Handbook of Research on Mobility and Computing: Evolving Technologies and Ubiquitous Impacts

Handbook of Research on Mobility and Computing: Evolving Technologies and Ubiquitous Impacts Maria Manuela Cruz-Cunha Polytechnic Institute of Cávado e Ave, Portugal Fernando Moreira Universidade Portucalense, Portugal

Volume I

INFORMATION SCIENCE REFERENCE Hershey • New York

Senior Editorial Director: Kristin Klinger Director of Book Publications: Julia Mosemann Editorial Director: Lindsay Johnston Acquisitions Editor: Erika Carter Development Editor: Michael Killian Typesetters: Michael Brehm, Casey Conapitski, Keith Glazewski, Milan Vrarich Jr. & Deanna Zombro Production Coordinator: Jamie Snavely Cover Design: Nick Newcomer Published in the United States of America by Information Science Reference (an imprint of IGI Global) 701 E. Chocolate Avenue Hershey PA 17033 Tel: 717-533-8845 Fax: 717-533-8661 E-mail: [email protected] Web site: http://www.igi-global.com Copyright © 2011 by IGI Global. All rights reserved. No part of this publication may be reproduced, stored or distributed in any form or by any means, electronic or mechanical, including photocopying, without written permission from the publisher. Product or company names used in this set are for identification purposes only. Inclusion of the names of the products or companies does not indicate a claim of ownership by IGI Global of the trademark or registered trademark. Library of Congress Cataloging-in-Publication Data Handbook of research on mobility and computing : evolving technologies and ubiquitous impacts / Maria Manuela Cruz-Cunha and Fernando Moreira, editors. p. cm. Includes bibliographical references and index. ISBN 978-1-60960-042-6 (hbk.) -- ISBN 978-1-60960-043-3 (ebook) 1. Mobile computing. 2. Wireless communication systems. I. Cruz-Cunha, Maria Manuela, 1964- II. Moreira, Fernando, 1969 Aug. 16QA76.59.H35 2011 004.165--dc22 2010036723 British Cataloguing in Publication Data A Cataloguing in Publication record for this book is available from the British Library. All work contributed to this book is new, previously-unpublished material. The views expressed in this book are those of the authors, but not necessarily of the publisher.

Editorial Advisory Board Goran D. Putnik, University of Minho, Portugal João Varajão, University of Trás-os-Montes e Alto Douro, Portugal Joaquim Arnaldo Martins, Universidade de Aveiro, Portugal Nuno Lopes, Polytechnic Institute of Cávado e Ave, Portugal Patrícia Gonçalves, Polytechnic Institute of Cávado e Ave, Portugal Paulo Ferreira, INESC Porto, Portugal Vítor Carvalho, Polytechnic Institute of Cávado e Ave, Portugal

List of Reviewers Abdellah Touhafi, Erasmushogheschool, Belgium Abid Yahya, Universiti Sains Malaysia, Malaysia Abilio Azenha, ISR – Porto, Portugal Adriano Carvalho, ISR-Porto, Portugal Agostino Poggi, University of Parma, Italy An Braken, Erasmushogeschool, Belgium Ana Isabel González-Tablas, Univ. Carlos III of Madrid, Spain Ana María Fermoso García, Universidad Pontificia de Salamanca, Spain Andrea Zanda, Universidad Politecnica Madrid, Spain Andreas Ahrens, Hochschule Wismar - University of Technology, Germany Anselmo Cardoso de Paiva, University of the Maranhão, Brazil Antonio Foncubierta Rodriguez, University of Seville, Spain Aparecido Fabiano de Carvalho, University of Limerick, Republic of Ireland Arturo Molina, Tecnologico de Monterrey, Mexico Aysegul Cayci, Sabanci University, Turkey Bahram Lotfisadigh, Middle East Technical University, USA Bart De Decker, K.U. Leuven, Belgium Bruno Kimura, University of São Paulo, Brazil C. Brad Crisp, Abilene Christian University, USA Carlo Bertolli, University of Pisa, Italy Carlos Ferras, University of Santiago, Spain

Carlos Quental, ESTV / IPV, Portugal Carmen Ruthenbeck, University of Bremen, Germany Castor Sánchez Chao, University of Vigo, Spain Célia Menezes, Escola Sec. Inf. D. Henrique, Portugal César Benavente-Peces, EUIT Telecomunicacion - Universidad Politécnica de Madrid, Spain Chad Lin, Curtin University of Technology, Australia Chongming Zhang, Shanghai Normal University, China Cláudio de Souza Baptista, University of Campina Grande, Brazil Crescenzio Gallo, University of Foggia, Italy Cristiano Costa, Universidade do Vale do Rio dos Sinos (Unisinos), Brazil Cristina Rodriguez-Sanchez, Universidad Rey Juan Carlos, Spain Daniel Câmara, EURECOM, France Daniele Buono, University of Pisa, Italy David Santos, Universidad San Pablo-CEU, Spain Dennis Viehland, Massey University, New Zeland Diana Bri, Polytechnic University of Valencia, Spain Edson Moreira, University of Sao Paulo, Brazil Elena Nazzi, IT-University of Copenhagen, Denmark Emanuel Angelescu, RWTH, Germany Emanuel Correia, University of Trás-os-Montes e Alto Douro, Portugal Emanuel Soares Peres Correia, Universidade de Trás-os-Montes e Alto Douro, Portugal Encarnación Beato Gutierrez, Universidad Pontificia de Salamanca, Spain Ernestina Menasalvas, Universitad Politecnica Madrid, Spain Ernesto Morales Kluge, BIBA, Gremany Fernando Lopez-Colino, UAM, Spain Fernando Moreira, Universidade Portucalense Infante D. Henrique, Portugal Florian Harjes, BIBA, Germany Francisco Rodríguez-Díaz, University of Granada, Spain Francisco Amaral, University of the Azores, Portugal Gabriele Mencagli, University of Pisa, Italy Gianluca Cornetta, Universidad San Pablo-CEU, Spain Gonzalo Aranda-Corral, University of Sevilla, Spain Greg Wilson, Virginia Tech, USA Gustavo Lermen, Universidade do Vale do Rio dos Sinos (Unisinos), Brasil Hakki Unver, TOBB-ETU, USA Hamilton Turner, Vanderbilt University, USA Hassan Karimi, University of Pittsburgh, USA Heather Katz Hugo Coll, Polytechnic University of Valencia, Spain Hugo Feitosa de Figueirêdo, University of Campina Grande, Brazil Ismael Bouassida Rodriguez, LAAS-CNRS, France Isidro Vila Verde, Faculdade de Engenharia da Universidade do Porto, Portugal J. Stephanie Collins, Southern NH University, USA Jaime Lloret, Polytechnic University of Vale, Spain

Jairo Gutierrez, University of Auckland, New Zealand Javier Pereira Loureiro, University of A Coruna, Spain Jérôme Lacouture, LAAS-CNRS, France Jiaxiang Gan, University of Auckland, New Zealand João Barreto, Inesc-ID/UTL, Portugal João Bartolo Gomes, Universitad Politecnica Madrid, Spain João Carmo, University of Minho, Portugal João Eduardo Quintela Alves de Sousa Varajão, Universidade de Trás-os-Montes e Alto Douro, Portugal João Rosa, Universidade do Vale do Rio dos Sinos (Unisinos), Brazil Joaquim Arnaldo Martins, Universidade de Aveiro, Portugal Joaquín Borrego-Díaz, University of Sevilla, Spain Joaquim Sousa Pinto, Universidade de Aveiro, Portugal Jorge Ferreira, E-Geo FCSH, Portugal Jorge Luís Barbosa, Universidade do Vale do Rio dos Sinos (Unisinos), Brazil José Colas, UAM, Spain José Delgado, Instituto Superior Técnico, Portugal José Eustáquio Rangel de Queiroz, University of Campina Grande, Brazil José María De Fuentes, University Carlos III of Madrid, Spain José Maria Monguet-Fierro, UPC, Spain Juan Álvaro Muñoz Naranjo, Universidad de Almería, Spain Juan Antonio López Ramos, Universidad de Almería, Spain Juan Carlos González Moreno, University of Vigo, Spain Juan Hernandez-Tamames, Universidad Rey Juan Carlos, Spain Júlio cesar Melo, UFRN, Brazil Justo Peralta López, Universidad de Almería, Spain Kathryn Hayes, UWS, Australia Kelvin Bwalya, University of Botswana, Botswana Koong Lin, National University of Tainan, Taiwan Krassie Petrova, Auckland University of Technology, New Zeland Kris Steenhaut, Vrije Universiteit Brussel, Belgium Laerte Franco, Unisinos, Brazil Laura Nieto Riveiro, University of A Coruna, Spain Nuno Lopes, Polytechnic Institute of Cávado and Ave, Portugal Luca Vassena, University of Udine, Italy Luciana Arantes, LIP6/INRIA, France Luís Villaseñor, CICESE Research Centre, Mexico Luís Gouveia, Univ.Fernando Pessoa, Portugal Luís Peneda, ISR-Porto, Portugal Luís Serrano, Polithecnic Institute Leiria, Portugal Luís Veiga, INESC ID / Technical University of Lisbon, Portugal Luiz Marcos Gonçalves, UFRN, Brazil Mahsa Ghafourian, University of Pittsburgh, USA Mamun Abu-Tair, ECIT, UK Marco Vanneschi, University of Pisa, Italy

