulation processes, 23233 See also Spread-spectrum systems Frequency reuse, 78 Frequency shift keying (FSK), 37 Fresnel zone, 462 Fronthaul, 143, 293 Future Public Land Mobile Telecommunications Systems (FPLMTS), 17 Gain antenna, 490 multiplexing, 153, 165 passive MW repeater, 22324 Gateway Control Protocol (GCP), 142 Gaussian minimum-shift keying (GMSK), 13 General Packet Radio Services (GPRS) advantages, 11 bandwidth sharing, 12 cell capacity and, 12 defined, 11 facilitation, 1112 GMSK, 13 instant connections, 11 service developments, 13 terminals, 12 Geographic partitioning, 516 Geometric Earth models, 24142 Global framing formats, 127 Global Positioning System (GPS), 154, 24446 antennas, 405 basis, 241 carrier-phase measurements, 246 defined, 244 differential (DGPS), 24546 differential techniques, 24546 holdover time, 317 PPS predicatable accuracy, 244 project costs, 246 satellites, 244 SPS predicatable accuracy, 245 time standard, 317 timing, 316
Index Global System for Mobile Communications (GSM), 6 3G in, 14142 defined, 10 system, 10 Grounding, 39299 basics, 39294 chemical wells, 397 critical nature of, 393 defined, 392 driven rods, 39697 electrolytic rods, 39798 for PCS sites, 398 planning stage, 394 soil resistivity and, 395 system design fundamentals, 39596 system functions, 394 types of, 39698 ufer, 397 water pipes, 397 for wireless cell sites, 39899 See also Transmission equipment Ground potential rise (GPR), 347 Guyed towers, 47679 defined, 476 materials, 476 material selection, 479 surface requirements, 477 See also Towers H.323 components, 92 terminals, 92, 94 zones, 94 HDB3, 12122 coding sequences, 121 defined, 121 line coding, 122 Head and torso simulator (HATS), 384 High-bit-rate digital subscriber line (HDSL), 293, 296, 376 bidirectional service, 377 circuits, 377 HDSL2, 377 See also Digital subscriber line (DSL) High-performance local area network (HIPERLAN), 36
597
High-speed circuit-switched data (HSCSD), 13 Horizontally layered network architecture, 13640 application layer, 13638 connectivity layer, 13840 illustrated, 137 layers, 136 network control layer, 138 Hub multiplexers, 366 Human resources management, 431 Hybrid diversity (HD), 212 Hydro towers, 466 Iub traffic load, 33739 circuit-switched, 33839 defined, 337 packet-switched, 339 total, 338 voice, 338 See also Traffic calculations Iur traffic load, 34041 calculation, 340 circuit-switched, 341 defined, 340 packet-switched, 341 voice, 34041 See also Traffic calculations Iu traffic load, 33537 circuit-switched data, 336 components, 335 packet-switched, 337 voice, 336 See also Traffic calculations Improved Mobile Telephone Service (IMTS), 5 Infrastructure housing, 46465 In-service monitoring, 526 Instructional Television Fixed Service (ITFS), 3334 Integrated services digital network (ISDN), 10 broadband, 80 digital subscriber line (IDSL), 296 Integrated voice and data services, 25 Integration management, 431 Interactive class, 14748
598
Transmission Systems Design Handbook for Wireless Networks
Interexchange carriers (IXCs), 48 Interference fade margin (IFM), 2089 RF, 454 International spare bits, 12425 International Telecommunications Union (ITU), 16 data rates, 69 ITU-R, 17, 207, 213 ITU-T, 213 ITU-T fiber optic equipment recommendations, 36768 microwave link design objectives, 21215 Internet chat groups, 262 e-mail services, 263 Engineering Task Force (IETF), 100 technologies, 2425 Interoffice distribution, 31012 Interworking functions (IWFs), 158 Intraoffice distribution, 31213 Inverse multiplexing, 7479 adding/removing channels, 76 of ATM, 166 bit-based, 75 delay issues, 76 development, 74 features, 7578 metaframing, 7879 See also Multiplexing Inverse multiplexing for ATM protocol (IMA), 8687 control protocol (ICP), 87 defined, 86 group, 87 IP-based wireless networks, 99102 admission control, 100 deployment, 102 illustrated, 101 QoS control, 100 IP telephony (IPT), 13536 ISM bands, 37, 39 Japanese Digital Cellular (JDC), 9 Japanese digital hierarchy (J1), 118 Jitter, 546