Marcus Bjelkemyr, The Royal Institute of Technology, Sweden Maria Borges Tiago, University of the Azores, Portugal Maria João Ferreira, Universidade Portucalense, Portugal Maria Manuela Cruz-Cunha, Polytechnic Institute of Cávado e Ave, Portugal Martin Stigmar, Växjö University, Sweden Mauro Onori, The Royal Institute of Technology, Sweden Maximino Esteves Correia Bessa, Universidade de Trás-os-Montes e Alto Douro, Portugal Michael Decker, University of Karlsruhe, Germany Michael Williams, Pepperdine University, USA Michelangelo De Bonis, University of Foggia, Italy Michele Perilli, University of Foggia, Italy Michele Tomaiuolo, University of Parmaz, Italy Miguel Angel Sánchez Vidales, Universidad Pontificia de Salamanca, Spain Miguel Garcia, Polytechnic University of Valencia, Spain Miguel Monteiro, Universidade do Porto, Portugal Milos Stojmenovic, University of Ottawa, Canada Montserrat Mateos Sánchez, Universidad Pontificia de Sala, Spain Nabeel Ahmad, IBM, USA Naoe Tatara, Norwegian Centre for Integrated Care and Telemedicine, Norway Natalia Padilla Zea, University of Granada, Spain Nicholas Pagden, Växjö University, Sweden Nikolaos Frangiadaki, University of Maryland, USA Nuno André Osório Liberato, Universidade de Trás-os-Montes e Alto Douro, Portugal Nuno Flores, Universidade do Porto, Portugal Nuno Lau, Universidade de Aveiro, Portugal Nuno Lopes, Polytechnic Institute of Cávado e Ave, Portugal Nuno Veiga, IPLeiria, Portugal Patrícia Gonçalves, Polytechnic Institute of Cávado and Ave, Portugal Paula Vicente, ISCTE-IUL, Portugal Paulo Ferreira, INESC ID / IST, Portugal Pedro Campos, University of Madeira, Portugal Pethuru Raj Chelliah, Wipro Technologies, USA Pierre Kirisci, BIBA, Germany Pierre Sens, LIP6/INRIA, France Qixiang Pang, General Dynamics Canada, Canada Raúl Aquino, University of Colima, Mexico Riikka Vuokko, Åbo Akademi University, Finland Roberto Berjón Gallinas, Universidad Pontificia de Salamanca, Spain Roberto Yokoyama, University of Sao Paulo, Brazil Rubén Romero González, University of Vigo, Portugal Rui Rijo, IPLeiria, Portugal Rummenigge Dantas, UFRN, Brasil Sandra Sendra, Polytechnic University of Valencia, Spain Santiago Eibe, Universitad Politecnica Madrid, Spain

Sílvio Bernardes, IPLeiria, Portugal Sohail Anwar, Penn State University, USA Thierry Desprats, IRIT, France Tom Chan, Southern NH University, USA Victor Gonzalez, University of Manchester, UK Víctor Rangel, National Autonomous University of Mexico, Mexico Vincenzo Pallotta, University of Geneva, Switzerland Violeta Chirino-Barceló, Tecnológico de Monterrey, Mexico Xinyu Lu, General Dynamics, Canada Yelena Zascerinska, University of Latvia, Republic of Latvia

List of Contributors

Abreu, Jorge / University of Aveiro, Portugal.................................................................................... 414 Ahmad, Nabeel / IBM Center for Advanced Learning, USA.............................................................. 871 Ahmad, R. Badlishah / Universiti Malaysia Perlis, Malaysia.......................................................... 157 Ahrens, Andreas / University of Technology, Business and Design, Germany.......................... 100, 760 Almeida, Margarida / University of Aveiro, Portugal....................................................................... 414 Almeida, Pedro / University of Aveiro, Portugal................................................................................ 414 Amaral, Francisco / University of the Azores, Portugal.................................................................... 327 Angelescu, Emanuel / Bremer Institut für Produktion und Logistik GmbH, Germany.................... 1226 Antunes, Maria / University of Aveiro, Portugal............................................................................... 414 Anwar, Sohail / Penn State University, USA...................................................................................... 237 Aquino, Raúl / University of Colima, México.................................................................................... 595 Aranda-Corral, Gonzalo / University of Sevilla, Spain.................................................................... 667 Arantes, Luciana / University Paris 6, France................................................................................ 1039 Arcangeli, Jean-Paul / Paul Sabatier University, France................................................................ 1056 Årsand, Eirik / Norwegian Centre for Integrated Care and Telemedicine, Norway.......................... 136 Astray, Loxo Lueiro / University of Vigo, Spain................................................................................ 445 Augustin, Iara / Federal University of Santa Maria, Brazil ........................................................... 1077 Ayu, Media / International Islamic University Malaysia, Malaysia.......................................... 539, 558 Azevedo, Renato / Federal University of Santa Maria, Brazil........................................................ 1077 Barbosa, Jorge / Universidade do Vale do Rio dos Sinos (Unisinos), Brazil..................................... 634 Barreto, João / INESC-ID/Technical University Lisbon, Portugal.......................................................1132 Bártolo Gomes, João / Universitad Politecnica, Spain...................................................................... 576 Bassus, Olaf / University of Technology, Business and Design, Germany......................................... 760 Benavente-Peces, César / Universidad Politécnica de Madrid, Spain.............................................. 100 Bernardes, Silvio / IPLeiria, Portugal............................................................................................... 487 Bertolli, Carlo / University of Pisa, Italy............................................................................................ 617 Bjelkemyr, Marcus / The Royal Institute of Technology, Sweden...................................................... 821 Bonnet, Christian / EURECOM Sophia Antipolis, France........................................................ 267, 356 Borrego-Díaz, Joaquín / University of Sevilla, Spain........................................................................ 667 Borromeo, Susana / Universidad Rey Juan Carlos, Spain................................................................ 472 Braeken, An / Erasmushogeschool Brussel, Belgium......................................................................... 930 Bri, Diana / Polytechnic University of Valencia, Spain.................................................................... 1155 Buono, Daniele / University of Pisa, Italy.......................................................................................... 617

Burlamaqui, Aquiles / Universidade Federal do Rio Grande do Norte, Brasil................................ 397 Busch, Cristina Diaz / University of A Coruna, Spain..................................................................... 1169 Cabrera, Marcelino / University of Granada, Spain......................................................................... 368 Câmara, Daniel / EURECOM Sophia Antipolis, France........................................................... 267, 356 Campos, Pedro / University of Madeira, Portugal............................................................................ 793 Canovas, Alejandro / Polytechnic University of Valencia, Spain...................................................... 426 Carmo, João Paulo / University of Minho, Portugal....................................................................... 1021 Cayci, Aysegul / Sabanci University, Turkey...................................................................................... 576 Cerquides Bueno, José Ramón / University of Seville, Spain......................................................... 1199 Chapman, Ross / Deakin University Melbourne, Australia................................................................. 65 Chassot, Christophe / Université de Toulouse, France................................................................... 1056 Chawla, Sheenu / SUSH Global Solutions, New Zealand.................................................................. 314 Chirino-Barceló, Violeta / Tecnologico de Monterrey Mexico, Mexico............................................ 774 Ciolfi, Luigina / University of Limerick, Republic of Ireland............................................................. 381 Colás, Jose / Universidad Autónoma de Madrid, Spain....................................................................... 83 Coll, Hugo / Polytechnic University of Valencia, Spain................................................................... 1155 Cornetta, Gianluca / Universidad San Pablo-CEU, Spain...................................................................... Cornetta, Gianluca / Universidad San Pablo-CEU, Spain........................................................ 930, 994 Correia, José Higino / University of Minho, Portugal..................................................................... 1021 Costa, Cristiano / Universidade do Vale do Rio dos Sinos (Unisinos), Brazil................................... 634 Costa, José / University Coimbra, Portugal....................................................................................... 703 Crisp, C. Brad / Abilene Christian University, USA........................................................................ 1213 Cunha, Maria / Pontificia Universidade Católica do PR, Brazil..................................................... 1091 da Silva, Tiago Eduardo / University of Campina Grande, Brazil.................................................. 1104 Dantas, Rummenigge / Universidade Federal do Rio Grande do Norte, Brasil............................... 397 de Bonis, Michelangelo / University of Foggia, Italy.......................................................................... 31 de Carvalho, Aparecido Fabiano Pinatti / University of Limerick, Republic of Ireland................. 381 de Decker, Bart / Katholieke Universiteit Leuven, Belgium............................................................. 1246 de Figueirêdo, Hugo Feitosa / University of Campina Grande, Brazil........................................... 1104 de Fuentes, José María / Carlos III University of Madrid, Spain..................................................... 894 de Paiva, Anselmo Cardoso / University of Maranhão, Brazil....................................................... 1104 de Queiroz, José Eustáquio Rangel / University of Campina Grande, Brazil................................ 1104 de Sousa Varajão, João Eduardo Quintela Alves / UTAD, Portugal.............................................. 881 de Souza Baptista, Cláudio / University of Campina Grande, Brazil............................................. 1104 Decker, Michael / Karlsruhe Institute of Technology (KIT), Germany.............................................. 912 Delgado, Jose / Instituto Superior Técnico, Portugal......................................................................... 853 Demetriou, Antonis / University of Manchester, UK....................................................................... 1262 Desprats, Thierry / Paul Sabatier University, France..................................................................... 1056 Dias, Ricardo / Universidade Federal do Rio Grande do Norte, Brasil............................................ 397 Dillenburg, Fabiane / Universidade Federal do Rio Grande do Sul, Brazil..................................... 634 Dougherty, Brian / Vanderbilt University, USA................................................................................. 502 Drira, Khalil / Université de Toulouse, France................................................................................ 1056 Edo, Miguel / Polytechnic University of Valencia, Spain................................................................... 426 Edwards, Artur / University of Colima, México................................................................................ 595

Eibe, Santiago / Universitad Politecnica, Spain................................................................................. 576 Esteves Correia Bessa, Maximino / UTAD, Portugal....................................................................... 881 Fermoso García, Ana María / Universidad Pontificia de Salamanca, Spain................................... 216 Ferrás, Carlos / University of Santiago Compostela, Spain............................................................. 1182 Ferreira, Paulo / INESC ID / Technical University of Lisbon, Portugal...................................719, 1132 Filali, Fethi / EURECOM, Qatar................................................................................................. 267,356 Foncubierta Rodriguez, Antonio / University of Seville, Spain...................................................... 1199 Frangiadakis, Nikolaos / University of Maryland, USA.................................................................... 356 Freitas, Leandro / Federal University of Santa Maria, Brazil......................................................... 1077 Gallinas, Roberto Berjón / Universidad Pontificia de Salamanca, Spain........................................ 216 Gallo, Crescenzio / University of Foggia, Italy.................................................................................... 31 Gan, Jiaxiang / University of Auckland, New Zealand....................................................................... 837 Garcia, Miguel / Polytechnic University of Valencia, Spain...................................................... 426, 595 Garcia, Osvaldo / Pontificia Universidade Católica do PR, Brazil................................................. 1091 García, Yolanda / University of Santiago Compostela, Spain.......................................................... 1182 Garijo, Francisco / Université de Toulouse, France........................................................................ 1056 Garzão, Alex / Universidade do Vale do Rio dos Sinos (Unisinos), Brazil........................................ 634 Gassen, Jonas / Federal University of Santa Maria, Brazil............................................................. 1077 Ghafourian, Mahsa / University of Pittsburgh, USA................................................................. 203, 298 Ghani, Farid / Universiti Malaysia Perlis, Malaysia......................................................................... 157 Gonçalves, Luiz Marcos / Universidade Federal do Rio Grande do Norte, Brasil.......................... 397 González Moreno, Juan Carlos / University of Vigo, Spain............................................................. 445 González, Betania Groba / University of A Coruna, Spain............................................................. 1169 Gonzalez, Victor / University of Manchester, UK............................................................................ 1262 González-Tablas, Ana Isabel / Carlos III University of Madrid, Spain............................................ 894 Gouveia, Luis / University Fernando Pessoa, Portugal..................................................................... 974 Grahn, Kaj / Arcada University of Applied Sciences, Finland.......................................................... 952 Gray, Breda / University of Limerick, Republic of Ireland................................................................ 381 Greve, Fabíola / Federal University of Bahia, Brazil...................................................................... 1039 Gutierrez, Encarnación Beato / Universidad Pontificia de Salamanca, Spain................................ 216 Gutiérrez, Jairo A. / Universidad Tecnológica de Bolívar, Colombia............................................... 837 Harjes, Florian / Bremer Institut für Produktion und Logistik, Germany.......................................... 738 Hartvigsen, Gunnar / University of Tromsø, Norway........................................................................ 136 Hayes, Kathryn J. / University of Western Sydney, Australia.............................................................. 65 Hernandez-Tamames, Juan / Universidad Rey Juan Carlos, Spain................................................. 472 Huang, Yu-An / National Chi Nan University, Taiwan...................................................................... 175 Iapichino, Giuliana / EURECOM, France......................................................................................... 267 Jalleh, Geoffrey / Curtin University of Technology, Australia........................................................... 175 Joseph, Bwalya Kelvin / University of Johannesburg, South Africa................................................... 48 Karimi, Hassan / University of Pittsburgh, USA........................................................................ 203, 298 Karlsson, Jonny / Arcada University of Applied Sciences, Finland.................................................. 952 Kimura, Bruno / University of São Paulo, Brazil.............................................................................. 522 Kirisci, Pierre / Bremer Institut für Produktion und Logistik GmbH, Germany.............................. 1226 Koong Lin, Hao-Chiang / National University of Tainan, Taiwan.................................................... 175

Kurtz, Guilherme / Federal University of Santa Maria, Brazil...................................................... 1077 Lacouture, Jérôme / Université de Toulouse, France...................................................................... 1056 Lermen, Gustavo / Universidade do Vale do Rio dos Sinos (Unisinos), Brazil................................. 634 Librelotto, Giovani / Federal University of Santa Maria, Brazil.................................................... 1077 Lin, Chad / Curtin University of Technology, Australia..................................................................... 175 Lloret, Jaime / Polytechnic University of Valencia, Spain....................................................... 426, 1155 López Ramos, Juan Antonio / Universidad de Almería, Spain......................................................... 188 López, Justo Peralta / Universidad de Almería, Spain...................................................................... 188 López-Colino, Fernando / Universidad Autónoma de Madrid, Spain................................................. 83 Maffei, Antonio / The Royal Institute of Technology, Sweden............................................................ 821 Mantoro, Teddy / International Islamic University Malaysia, Malaysia................................... 539, 558 Martini, Ricardo / Federal University of Santa Maria, Brazil........................................................ 1077 McCrickard, Scott / Virginia Tech, USA............................................................................................... 459 Melo, Julio cesar / Universidade Federal do Rio Grande do Norte, Brasil....................................... 397 Menasalvas, Ernestina / Universitad Politecnica, Spain................................................................... 576 Mencagli, Gabriele / University of Pisa, Italy.................................................................................... 617 Menegon, Davide / University of Udine, Italy........................................................................................ 1 Menezes, Célia / Escola Sec. Inf. D. Henrique, Portugal................................................................... 250 Mentens, Nele / Katholieke Universiteit Leuven, Belgium................................................................. 930 Mizzaro, Stefano / University of Udine, Italy........................................................................................ 1 Molina, Arturo / Tecnologico de Monterrey Mexico, Mexico............................................................ 774 Monguet-Fierro, Jose María / Universidad Politecnica de Catalunya, Spain................................. 285 Morales Kluge, Ernesto / Bremer Institut für Produktion und Logistik GmbH, Germany.............. 1226 Moreira, Edson / University of São Paulo, Brazil............................................................................. 522 Moreira, Fernando / Universidade Portucalense, Portugal.............................................................. 250 Moreiras Lorenzo, Alberto / University of A Coruna, Spain.......................................................... 1169 Mourelos Sánchez, Iván / University of A Coruna, Spain............................................................... 1169 Naessens, Vincent / Katholieke Hogeschool Sint-Lieven, Belgium.................................................. 1246 Naranjo, Juan Álvaro Muñoz / Universidad de Almería, Spain....................................................... 188 Nazzi, Elena / University of Copenhagen, Denmark.............................................................................. 1 Ndlovu, Mandla / Botswana Accountancy College, Botswana............................................................ 48 Nieto Riveiro, Laura / University of A Coruna, Spain..................................................................... 1169 Noel, Victor / Paul Sabatier University, France............................................................................... 1056 Onori, Mauro / The Royal Institute of Technology, Sweden............................................................... 821 Osório Liberato, Nuno André / UTAD, Portugal.............................................................................. 881 Padilla Zea, Natalia / University of Granada, Spain......................................................................... 368 Pallotta, Vincenzo / Webster University, Switzerland........................................................................ 689 Parguiñas Portas, Cesar / University of Vigo, Spain......................................................................... 445 Pereira Loureiro, Javier / University of A Coruna, Spain............................................................... 1169 Pérez-Guerrero, María Luisa / Universidad Politecnica de Catalunya, Spain................................ 285 Perilli, Michele / University of Foggia, Italy........................................................................................ 31 Piotrowski, Jakub / Bremer Institut für Produktion und Logistik, Germany..................................... 738 Poggi, Agostino / Università degli Studi di Parma, Italy.................................................................... 343 Pose, Mariña / University of Santiago Compostela, Spain............................................................... 1182