buffer depth, 9596 calculation, 97 Label-switched paths, 98 Laser transmission systems (LTS), 34849 Latency, 32123 access, 321 backbone, 95 CODEC, 95, 97 defined, 321 gateway contribution to, 546 jitter, 546 measuring, 9697, 546 network, 321 Lawful intercept (LI), 28081 Layered backbone network model, 132 Leased lines, 292302 access and core transmission networks, 29394 dedicated service, 29496 higher-speed switched/nonswitched services, 299 microwave comparison, 302 network interface unit (NIU) and, 392 owned vs., 3012 provider negotiations, 422 scheduling, 292 switched service, 29798 tracking process, 44143 xDSL, 29697 Leased lines network build-out, 299301 system deployment/testing, 300 system design, 300 system optimization, 301 system planning, 300 Life cycle, 28183 Lightning, 11012 aerial cables and, 110 defined, 110 protectors, 11012 strikes, 110 See also T1 system Line build-out (LBO), 105 Live-data emulation, 522 Local-area networks (LANs) defined, 49 internetworking, 77
Index traffic, 82 Local exchange carriers (LECs), 48 Local multipoint distribution service (LMDS), 31 band designation, 32 licensing, 32 Location-finding techniques, 51619 AOA systems, 51819 COO systems, 517 E-OTD systems, 518 implementation, 517 Location Interoperability Forum (LIF), 271 Loopback testing, 522, 525 Maintenance, 436, 50919 Major trading areas (MTAs), 261 Marketing, 27678 Mean opinion score (MOS), 275 Mean time between failures (MTBF), 407 Mean time between outages (MTBO), 91 Measurement procedures, 49495 Media Gateway Control Protocol (MGCP), 379 Media gateways (MGWs), 132, 133 Metaframe facility data link (MFDL), 78 Metaframes, 7879 Metropolitan area networks (MANs), 196 Metropolitan statistical areas (MSAs), 261 Microwave antenna mounting structures, 46685 billboards, 468 building elevator shaft mounts, 468 building flat mounts, 467 building pole mounts, 468 building tower mounts, 468 ceilings, 468 cement poles, 467 chimney racks, 466 civil construction, 47074 county/regional planning role, 46870 environmental monopoles, 466 equipment mounting, 473 erosion and, 473 fake trees, 467 fall-prevention systems, 48183 guyed towers, 47679 hidden sites, 468
599
hydro towers, 466 lamp standards, 468 microwave/radar towers, 466 minimum-visual-impact, 48081 monopoles, 466, 471, 47679 overview, 46668 real trees, 467 self-support towers, 47679 stability requirements, 47273 tower lights, 471 tower-strength requirements, 474 tripod, 47980 See also Microwave deployment Microwave antennas, 35861 beamwidth, 259 defined, 358 deflection formula, 472 illustrated, 360 maximum allowed deflection, 47576 options, 36061 parabolic, 359 parameters, 35960 sizes, 358 types of, 360 See also Microwave antenna mounting structures Microwave colocate, 45556 Microwave deployment, 45485 antenna mounting structures, 46685 infrastructure housing, 46465 path survey, 46164 scope of work, 45456 site surveys, 45661 Microwave-link design, 20630 availability, 21516 diversity improvement, 21012 fade margins, 20710 frequency coordination, 22021, 22526 guidelines, 22628 illustrated, 227 interference analysis, 22021, 22526 microwave lookup table, 22930 objectives, 21215 output, 221 overview, 21921 path calculations, 2067, 22425
600
Transmission Systems Design Handbook for Wireless Networks