Pousada García, Thais / University of A Coruna, Spain................................................................. 1169 Pulkkis, Göran / Arcada University of Applied Sciences, Finland.................................................... 952 Quental, Carlos / Polytechnic Institute of Viseu, Portugal................................................................ 974 Ramos, Fernando / University of Aveiro, Portugal............................................................................ 414 Rangel, Víctor / National Autonomous University of Mexico, México.............................................. 595 Reis, Elizabeth / UNIDE, ISCTE – Lisbon University Institute, Portugal......................................... 805 Rensleigh, Chris / University of Johannesburg, South Africa.............................................................. 48 Ribagorda, Arturo / Carlos III University of Madrid, Spain............................................................. 894 Rijo, Rui / IPLeiria, Portugal............................................................................................................. 487 Rodrigues, Vitor / Movensis, Portugal............................................................................................... 719 Rodriguez, Ismael Bouassida / Université de Toulouse, France..................................................... 1056 Rodríguez-Díaz, Francisco / University of Granada, Spain.............................................................. 368 Rodriguez-Sanchez, Cristina / Universidad Rey Juan Carlos, Spain............................................... 472 Romero González, Rubén / University of Vigo, Spain...................................................................... 445 Ruthenbeck, Carmen / Bremer Institut für Produktion und Logistik, Germany............................... 738 Sadigh, Bahram Lotfi / Middle East Technical University, Turkey................................................... 649 Saldaña-García, Carmina / Universidad de Barcelona, Spain......................................................... 285 Salleh, M. F. M. / Universiti Sains Malaysia, Malaysia..................................................................... 157 Sánchez Chao, Castor / University of Vigo, Spain............................................................................. 445 Sánchez Vidales, Miguel Angel / Universidad Pontificia de Salamanca, Spain............................... 216 Sánchez, Montserrat Mateos / Universidad Pontificia de Salamaca, Spain.................................... 216 Santos, David J. / Universidad San Pablo-CEU, Spain..................................................................... 994 Saraiva, Melissa / University of Aveiro, Portugal.............................................................................. 414 Schmidt, Doug / Vanderbilt University, USA...................................................................................... 502 Schneider, Claudio / Universidade Federal do Rio Grande do Norte, Brasil................................... 397 Scholz-Reiter, Bernd / Bremer Institut für Produktion und Logistik, Germany................................. 738 Sena, Hugo / Universidade Federal do Rio Grande do Norte, Brasil................................................ 397 Sendra, Sandra / Polytechnic University of Valencia, Spain........................................................... 1155 Sens, Pierre / University Paris 6, France......................................................................................... 1039 Serrano, Luís / Polithecnic Institute Leiria, Portugal........................................................................ 703 Sibilla, Michelle / Paul Sabatier University, France........................................................................ 1056 Sidek, Othman / Universiti Sains Malaysia, Malaysia...................................................................... 157 Silva, Lidia / University of Aveiro, Portugal....................................................................................... 414 Silva, Manuel / University Coimbra, Portugal................................................................................... 703 Simões, Diogo / Movensis, Portugal................................................................................................... 719 Soares Peres Correia, Emanuel / UTAD, Portugal........................................................................... 881 Steenhaut, Kris / Vrije Universiteit Brussel, Belgium........................................................................ 930 Stojmenovic, Milos / University of Ottawa, Canada............................................................................ 16 Tatara, Naoe / Norwegian Centre for Integrated Care and Telemedicine, Norway........................... 136 Teixeira, Jorge / University of Aveiro, Portugal................................................................................. 414 Tessier, Catherine / ONERA Centre de Toulouse – DCSD, France................................................. 1056 Thoben, Klaus-Dieter / Bremer Institut für Produktion und Logistik GmbH, Germany................. 1226 Tiago, Maria Borges / University of the Azores, Portugal................................................................. 327 Tomaiuolo, Michele / Università degli Studi di Parma, Italy............................................................ 343

Touhafi, Abdellah / Erasmushogheschool Brussel, Belgium...................................................... 930, 994 Turner, Hamilton / Vanderbilt University, USA................................................................................. 502 Ünver, Özgür / TOBB-University of Economics and Technology, Turkey.......................................... 649 Vanneschi, Marco / University of Pisa, Italy...................................................................................... 617 Vanni, Renata Maria / University of São Paulo, Brazil.................................................................... 522 Vassena, Luca / University of Udine, Italy............................................................................................. 1 Vázquez, José Manuel / Universidad San Pablo-CEU, Spain........................................................... 994 Veiga, Luis / INESC ID / Technical University of Lisbon, Portugal................................................... 719 Veiga, Nuno / IPLeiria, Portugal........................................................................................................ 487 Vicente, Paula / UNIDE, ISCTE – Lisbon University Institute, Portugal.......................................... 805 Viehland, Dennis / Massey University, New Zealand......................................................................... 314 Villaseñor, Luis / CICESE Research Centre, México......................................................................... 595 Vuokko, Riikka / Åbo Akademi University, Finland........................................................................ 1119 Wang, Zhonghai / Michigan Tech University, USA........................................................................... 115 Weyn, Maarten / Artesis University College of Antwerpen, Belgium................................................ 539 White, Jules / Vanderbilt University, USA.......................................................................................... 502 Williams, Michael / Pepperdine University, USA............................................................................ 1213 Wilson, Greg / Virginia Tech, USA......................................................................................................... 459 Yahya, Abid / Universiti Malaysia Perlis, Malaysia.......................................................................... 157 Yokoyama, Roberto / University of São Paulo, Brazil...................................................................... 522 Zanda, Andrea / Universitad Politecnica, Spain............................................................................... 576 Zaščerinska, Jeļena / University of Latvia, Latvia............................................................................. 760 Zekavat, Seyed (Reza) / Michigan Tech University, USA.................................................................. 115 Zhang, Chongming / Shanghai Normal University, China................................................................ 237

Table of Contents

Foreword . ............................................................................................................................................. lx Preface . ..............................................................................................................................................lxiii Acknowledgment............................................................................................................................. lxxxv Volume I Section 1 Mobile Technologies Chapter 1 Evaluating the Context Aware Browser: A Benchmark for Proactive, Mobile, and Contextual Web Search............................................................................................................................ 1 Davide Menegon, University of Udine, Italy Stefano Mizzaro, University of Udine, Italy Elena Nazzi, University of Copenhagen, Denmark Luca Vassena, University of Udine, Italy Chapter 2 Routing in Wireless Ad Hoc and Sensor Networks............................................................................... 16 Milos Stojmenovic, University of Ottawa, Canada Chapter 3 Mobile Ad Hoc Networks: Protocol Design and Implementation......................................................... 31 Crescenzio Gallo, University of Foggia, Italy Michele Perilli, University of Foggia, Italy Michelangelo De Bonis, University of Foggia, Italy