Microwave-link design (continued) practical aspects, 21930 protected/nonprotected, 22122 questions, 220 rain and, 21719 reliability, 21516 repeaters, 22224 ring configuration, 22728 routing, 220 steps, 228 theoretical aspects, 20619 Microwave lookup table, 22930 defined, 229 illustrated, 229 worksheet, 229, 231 Microwave network management, 24754 alarm-management application, 248 application software modules, 248 configuration-management application, 24849 domain manager, 24748 element-management application, 248 network configurations, 247 performance-monitoring application, 249 with SNMP, 24954 Microwave path calculations, 22425 Microwave path survey, 46164 physical, 464 recommended equipment, 463 undertaking, 463 Microwave/radar towers, 466 Microwave radio equipment, 35163 antennas, 35861 environmental/quality issues, 36263 PDH/SDH microwave radios, 35152 rack, 356 split configuration, 35358 standard configuration, 35253 standards and recommendations, 363 standards compliance, 36465 transmission lines, 361 See also Transmission equipment Microwave relocation, 267 Microwave repeaters, 22224 active, 222 passive, 22224
site selection, 45253 Microwave safety, 23640 guidelines, 240 maximum-permissible-exposure (MPE), 236, 237, 239 Microwave sites checklist, 459, 460 feasibility analysis, 458 surveys, 45661 Microwave systems, 2844 ATM in, 16365 compatibility and safety, 23640 coordinate, 24047 design guidelines, 22628 design illustration, 227 disadvantages, 302 feasibility analysis, 458 frequencies, 28 frequency effects, 205 link design, 20630 managing, 24754 multipath effects, 206 nonprotected, 22122 planning, 457 point-to-multipoint, 3144 point-to-point, 2831, 20354 protected, 22122 scope of work, 45456 spread-spectrum, 23036 terrain effects, 205 testing, 52728 transmission theory, 2036 tropospheric effects, 205 See also Wireless networks MIL-HDBK-217, 40710 Mini-DACS, 374 Mobile broadband systems (MBS), 33 Mobile information society (MIS), 148 Mobile positioning, 27172 Mobile QoS (M-QoS), 16061 Mobile satellite systems, 2728 Modified final judgment (MFJ), 58 Monitoring Center (MC), 28081 Monopoles, 466, 471 Multiband excitation (MBE) speech model, 320
Index Multichannel multipoint distribution service (MMDS), 39 broadband networks, 39 integration of, 39 Multiframe-alignment signal (MAS), 12324 Multiplexers, 36566 defined, 365 DWDM, 36667 hub, 366 terminal, 366 Multiplexing, 7079 2G, 152 3/1, 7374 defined, 70 frequency-division (TDM), 59, 71 gain, 153, 165 implementation, 7071 inverse, 7479 on packet level, 165 on transaction level, 165 statistical, 7273 subrate, 7374 time-division (TDM), 59, 71 traffic streams, 165 Multiprotocol label switching (MPLS), 9798 defined, 97 domain, 98 explicit routing support, 98 label-switched paths, 98 Multiservice networks, 50911 MW LOS survey report, 291 MW path and frequency-planning report, 291 MW radio tests, 36263 National bits, 12527 National Council for Radiation Protection and Measurements (NCRP), 236 National Electric Code (NEC) cable categorization, 38486 defined, 384 guidelines, 384 National Table of Frequency Allocations (NTFA), 428, 429 Near-field region, 48889
601
Negotiations contract, 421 equipment/service suppliers, 425 statement of work, 42122 telecommunications providers, 42225 Network control layer, 138 Network Equipment Building Standard (NEBS), 41112 Bellcore/Telcordia documents, 41112 defined, 411 Network interface units (NIUs), 112 as demarcation point, 392 leased lines and, 392 Network loopback, 37576 Network management system, 51115 accounting management, 515 configuration management, 513 fault management, 51314 performance management, 51415 planning, 512 security management, 515 Network performance, 27376 Network-to-network interface (NNI), 81 Nonvoice wireless services, 26265 corporate e-mail, 363 document sharing, 263 file transfer applications, 264 home automation applications, 264 Internet chat groups, 262 Internet e-mail services, 263 moving images, 262 still images, 262 textural/visual information, 262 vehicle positioning, 26364 Web browsing, 26263 wireless LAN access, 264 Nordic Mobile Telephone (NMT), 15 North American Digital Cellular (NADC), 9 North American digital hierarchy, 6365 data rates, 68 pulse-code modulation, 64 T1, 65 Not-word frames, 123 Open service architecture (OSA), 136 Operations and maintenance, 50919
602
Transmission Systems Design Handbook for Wireless Networks
Operations and maintenance (continued) geographic positioning, 516 location-finding techniques, 516 multiservice network growth, 50911 network management system, 51115 Optical carrier level 1 (OC-1), 193, 194 Optical Domain Service Interconnect (ODSI), 202 Optical laser communications, 34849 Optical losses measuring, 531 test sets (OLTSs), 535 Optical power budgets, 18589 calculating, 185 calculator, 188 calculator example, 189 See also Fiber-optic transmission Optical power ground wire (OPGW), 348, 36870 advantages, 36869 cable, 368 cross section, 370 design, 368 installation, 369 Optical return loss (ORL), 536, 537 Optical spectrum analyzers (OSAs), 534 Optical supervisory channels (OSCs), 535 Optical switching, 2012 cross connects, 201 ODSI, 202 Optical time domain reflectors (OTDRs), 53132 capabilities, 531 defined, 531 test procedure, 532 Optimization, 31724 activities management, 44345 daisy chaining, 31718 network example, 32324 signal propagation delay, 32123 traffic grooming, 31718 voice compression, 31821 See also Transmission-network design Order wire, 110 Outdoor equipment cabinets, 464 Out-of-service testing, 52426 conducting, 524
defined, 524 Outsourcing services, 43234 decision, 434 feasibility studies, 434 Packet core network (PCN), 20, 99 Packet data serving node (PDSN), 2022 Packetization delay, 538 Packet-network testing, 53747 ATM network, 53843 voice traffic, 53738 VoIP networks, 54347 See also Transmission-network testing Packet-switched networks, 298 echo cancellation in, 383 frame relay, 298 Packet switching (PS), 50 Packet-type services, 166, 167 Passive MW repeaters, 22224 correction factor, 224 function of, 222 gain calculations, 22324 See also Microwave repeaters Patch cords, 500 PathLoss 4.0, 229, 231 Performance DS1/DS3, 52829 management, 51415 measurement, 432 SONET, 191 transmission network, 17879 Performance-monitoring units (PMUs), 12627 Personal communications service (PCS), 6, 31416 Personal Digital Cellular (PDC), 10 Physical layer, 16165 defined, 161 sublayers, 161 See also Asynchronous transfer mode (ATM) Plain old telephone service (POTS), 7, 47 Planning agencies, 46870 Plesiochronous digital hierarchy (PDH), 6870 limitations, 190 microwave radio, 35152
Index network design example, 32425 transmission systems, 286 Point-to-multipoint systems, 3144 Bluetooth, 3944 broadband wireless access, 3135 characteristics, 32 LOS problem, 3435 services, 3334 system architecture, 34 TDD vs. FDD in, 35 U-NII band, 3539 See also Microwave systems Point-to-Point Protocol (PPP), 20 Point-to-point systems, 2831, 20354 advantages, 30 compatibility and safety, 23640 coordinate, 24047 defined, 28 growth areas, 2930 illustrated, 29 link design (practical aspects), 21930 link design (theoretical aspects), 20619 managing, 24754 performance in U-NII band, 38 for radio base stations (RBSs), 31 spread-spectrum, 23036 testing, 524 transmission theory, 2036 See also Microwave systems Polarization mode dispersion (PMD), 533 Postinstallation activities, 44345 Power, 399404 ac, 399 backup, 401 batteries, 4013 dc, 400 density, 48990 levels, 48990 loss, 401 solar energy, 4034 See also Transmission equipment Precise Positioning System (PPS), 244 Preliminary equipment list, 291 Preliminary network layout, 290 Preliminary network report, 28990 Private branch exchanges (PBXs), 51
603