Chapter 4 Convergence of Wireless Technologies in Consolidating E-Government Applications in Sub-Saharan Africa............................................................................................................................ 48 Bwalya Kelvin Joseph, University of Johannesburg, South Africa Chris Rensleigh, University of Johannesburg, South Africa Mandla Ndlovu, Botswana Accountancy College, Botswana Chapter 5 Process Innovation with Ambient Intelligence (AmI) Technologies in Manufacturing SMEs: Absorptive Capacity Limitations........................................................................................................... 65 Kathryn J. Hayes, University of Western Sydney, Australia Ross Chapman, Deakin University Melbourne, Australia Chapter 6 Providing Ubiquitous Access to Synthetic Sign Language Contents over Multiple Platforms............. 83 Fernando López-Colino, Universidad Autónoma de Madrid, Spain Jose Colás, Universidad Autónoma de Madrid, Spain Chapter 7 The Impact of MIMO Communication on Non-Frequency Selective Channels Performance............ 100 Andreas Ahrens, Hochschule Wismar, Germany César Benavente-Peces, Universidad Politécnica de Madrid, Spain Chapter 8 Node Localization in Ad-Hoc Networks.............................................................................................. 115 Zhonghai Wang, Michigan Tech University, USA Seyed (Reza) Zekavat, Michigan Tech University, USA Chapter 9 Wireless and Mobile Technologies Improving Diabetes Self-Management........................................ 136 Eirik Årsand, Norwegian Centre for Integrated Care and Telemedicine, Norway Naoe Tatara, Norwegian Centre for Integrated Care and Telemedicine, Norway Gunnar Hartvigsen, University of Tromsø, Norway Chapter 10 Adaptive Multicarrier Frequency Hopping Spread Spectrum Combined with Channel Coding........ 157 Othman Sidek, Universiti Sains Malaysia, Malaysia Abid Yahya, Universiti Malaysia Perlis, Malaysia Farid Ghani, Universiti Malaysia Perlis, Malaysia R. Badlishah Ahmad, Universiti Malaysia Perlis, Malaysia M. F. M. Salleh, Universiti Sains Malaysia, Malaysia

Chapter 11 Key Adoption Challenges and Issues of B2B E-Commerce in the Healthcare Sector........................ 175 Chad Lin, Curtin University of Technology, Australia Hao-Chiang Koong Lin, National University of Tainan, Taiwan Geoffrey Jalleh, Curtin University of Technology, Australia Yu-An Huang, National Chi Nan University, Taiwan Chapter 12 A Pervasive Polling Secret-Sharing Based Access Control Protocol for Sensitive Information......... 188 Juan Álvaro Muñoz Naranjo, Universidad de Almería, Spain Justo Peralta López, Universidad de Almería, Spain Juan Antonio López Ramos, Universidad de Almería, Spain Chapter 13 Mobile Location-Based Recommenders: An Advertisement Case Study............................................ 203 Mahsa Ghafourian, University of Pittsburgh, USA Hassan Karimi, University of Pittsburgh, USA Chapter 14 Success Cases for Mobile Devices in a Real University Scenario...................................................... 216 Montserrat Mateos Sánchez, Universidad Pontificia de Salamaca, Spain Roberto Berjón Gallinas, Universidad Pontificia de Salamanca, Spain Encarnación Beato Gutierrez, Universidad Pontificia de Salamanca, Spain Miguel Angel Sánchez Vidales, Universidad Pontificia de Salamanca, Spain Ana María Fermoso García, Universidad Pontificia de Salamanca, Spain Chapter 15 Event Detection in Wireless Sensor Networks.................................................................................... 237 Sohail Anwar, Penn State University, USA Chongming Zhang, Shanghai Normal University, China Chapter 16 M-English Podcast: A Tool for Mobile Devices.................................................................................. 250 Célia Menezes, Escola Sec. Inf. D. Henrique, Portugal Fernando Moreira, Universidade Portucalense, Portugal Chapter 17 Public Safety Networks........................................................................................................................ 267 Giuliana Iapichino, EURECOM, France Daniel Câmara, EURECOM, France Christian Bonnet, EURECOM, France Fethi Filali, EURECOM, Qatar

Chapter 18 Mobile Applications as Mobile Learning and Performance Support Tools in Psychotherapy Activitites..................................................................................................................... 285 María Luisa Pérez-Guerrero, Universidad Politecnica de Catalunya, Spain Jose María Monguet-Fierro, Universidad Politecnica de Catalunya, Spain Carmina Saldaña-García, Universidad de Barcelona, Spain Chapter 19 CampusLocator: A Mobile Location-Based Service for Learning Resources..................................... 298 Hassan Karimi, University of Pittsburgh, USA Mahsa Ghafourian, University of Pittsburgh, USA Chapter 20 The Future of WiMAX......................................................................................................................... 314 Dennis Viehland, Massey University, New Zealand Sheenu Chawla, SUSH Global Solutions, New Zealand Chapter 21 Determinants of Loyalty Intention in Portuguese Mobile Market....................................................... 327 Maria Borges Tiago, University of the Azores, Portugal Francisco Amaral, University of the Azores, Portugal Chapter 22 Mobile Agents: Concepts and Technologies........................................................................................ 343 Agostino Poggi, Università degli Studi di Parma, Italy Michele Tomaiuolo, Università degli Studi di Parma, Italy Chapter 23 Vehicular Delay Tolerant Networks..................................................................................................... 356 Daniel Câmara, EURECOM Sophia Antipolis, France Nikolaos Frangiadakis, University of Maryland, USA Fethi Filali, QU Wireless Innovations Center, Qatar Christian Bonnet, EURECOM Sophia Antipolis, France Chapter 24 Monitoring the Learning Process through the Use of Mobile Devices............................................... 368 Francisco Rodríguez-Díaz, University of Granada, Spain Natalia Padilla Zea, University of Granada, Spain Marcelino Cabrera, University of Granada, Spain

Chapter 25 The Making of Nomadic Work: Understanding the Mediational Role of ICTs................................... 381 Aparecido Fabiano Pinatti de Carvalho, University of Limerick, Republic of Ireland Luigina Ciolfi, University of Limerick, Republic of Ireland Breda Gray, University of Limerick, Republic of Ireland Chapter 26 I-Gate: Interperception – Get all the Environments............................................................................. 397 Rummenigge Dantas, Universidade Federal do Rio Grande do Norte, Brasil Luiz Marcos Gonçalves, Universidade Federal do Rio Grande do Norte, Brasil Claudio Schneider, Universidade Federal do Rio Grande do Norte, Brasil Aquiles Burlamaqui, Universidade Federal do Rio Grande do Norte, Brasil Ricardo Dias, Universidade Federal do Rio Grande do Norte, Brasil Hugo Sena, Universidade Federal do Rio Grande do Norte, Brasil Julio cesar Melo, Universidade Federal do Rio Grande do Norte, Brasil Chapter 27 CONNECTOR: A Geolocated Mobile Social Service......................................................................... 414 Pedro Almeida, University of Aveiro, Portugal Jorge Abreu, University of Aveiro, Portugal Margarida Almeida, University of Aveiro, Portugal Maria Antunes, University of Aveiro, Portugal Lidia Silva, University of Aveiro, Portugal Melissa Saraiva, University of Aveiro, Portugal Jorge Teixeira, University of Aveiro, Portugal Fernando Ramos, University of Aveiro, Portugal Chapter 28 Providing VoIP and IPTV Services in WLANs................................................................................... 426 Miguel Edo, Polytechnic University of Valencia, Spain Alejandro Canovas, Polytechnic University of Valencia, Spain Miguel Garcia, Polytechnic University of Valencia, Spain Jaime Lloret, Polytechnic University of Valencia, Spain Chapter 29 SIe-Health, e-Health Information System............................................................................................ 445 Juan Carlos González Moreno, University of Vigo, Spain Loxo Lueiro Astray, University of Vigo, Spain Rubén Romero González, University of Vigo, Spain Cesar Parguiñas Portas, University of Vigo, Spain Castor Sánchez Chao, University of Vigo, Spain

Chapter 30 Combining Location Tracking and RFID Tagging toward an Improved Research Infrastructure...........459 Greg Wilson, Virginia Tech, USA Scott McCrickard, Virginia Tech, USA Chapter 31 Model and Infrastructure for Communications in Context-Aware Services........................................ 472 Cristina Rodriguez-Sanchez, Universidad Rey Juan Carlos, Spain Susana Borromeo, Universidad Rey Juan Carlos, Spain Juan Hernandez-Tamames, Universidad Rey Juan Carlos, Spain Chapter 32 Network Mobility and Mobile Applications Development................................................................. 487 Rui Rijo, IPLeiria, Portugal Nuno Veiga, IPLeiria, Portugal Silvio Bernardes, IPLeiria, Portugal Chapter 33 Building Mobile Sensor Networks Using Smartphones and Web Services: Ramifications and Development Challenges.............................................................................................................. 502 Hamilton Turner, Vanderbilt University, USA Jules White, Vanderbilt University, USA Brian Dougherty, Vanderbilt University, USA Doug Schmidt, Vanderbilt University, USA Chapter 34 Technologies to Improve the Quality of Handovers: Ontologies, Contexts and Mobility Management.......................................................................................................................... 522 Edson Moreira, University of São Paulo, Brazil Bruno Kimura, University of São Paulo, Brazil Renata Maria Vanni, University of São Paulo, Brazil Roberto Yokoyama, University of São Paulo, Brazil Chapter 35 Making Location-Aware Computing Working Accurately in Smart Spaces....................................... 539 Teddy Mantoro, International Islamic University Malaysia, Malaysia Media Ayu, International Islamic University Malaysia, Malaysia Maarten Weyn, Artesis University College of Antwerpen, Belgium Chapter 36 User Pro-Activities Based on Context History.................................................................................... 558 Teddy Mantoro, International Islamic University Malaysia, Malaysia Media Ayu, International Islamic University Malaysia, Malaysia