Procurement management, 431 Project management, 43032, 43646 change orders, 443 defined, 430 definitions, 43638 leased-lines tracking process, 44143 organizational issues, 43840 postinstallation/optimization activities, 44345 processes, 431 project stages, 44041 S-curve, 438 services, 437 successful, 43637 tools, 44546 in wireless networks, 43646 Protection ATM, 163 DWDM, 200201 microwave systems, 22122 SDH transmission-network, 28687 SONET/SDH, 173 Public Switched Telephone Network (PSTN), 7, 49 signaling, 49 timing, 31516 QoS-based differential services, 160 Quadrature-amplitude modulation (QAM), 236 Quadrature phase-shift keying (QPSK), 37 Quality control, 431 Quality of Service (QoS) in ATM networks, 8789 CBR, 156 characteristics of classic telephony networks, 13334 classes, 146 concept, 15861 concerns addressed by, 158 IP-based wireless networks, 100 large-scale telephony solutions, 131 mechanisms, 159, 160 mobile (M-QoS), 16061 toll, 27273 as underlying guide, 159 Queuing delay, 15960
604
Transmission Systems Design Handbook for Wireless Networks
Radiation patterns, 49091 Radio access network (RAN) defined, 14243 infrastructure, 133 Radio base stations (RBSs), 31 clusters, 175 traffic load, 167 trunk calculation, 167 Radio frequency coordination, 22526 Radio-frequency field measurements, 48595 electromagnetic field sources, 49192 emission sources, 48687 health/safety issues, 48586 induced and contact currents, 49293 instrumentation, 49394 near-field/far-field regions, 48889 power levels and power density, 48990 procedures, 49495 radiation patterns and polarization, 49091 Radio-frequency radiation (RFR), 485 Radio network controller (RNC), 141 dimensioning, 32743 distributed topology, 331 interfaces, 32932 locations, 331 traffic calculation, 33443 Radio transmission technology (RTT), 17 Rain, 21719, 230 attenuation, 21719 effects on microwave propagation, 21719 fade margins and, 218 ITU-R model, 217 outages, 219 Reference ellipsoids, 24243 Reliability equipment, 40710 microwave-link design, 21516 Remote BSCs, 17778 Remote-frame alarm bits, 125 Request for information (RFI), 414 Request for quotes (RFQs), 41519 defined, 415 issues, 416 main activities, 417
process, 41719 proposals, 415 structure, 416 technical requirements, 419 terminology, 416 transmission network, 41718 Responsibility matrix, 420 RF cell-site compliance, 45354 RF design, 26870 defined, 268 geographic characteristics, 268 for new wireless networks, 268 result, 269 See also Transmission-network design Ring topology, 17273 in microwave design, 22728 protection, 28889 Risk management, 431 RNC dimensioning, 32743 Rural statistical areas (RSAs), 261 Satellite networks, 2628 fixed systems, 2627 geostationary, 26 GPS, 244 low-Earth orbit (LEO), 27 mobile systems, 2728 SC connectors, 505 Scheduling tool, 445 Scope management, 431 S-curve, 438 Security management, 515 Synchronous Optical Network (SONET), 191 Self-support towers, 47679 defined, 476 materials, 476 material selection, 479 surface requirements, 478 See also Towers Service level agreements (SLAs), 510 Services, 43036 delivery, 427 due diligence, 43536 engineering, 430 network maintenance, 436
Index order processing, 41325 outsourcing, 43234 project management, 43032, 437 supplier negotiations, 425 Short-message service (SMS), 11, 13 Sidehaul, 143 Signaling System 7 (SS7), 5456 data packets, 56 defined, 49, 55 functions, 55 hardware/software functions, 57 switches, 56 Signal propagation delay. See Latency Simplex power design, 11214 Simulation tools, 16568 Sites acquisition process, 44849 acquisition team, 447 cell, selection, 45152 information, 446 microwave, surveys, 45661 microwave repeater, selection, 45253 NOC, selection, 449 selection, 44654 switch, 44951 zoning issues, 447 SI units, 55051 additional units, 551 base units, 550 defined, 550 derived units, 551 See also Units of measurement SMA connectors, 505 SNMP agent, 252 defined, 251, 511 direct connection, 25253 implementation methods, 251 microwave network management with, 24954 network components, 25152 network illustration, 251 NMS connection methods, 25354 overview, 25152 traps, 252 Soil resistivity, 395 Solar energy, 4034
605