Section 2 Emerging Technologies Chapter 37 Research Challenge of Locally Computed Ubiquitous Data Mining.................................................. 576 Aysegul Cayci, Sabanci University, Turkey João Bártolo Gomes, Universitad Politecnica, Spain Andrea Zanda, Universitad Politecnica, Spain Ernestina Menasalvas, Universitad Politecnica, Spain Santiago Eibe, Universitad Politecnica, Spain Chapter 38 Emerging Wireless Networks for Social Applications......................................................................... 595 Raúl Aquino, University of Colima, México Luis Villaseñor, CICESE Research Centre, México Víctor Rangel, National Autonomous University of Mexico, México Miguel García, University of Colima, México Artur Edwards, University of Colima, México Chapter 39 An Approach to Mobile Grid Platforms for the Development and Support of Complex Ubiquitous Application........................................................................................................ 617 Carlo Bertolli, University of Pisa, Italy Daniele Buono, University of Pisa, Italy Gabriele Mencagli, University of Pisa, Italy Marco Vanneschi, University of Pisa, Italy Chapter 40 Towards a Programming Model for Ubiquitous Computing............................................................... 634 Jorge Barbosa, Universidade do Vale do Rio dos Sinos (Unisinos), Brazil Fabiane Dillenburg, Universidade Federal do Rio Grande do Sul, Brazil Alex Garzão, Universidade do Vale do Rio dos Sinos (Unisinos), Brazil Gustavo Lermen, Universidade do Vale do Rio dos Sinos (Unisinos), Brazil Cristiano Costa, Universidade do Vale do Rio dos Sinos (Unisinos), Brazil Volume II Chapter 41 An Agent-Based Operational Virtual Enterprise Framework Enabled by RFID................................. 649 Özgür Ünver, TOBB-University of Economics and Technology, Turkey Bahram Lotfi Sadigh, Middle East Technical University, Turkey

Chapter 42 Ontological Dimensions of Semantic Mobile Web 2.0: First Principles............................................. 667 Gonzalo Aranda-Corral, University of Sevilla, Spain Joaquín Borrego-Díaz, University of Sevilla, Spain Chapter 43 Unobtrusive Interaction with Mobile and Ubiquitous Computing Systems through Kinetic User Interfaces..................................................................................................................................... 689 Vincenzo Pallotta, Webster University, Switzerland Chapter 44 Impact of Advances on Computing and Communication Systems in Automotive Testing................. 703 Luís Serrano, Polithecnic Institute Leiria, Portugal José Costa, University Coimbra, Portugal Manuel Silva, University Coimbra, Portugal Chapter 45 RFID and NFC in the Future of Mobile Computing........................................................................... 719 Diogo Simões, Movensis, Portugal Vitor Rodrigues, Movensis, Portugal Luis Veiga, INESC ID / Technical University of Lisbon, Portugal Paulo Ferreira, INESC ID / Technical University of Lisbon, Portugal Chapter 46 A Multi-Loop Development Process for a Wearable Computing System in Autonomous Logistics......................................................................................................................... 738 Jakub Piotrowski, Bremer Institut für Produktion und Logistik, Germany Carmen Ruthenbeck, Bremer Institut für Produktion und Logistik, Germany Florian Harjes, Bremer Institut für Produktion und Logistik, Germany Bernd Scholz-Reiter, Bremer Institut für Produktion und Logistik, Germany Section 3 Critical Success Factors Chapter 47 Collaboration within Social Dimension of Computing: Theoretical Background, Empirical Findings and Practical Development.................................................................................. 760 Andreas Ahrens, University of Technology, Business and Design, Germany Jeļena Zaščerinska, University of Latvia, Latvia Olaf Bassus, University of Technology, Business and Design, Germany

Chapter 48 Critical Factors in Defining the Mobile Learning Model: An Innovative Process for Hybrid Learning at the Tecnologico de Monterrey, a Mexican University......................................... 774 Violeta Chirino-Barceló, Tecnologico de Monterrey Mexico, Mexico Arturo Molina, Tecnologico de Monterrey Mexico, Mexico Chapter 49 Critical Human Factors on Mobile Applications for Tourism and Entertainment............................... 793 Pedro Campos, University of Madeira, Portugal Chapter 50 Internet Surveys: Opportunities and Challenges.................................................................................. 805 Paula Vicente, UNIDE, ISCTE – Lisbon University Institute, Portugal Elizabeth Reis, UNIDE, ISCTE – Lisbon University Institute, Portugal Chapter 51 Evolvable Production Systems: A Coalition-Based Production Approach.......................................... 821 Marcus Bjelkemyr, The Royal Institute of Technology, Sweden Antonio Maffei, The Royal Institute of Technology, Sweden Mauro Onori, The Royal Institute of Technology, Sweden Section 4 New Business Models Chapter 52 Viable Business Models for M-Commerce: The Key Components..................................................... 837 Jiaxiang Gan, University of Auckland, New Zealand Jairo A. Gutiérrez, Universidad Tecnológica de Bolívar, Colombia Chapter 53 A Service-Based Framework to Model Mobile Enterprise Architectures........................................... 853 Jose Delgado, Instituto Superior Técnico, Portugal Chapter 54 Research-Based Insights Inform Change in IBM M-Learning Strategy............................................. 871 Nabeel Ahmad, IBM Center for Advanced Learning, USA Chapter 55 Location Based E-Commerce System: An Architecture...................................................................... 881 Nuno André Osório Liberato, UTAD, Portugal João Eduardo Quintela Alves de Sousa Varajão, UTAD, Portugal Emanuel Soares Peres Correia, UTAD, Portugal Maximino Esteves Correia Bessa, UTAD, Portugal

Section 5 Security Chapter 56 Overview of Security Issues in Vehicular Ad-Hoc Networks.............................................................. 894 José María De Fuentes, Carlos III University of Madrid, Spain Ana Isabel González-Tablas, Carlos III University of Madrid, Spain Arturo Ribagorda, Carlos III University of Madrid, Spain Chapter 57 Modelling of Location-Aware Access Control Rules.......................................................................... 912 Michael Decker, Karlsruhe Institute of Technology (KIT), Germany Chapter 58 Secure Techniques for Remote Reconfiguration of Wireless Embedded Systems.............................. 930 Abdellah Touhafi, Vrije Universiteit Brussel, Belgium An Braeken, Erasmushogeschool Brussel, Belgium Gianluca Cornetta, Universidad San Pablo-CEU, Spain Nele Mentens, Katholieke Universiteit Leuven, Belgium Kris Steenhaut, Vrije Universiteit Brussel, Belgium Chapter 59 Secure Routing and Mobility in Future IP Networks.......................................................................... 952 Kaj Grahn, Arcada University of Applied Sciences, Finland Jonny Karlsson, Arcada University of Applied Sciences, Finland Göran Pulkkis, Arcada University of Applied Sciences, Finland Section 6 Applications, Surveys and Case Studies Chapter 60 Evaluation of a Mobile Platform to Support Collaborative Learning: Case Study............................. 974 Carlos Quental, Polytechnic Institute of Viseu, Portugal Luis Gouveia, University Fernando Pessoa, Portugal Chapter 61 Power Issues and Energy Scavenging in Mobile Wireless Ad-Hoc and Sensor Networks................. 994 Gianluca Cornetta, Universidad San Pablo-CEU, Spain Abdellah Touhafi, Erasmushogheschool Brussel, Belgium David J. Santos, Universidad San Pablo-CEU, Spain José Manuel Vázquez, Universidad San Pablo-CEU, Spain

Chapter 62 A Low Cost Wireless Sensors Network with Low-Complexity and Fast-Prototyping...................... 1021 João Paulo Carmo, University of Minho, Portugal José Higino Correia, University of Minho, Portugal Chapter 63 Unreliable Failure Detectors for Mobile Ad-Hoc Networks.............................................................. 1039 Luciana Arantes, University Paris 6, France Fabíola Greve, Federal University of Bahia, Brazil Pierre Sens, University Paris 6, France Chapter 64 Mission-Aware Adaptive Communication for Collaborative Mobile Entities................................... 1056 Jérôme Lacouture, Université de Toulouse, France Ismael Bouassida Rodriguez, Université de Toulouse, France Jean-Paul Arcangeli, Paul Sabatier University, France Christophe Chassot, Université de Toulouse, France Thierry Desprats, Paul Sabatier University, France Khalil Drira, Université de Toulouse, France Francisco Garijo, Université de Toulouse, France Victor Noel, Paul Sabatier University, France Michelle Sibilla, Paul Sabatier University, France Catherine Tessier, ONERA Centre de Toulouse – DCSD, France Chapter 65 OntoHealth: An Ontology Applied to Pervasive Hospital Environments.......................................... 1077 Giovani Librelotto, Federal University of Santa Maria, Brazil Iara Augustin, Federal University of Santa Maria, Brazil Jonas Gassen, Federal University of Santa Maria, Brazil Guilherme Kurtz, Federal University of Santa Maria, Brazil Leandro Freitas, Federal University of Santa Maria, Brazil Ricardo Martini, Federal University of Santa Maria, Brazil Renato Azevedo, Federal University of Santa Maria, Brazil Chapter 66 Adoption of Mobile and Information Technology in an Energy Utility in Brazil............................. 1091 Osvaldo Garcia, Pontificia Universidade Católica do PR, Brazil Maria Cunha, Pontificia Universidade Católica do PR, Brazil

Chapter 67 Infrastructures for Development of Context-Aware Mobile Applications......................................... 1104 Hugo Feitosa de Figueirêdo, University of Campina Grande, Brazil Tiago Eduardo da Silva, University of Campina Grande, Brazil Anselmo Cardoso de Paiva, University of Maranhão, Brazil José Eustáquio Rangel de Queiroz, University of Campina Grande, Brazil Cláudio De Souza Baptista, University of Campina Grande, Brazil Chapter 68 A Practice Perspective on Transforming Mobile Work..................................................................... 1119 Riikka Vuokko, Åbo Akademi University, Finland Chapter 69 Data Replication Support for Collaboration in Mobile and Ubiquitous Computing Environments.......... 1132 João Barreto, INESC-ID/Technical University Lisbon, Portugal Paulo Ferreira, INESC-ID/Technical University Lisbon, Portugal Chapter 70 Providing Outdoor and Indoor Ubiquity with WLANs..................................................................... 1155 Diana Bri, Polytechnic University of Valencia, Spain Hugo Coll, Polytechnic University of Valencia, Spain Sandra Sendra, Polytechnic University of Valencia, Spain Jaime Lloret, Polytechnic University of Valencia, Spain Chapter 71 In-TIC for Mobile Devices: Support System for Communication with Mobile Devices for the Disabled.................................................................................................................................. 1169 Cristina Diaz Busch, University of A Coruna, Spain Alberto Moreiras Lorenzo, University of A Coruna, Spain Iván Mourelos Sánchez, University of A Coruna, Spain Betania Groba González, University of A Coruna, Spain Thais Pousada García, University of A Coruna, Spain Laura Nieto Riveiro, University of A Coruna, Spain Javier Pereira Loureiro, University of A Coruna, Spain Chapter 72 New Ways to Buy and Sell: An Information Management Web System for the Commercialization of Agricultural Products from Family Farms without Intermediaries................ 1182 Carlos Ferrás, University of Santiago Compostela, Spain Yolanda García, University of Santiago Compostela, Spain Mariña Pose, University of Santiago Compostela, Spain

Chapter 73 Broadcast Quality Video Contribution in Mobility........................................................................... 1199 José Ramón Cerquides Bueno, University of Seville, Spain Antonio Foncubierta Rodriguez, University of Seville, Spain Chapter 74 Mobile Device Selection in Higher Education: iPhone vs. iPod Touch............................................ 1213 C. Brad Crisp, Abilene Christian University, USA Michael Williams, Pepperdine University, USA Chapter 75 Design of Wearable Computing Systems for Future Industrial Environments.................................. 1226 Pierre Kirisci, Bremer Institut für Produktion und Logistik GmbH, Germany Ernesto Morales Kluge, Bremer Institut für Produktion und Logistik GmbH, Germany Emanuel Angelescu, Bremer Institut für Produktion und Logistik GmbH, Germany Klaus-Dieter Thoben, Bremer Institut für Produktion und Logistik GmbH, Germany Chapter 76 Extending the Scope of eID Technology: Threats and Opportunities in a Commercial Setting........ 1246 Vincent Naessens, Katholieke Hogeschool Sint-Lieven, Belgium Bart De Decker, Katholieke Universiteit Leuven, Belgium Chapter 77 Mobility and Connectivity: On the Character of Mobile Information Work..................................... 1262 Victor Gonzalez, University of Manchester, UK Antonis Demetriou, University of Manchester, UK

Detailed Table of Contents

Foreword . ............................................................................................................................................. lx Preface . ..............................................................................................................................................lxiii Acknowledgment............................................................................................................................. lxxxv Section 1 Mobile Technologies Chapter 1 Evaluating the Context Aware Browser: A Benchmark for Proactive, Mobile, and Contextual Web Search............................................................................................................................ 1 Davide Menegon, University of Udine, Italy Stefano Mizzaro, University of Udine, Italy Elena Nazzi, University of Copenhagen, Denmark Luca Vassena, University of Udine, Italy Chapter 1 discusses the evaluation of highly interactive and novel context-aware system with a methodology based on a TREC-like benchmark. The authors take as a case study an application for Web content perusal by means of context-aware mobile devices, named Context-Aware Browser. In this application, starting from the representation of the user’s current context, queries are automatically constructed and used to retrieve the most relevant Web contents. Since several alternatives for query construction exist, it is important to compare their effectiveness, and to this aim we developed a TREClike benchmark. The authors present their approach to an early stage evaluation, describing their aims and the techniques applied, underlining for, for the evaluation of context-aware retrieval systems, the benchmark methodology adopted could be an extensible and reliable tool. Chapter 2 Routing in Wireless Ad Hoc and Sensor Networks............................................................................... 16 Milos Stojmenovic, University of Ottawa, Canada Routing is the process of finding a path from a source node to a destination node. Proposed routing schemes can be divided into topological and position based, depending on the availability of geograph-

ic location for nodes. Topological routing may be proactive or reactive, while position based routing consists of greedy approaches applied when a neighbor closer to the destination (than the node currently holding the packet) exists, and recovery schemes otherwise. In order to preserve bandwidth and power which are critical resources in ad hoc and sensor networks, localized approaches are proposed, where each node acts based solely on the location of itself, its neighbors, and the destination. There are various measures of optimality which lead to various schemes which optimize hop count, power, network lifetime, delay, or other metrics. This second chapter describes a uniform solution based on ratio of cost and progress. Chapter 3 Mobile Ad Hoc Networks: Protocol Design and Implementation......................................................... 31 Crescenzio Gallo, University of Foggia, Italy Michele Perilli, University of Foggia, Italy Michelangelo De Bonis, University of Foggia, Italy Mobile communication networks have become an integral part of our society, significantly enhancing communication capabilities. Mobile ad hoc networks (MANETs) extend this capability to any time/ anywhere communication, providing connectivity without the need of an underlying infrastructure. Chapter 2 investigates the new coming realm of mobile ad hoc networks, focusing on research problems related to the design and development of routing protocols, both from a formal and technical point of view. Then the link stability in a high mobility environment is examined, and a route discovery mechanism is analyzed, together with a practical implementation of a routing protocol in ad hoc multirate environments which privileges link stability instead of traditional speed and minimum distance approaches. Chapter 4 Convergence of Wireless Technologies in Consolidating E-Government Applications in Sub-Saharan Africa............................................................................................................................ 48 Bwalya Kelvin Joseph, University of Johannesburg, South Africa Chris Rensleigh, University of Johannesburg, South Africa Mandla Ndlovu, Botswana Accountancy College, Botswana The convergence of wireless applications presents a greater hope for consolidating e-Government applications even in resource-constrained countries such as those in Africa. This chapter presents an exploratory study that aims at discussing the extent as to how convergence of wireless technologies from different vendors promises to contribute to the consolidation of e-Government applications in Sub-Saharan-Africa (SSA). This is done by reviewing the different adoption stages of ICT and eGovernment in SSA. It looks at challenges facing adoption of wireless technologies (GSMs, Wireless Internet Access, satellite transmission, etc.) across all the socio-economic value chains in SSA. The chapter looks at Botswana and South Africa as case studies by bringing out the different interventions that have been done in the realm of facilitating a conducive environment for the convergence of different wireless technologies. Out of the analysis of legal, regulatory, market and spectrum policies affecting the adoption of wireless communications in SSAs, the chapter draws out recommendations on how to consolidate wireless communications to be adopted in different socio-economic setups (e.g. e-government, e-Health, e-Banking, etc.).