PV panels, 404 solar thermal technologies, 404 SONET/SDH, 80 ATM in, 16263 basis, 70 defined, 49, 69, 190 evolution to, 70 fiber-optic equipment, 36368 IP over, 150 microwave radios, 164, 35152 protection and switching techniques, 173 terminals, 69 transmission products, 36365 use of, 191 See also Synchronous digital hierarchy (SDH); Synchronous Optical Network (SONET) Space-diversity (SD) systems, 21011 Spare bits, 12427 international, 12425 national bits, 12527 remote-frame alarm bits, 125 types of, 124 See also E1 system Specific Absorption Rate (SAR), 49293 Spectrum analyzers, 49394 auctions, 26667 clearing, 267 as valuable resource, 280 Split microwave radio configuration, 35358 block diagram, 355 equipment main characteristics, 35457 general description, 35354 illustrated, 354 mechanical characteristics, 357 power supply, 357 transmission interfaces, 35758 See also Microwave radio equipment Spread-spectrum systems, 23036 defined, 230 direct-sequence (DSSS), 232, 233 frequency-hopping (FHSS), 23233 illustrated, 234 modulation techniques, 232
606
Transmission Systems Design Handbook for Wireless Networks
Spread-spectrum systems (continued) special rules, 23435 See also Microwave systems Standard-configuration microwave system, 35253 Standard Positioning Service (SPS), 24445 Statistical multiplexing, 7273, 156 bandwidth efficiency and, 154 of data traffic, 152 use of, 154 See also Multiplexing ST-compatible connectors, 5045 Strategic planning, 25962 Stratum, 3048 clock requirements, 306 Stratum 1, 304, 307 Stratum 2, 304, 307 Stratum 3, 304, 307 Stratum 3E, 307 Stratum 4, 304, 3078 Stratum 4E, 3078 Streaming class, 147 Structured circuit emulation, 169 Subrate multiplexing, 7374 Supervisory control and data acquisition (SCADA), 348 Switched leased service, 29798 characteristics, 297 circuit-switched network, 29798 defined, 297 packet-switched network, 298 See also Leased lines Switched multimegabit digital service (SMDS), 74 Switch sites, 44951 Synchronization, 30217 CDMA and, 31617 cell-site, 31617 cell-site timing, 31617 historical overview, 3024 interoffice distribution, 31012 intraoffice distribution, 31213 introduction, 3024 network design rules, 31112 networks, 303 in PCS networks, 31416
SONET network timing, 31314 strata, 3048 timing planning rules, 30810 See also Transmission-network design Synchronous digital hierarchy (SDH) defined, 190 flexibility, 195 as international standard, 195 management layer, 196 microwave radio, 35152 terminal equipment, 294 transmission network protection, 28687 transmission systems, 286 See also SONET/SDH Synchronous Optical Network (SONET) architecture, 19194 availability requirements, 195 basic frame format, 19394 benefits, 191 defined, 190 disaster recovery, 191 functionality, 193 high-speed transmission, 191 integration, 191 network timing, 31314 performance, 191 security, 191 stability/robustness, 191 transmission layers, 19193 See also SONET/SDH Synchronous payload envelope (SPE), 81 Synchronous Transmission Module (STM) defined, 79 Level 1 (STM-1), 70 Synchronous Transmission Signal, Level 1 (STS-1), 70 channels, 193 frame format, 194 signals, 193, 194 T1
cables, 391 circuits, 295 impairments, 52223 pulse mark, 389 testing, 52324
Index T1 system, 10218 ABAM cable, 109 acceptance test procedure (ATP), 116 AMI, 1023, 104 B8ZS, 102, 104 BERs, 1056 BERT, 116 channels, 59 channel service units (CSUs), 1034, 112 connections, 38 data error rates, 11415 defined, 59 documentation, 116 dual-cable operation, 107 elements, 63 engineering, 116 facts, 65 fractional (F-T1), 118 framing, 118 installation, 116 in Japan (J1), 118 lightning, 11012 line codes, 118 network interface units (NIUs), 112 North American digital hierarchy, 65 order wire, 110 overall system length, 1067 problem classification, 11617 pulse transmission, 105 repeatered lines, 1079 service providers, 69 shortspan power feed, 113 signals in, 1024 simplex power design, 11214 single-cable operation, 107, 108 smart jacks, 112 span line, 113, 114 switch options, 118 troubleshooting, 11617 tutorial, 10218 typical cable layout, 108 use of, 62 voltage/temperature factors, 115 T3, 29596 Telecommunications Act of 1996, 58
607
Telecommunications management network (TMN) model, 511 Telecommunications provider negotiations, 42225 Terminal multiplexers, 366 Testing. See Transmission-network testing Third Generation Partnership Project (3GPP), 18 Tight-buffered cables, 499 Time-assignment speech-interpolation (TASI), 7172 defined, 71 multiplexers, 71, 72 trunks, 72 Time-division multiple access (TDMA), 6 air interface, 10 defined, 9 specification, 9 use environments, 10 Time-division multiplexing (TDM), 59 asynchronous (ATDM), 71 data traffic, 152 replacing, with ATM, 14854 switches, 74 Time management, 431 Timing BITS, 31213 distribution, 310 GPS, 316 interoffice distribution, 31012 intraoffice distribution, 31213 planning rules, 30810 private network, 315 PSTN, 31516 SONET network, 31314 See also Synchronization Toll QoS, 27273 Topology planning, 27071 Total Access Communications Systems (TACS), 15 Towers electric transmission, 345 fall-prevention systems, 48183 guyed, 47679 hydro, 466 lights, 471 material selection, 479
608
Transmission Systems Design Handbook for Wireless Networks
Towers (continued) microwave/radar, 466 minimum strength requirements, 48485 procurement, 48385 self-support, 47679 strength requirements, 474 See also Microwave antenna mounting structures Traffic calculations, 33443 example, 34243 guidelines, 33441 Iub traffic load, 33739 Iur traffic load, 34041 Iu traffic load, 33537 See also Radio network controller (RNC) Traffic classes, 14548 3G, 14548 background, 148 conversational, 14647 defined, 145 interactive, 14748 streaming, 147 Traffic engineering, 5154 busy-hour traffic, 51 capacity per circuit, 53 Erlang B, 51, 52 Erlang C, 5253 example, 5354 GoS, 52 Traffic grooming, 31718 Traffic modeling, 16568 Transmission and distribution (T&D) system, 346 Transmission ATP form, 52930 Transmission Control Protocol (TCP), 100 Transmission equipment, 351412 cabling, 38492 delivery, 426 digital microwave radio, 35163 environmental specifications, 41011 factory acceptance testing, 407 fiber-optic, 36370 first article test (FAT), 4067 GPS antennas, 405 grounding, 39299
Network Equipment Building Standard (NEBS), 41112 order processing, 41325 power/battery backup, 399404 procurement, 2 quality assurance, 4056 quality/reliability issues, 40512 supplier negotiations, 425 wireline, 37084 Transmission lines, 361 Transmission media planning, 27071 types, 3 Transmission-network architecture, 17079 Transmission-network deployment, 413519 equipment/services ordering, 41325 fiber-optic cable installation, 495509 microwave deployment, 45485 operations and maintenance, 50919 project management, 43646 radio-frequency fields measurement, 48595 regulatory issues, 42530 services, 43036 site selection, 44654 Transmission-network design, 257349 alternative solutions, 34449 clearing spectrum, 267 complex, 32527 customer requirements analysis, 26264 dark fiber/dark copper, 344 examples, 32427 final equipment list, 291 initial questions, 283 interconnection, 26566 leased lines and, 292302 list of approved sites, 291 media and topology planning, 27071 microwave relocation, 267 mobile positioning, 27172 MW LOS survey report, 291 MW path and frequency-planning report, 291 network life cycle and, 28183 network performance, 27376
Index optical laser communications, 34849 optimization, 31724 overview, 25759 partnership with utilities, 34448 PDH microwave example, 32425 PDH systems, 286 preliminary equipment list, 291 preliminary network layout, 290 preliminary network report, 28990 principles, 25983 process, 28385 process illustration, 284 regulatory issues, 27881 requirements, 28385 RF design, 26869 ring protection, 28889 RNC dimensioning, 32743 sales and marketing, 27678 SDH systems, 286 spectrum auctions, 26667 strategic planning and, 25962 synchronization, 30217 TND deliverables, 28991 toll QoS, 27273 transmission-network plan, 291 Transmission-network plan, 291 Transmission-network principles, 47127 ATM, 7991 digital technology, 5966 multiplexing, 7079 plesiochronous vs. synchronous digital hierarchy, 6670 voice over IP (VoIP), 91102 wireline side, 4758 Transmission-network testing, 52147 ATM network, 53843 BERT, 52226 definitions, 52122 DS1/DS3 performance objectives, 52829 DS1 procedure, 52930 end-to-end, 522 fiber-optic cable, 53037 leased facilities, 52627 long-term, 522 loopback, 522, 525 microwave systems, 52728
609
out-of-service, 52426 packet-network, 53747 T1/E1, 52324 voice traffic, 53738 VoIP networks, 54347 Tripods, 47980 with MW antenna, 481 nonpenetrating, 480 penetrating, 479 Trunk calculation, 167 Ufer grounds, 397 Unfaded receive-signal level (RSL), 204 U-NII band effective use of, 37 FCC rules for operation in, 36 point-to-point microwave path performance in, 38 subbands, 36 wireless local access in, 3539 Uninterrupted power supply (UPS), 399 Units of measurement common, 552 conversion, 54952 International System (SI), 55051 Universal Mobile Telephone System (UMTS), 15, 141 ATM and, 269 defined, 141 deployment, 141 origination, 142 Terrestrial Radio Access Network (UTRAN), 32829 Unstructured circuit emulation, 16970 User-plane protocol stack, 151 User traffic modeling, 33233 bandwidth formula, 333 defined, 332 example, 342 parameters, 33233 Utility partnerships, 34448 distribution poles, 34546 electric transmission towers, 345 power line carrier, 348 technical issues, 34648 UTRAN networks ATM in, 329
610
Transmission Systems Design Handbook for Wireless Networks
UTRAN networks (continued) dimensioning, 333 interfaces, 330 traffic classes, 32829 See also Universal Mobile Telephone System (UMTS) Valve-regulated lead acid (VRLA) batteries, 400, 4013 benefits, 4012 defined, 401 installation/maintenance rules, 4023 Variable packet fill delay, 156 VBR-real time (VBR-rt), 156 Virtual home environment (VHE), 136 Virtual private networks (VPNs), 510 Voice compression, 31821 coding overview, 31820 in wireless networks, 32021 Voice optimization, 156 Voice over frame relay (VoFR), 91 Voice over IP (VoIP), 91102 audio-quality tests, 545 billing systems, 547 H.323 and, 9294 illustrated, 93 IP-based wireless networks, 99102 jitter issue, 9597 latency issue, 9597 market, 91 MPLS and, 9798 network components, 544 quality tests, 54347 services, 92 VoX networks, 38283 Water pipes, 397 Wide-area networks (WANs), 49 Wideband CDMA, 16, 17 Wireless Application Protocol (WAP), 15 Wireless evolution, 135 Wireless local loop (WLL) applications, 7 competition and, 7 defined, 7 requirement of, 7 Wireless-network architecture, 12979 2G, 12931
3G, 13148 Wireless networks, 544 1G, 15 2G, 15, 12931, 16870 3G, 3, 1519, 144, 16870, 32743 architecture, 259 cell-site timing in, 31617 complexity, 258 echo cancellers in, 38082 existing technologies, 614 expense, 258 fixed microwave, 2844 historical background, 56 IP-based, 99102 leased lines, 292302 life cycle, 28183 planning/design principles, 25983 project management in, 43646 ring protection in, 28889 satellite, 2628 technology evolution, 1526 transmission (backhaul), 6162 voice compression in, 32021 Wireless service providers (WSPs), 31 Wireless technologies, 614 analog cellular systems, 79 CDMA, 1314 evolution of, 1526 FDMA, 9 future directions, 2426 GSM, 1013 PDC, 10 TDMA, 910 WLL, 67 Wireline equipment, 37084 CSU/DSU, 37576 DACS, 37074 DSL/ADSL, 37679 echo cancellers, 37984 See also Transmission equipment Wireline side, 4758 AIN, 5658 PSTN interconnect, 4750 SS7, 5456 step-by-step system, 4950 Telecommunications Act of 1996 and, 58
Index traffic engineering, 5154 See also Wireless networks
Word frames, 123 xDSL. See Digital subscriber line (DSL)
611