Chapter 5 Process Innovation with Ambient Intelligence (AmI) Technologies in Manufacturing SMEs: Absorptive Capacity Limitations........................................................................................................... 65 Kathryn J. Hayes, University of Western Sydney, Australia Ross Chapman, Deakin University Melbourne, Australia This Chapter 5 considers the potential for absorptive capacity limitations to prevent SME manufacturers benefiting from the implementation of Ambient Intelligence (AmI) technologies. The chapter also examines the role of intermediary organisations in alleviating these absorptive capacity constraints. In order to understand the context of the research, a review of the role of SMEs in the Australian manufacturing industry, plus the impacts of government innovation policy and absorptive capacity constraints in SMEs in Australia is provided. Advances in the development of ICT industry standards, and the proliferation of software and support for the Windows/Intel platform have brought technology to SMEs without the need for bespoke development. The results from the joint European and Australian AmI4-SME projects suggest that SMEs can successfully use “external research sub-units” in the form of industry networks, research organisations and technology providers to offset internal absorptive capacity limitations. Chapter 6 Providing Ubiquitous Access to Synthetic Sign Language Contents over Multiple Platforms............. 83 Fernando López-Colino, Universidad Autónoma de Madrid, Spain Jose Colás, Universidad Autónoma de Madrid, Spain Chapter 6 presents the design of distributed sign language synthesis architecture. The main objective of this design is to adapt the synthesis process to the diversity of user devices. The synthesis process has been divided into several independent modules that can be run either in a dedicated server or in the client device. Depending on the modules assigned to the server or to the client, four different scenarios have been defined. These scenarios may vary from a heavy client design which executes the whole synthesis process, to a light client design similar to a video player. These four scenarios will provide equivalent signed message quality independently of the device’s hardware and software resources. Chapter 7 The Impact of MIMO Communication on Non-Frequency Selective Channels Performance............ 100 Andreas Ahrens, Hochschule Wismar, Germany César Benavente-Peces, Universidad Politécnica de Madrid, Spain Chapter 7 reviews the basic concepts of multiple-input multiple-output (MIMO) communication systems and analyses their performance within non-frequency selective channels. The MIMO system model is established and by applying the singular value decomposition (SVD) to the channel matrix, the whole MIMO system can be transformed into multiple single-input single-output (SISO) channels having unequal gains. In order to analyze the system performance, the quality criteria needed to calculate the error probability of M-ary QAM (Quadrature Amplitude Modulation) are briefly reviewed and used as reference to measure the improvements when applying different signal processing techniques. Bit and power allocation is a well-known technique that allows improvement in the bit-error rate (BER)

by managing appropriately the different properties of the multiple SISO channels. It can be used to balance the BER’s in the multiple SISO channels when minimizing the overall BER. In order to compare the various results, the efficiency of fixed transmission modes is studied in this work regardless of the channel quality. It is demonstrated that only an appropriate number of MIMO layers should be activated when minimizing the overall BER under the constraints of a given fixed date rate. Chapter 8 Node Localization in Ad-Hoc Networks.............................................................................................. 115 Zhonghai Wang, Michigan Tech University, USA Seyed (Reza) Zekavat, Michigan Tech University, USA Chapter 8 introduces node localization techniques in ad-hoc networks including received signal strength (RSS), time-of-arrival (TOA) and direction-of-arrival (DOA). Wireless channels in ad-hoc networks can be categorized as LOS and NLOS. In LOS channels, the majority of localization techniques perform properly. However, in NLOS channels, the performance of these techniques reduces. Therefore, non-line-of-sight (NLOS) identification and mitigation techniques, and localization techniques for NLOS scenarios are briefly reviewed. Chapter 9 Wireless and Mobile Technologies Improving Diabetes Self-Management........................................ 136 Eirik Årsand, Norwegian Centre for Integrated Care and Telemedicine, Norway Naoe Tatara, Norwegian Centre for Integrated Care and Telemedicine, Norway Gunnar Hartvigsen, University of Tromsø, Norway The technological revolution that has created a vast health problem due to a drastic change in lifestyle also holds great potential for individuals to take better care of their own health. This is the focus of the presented overview of current applications, and prospects for future research and innovations. The presented overview and the main goals of the systems included are to utilize ICT as aids in self-management of individual health challenges, for the disease Diabetes, both for Type 1 and Type 2 diabetes. People with diabetes are generally as mobile as the rest of the population, and should have access to mobile technologies for managing their disease. Forty-seven relevant studies and prototypes of mobile, diabetes-specific self-management tools meeting our inclusion criteria have been identified; 27 publicly available products and services, nine relevant patent applications, and 31 examples of other diseaserelated mobile self-management systems are included to provide a broader overview of the state of the art. Finally, the reviewed systems are compared, and future research directions are suggested. Chapter 10 Adaptive Multicarrier Frequency Hopping Spread Spectrum Combined with Channel Coding........ 157 Othman Sidek, Universiti Sains Malaysia, Malaysia Abid Yahya, Universiti Malaysia Perlis, Malaysia Farid Ghani, Universiti Malaysia Perlis, Malaysia R. Badlishah Ahmad, Universiti Malaysia Perlis, Malaysia M. F. M. Salleh, Universiti Sains Malaysia, Malaysia

Chapter 10 presents an adaptive Multicarrier Frequency Hopping Spread Spectrum (MCFH-SS) system employing proposed Quasi Cyclic Low Density Parity Check (QC-LDPC) codes instead of the conventional LDPC codes. A new technique for constructing the QC-LDPC codes based on row division method is proposed. The new codes offer more flexibility in terms of girth, code rates and codeword length. Moreover, a new scheme for channel prediction in MCFH-SS system is also proposed. The technique adaptively estimates the channel conditions and eliminates the need for the system to transmit a request message prior to transmitting the packet data. The proposed adaptive MCFH-SS system uses PN sequences to spread out frequency spectrum, reduce the power spectral density and minimize the jammer effects. Chapter 11 Key Adoption Challenges and Issues of B2B E-Commerce in the Healthcare Sector........................ 175 Chad Lin, Curtin University of Technology, Australia Hao-Chiang Koong Lin, National University of Tainan, Taiwan Geoffrey Jalleh, Curtin University of Technology, Australia Yu-An Huang, National Chi Nan University, Taiwan Although B2B e-commerce provides healthcare organizations a wealth of new opportunities and ways of doing business, it also presents them with a series of challenges. B2B e-commerce adoption remain poorly understood and it is also a relatively under-researched area. Therefore, case studies were conducted to investigate the challenges and issues in adopting and utilizing B2B e-commerce systems in the healthcare sector. The major aims of the study presented in Chapter 11 are to: (a) identify and examine main B2B e-commerce adoption challenges and issues for healthcare organizations; and (b) develop a B2B e-commerce adoption challenges and issues table to assist healthcare organizations in identifying and managing them. Chapter 12 A Pervasive Polling Secret-Sharing Based Access Control Protocol for Sensitive Information......... 188 Juan Álvaro Muñoz Naranjo, Universidad de Almería, Spain Justo Peralta López, Universidad de Almería, Spain Juan Antonio López Ramos, Universidad de Almería, Spain Chapter 12 presents a novel access control mechanism for sensitive information which requires permission from different entities or persons to be accessed. The mechanism consists of a file structure and a protocol which extend the features of the OpenPGP Message Format standard by using secret sharing techniques. Several authors are allowed to work in the same file, while access is blocked for not authorized users. Access control rules can be set indicating the minimum number of authors that need to be gathered together in order to open the file. Furthermore, these rules can be different for each section of the document, allowing collaborative work. Non-repudiation and authentication are achieved by means of a shared signature. The scheme’s features are best appreciated when using it in a mobile scenario. Deployment in such an environment is easy and straight

Chapter 13 Mobile Location-Based Recommenders: An Advertisement Case Study............................................ 203 Mahsa Ghafourian, University of Pittsburgh, USA Hassan Karimi, University of Pittsburgh, USA Mobile devices, including cell phones, capable of geo-positioning (or localization) are paving the way for new computer assisted systems called mobile location-based recommenders (MLBRs). MLBRs are systems that combine information on user’s location with information about user’s interests and requests to provide recommendations that are based on “location”. MLBR applications are numerous and emerging. One MLBR application is in advertisement where stores announce their coupons and users try to find the coupons of their interests nearby their locations through their cell phones. Chapter 13 discusses the concept and characteristics of MLBRs and presents the architecture and components of a MLBR for advertisement. Chapter 14 Success Cases for Mobile Devices in a Real University Scenario...................................................... 216 Montserrat Mateos Sánchez, Universidad Pontificia de Salamaca, Spain Roberto Berjón Gallinas, Universidad Pontificia de Salamanca, Spain Encarnación Beato Gutierrez, Universidad Pontificia de Salamanca, Spain Miguel Angel Sánchez Vidales, Universidad Pontificia de Salamanca, Spain Ana María Fermoso García, Universidad Pontificia de Salamanca, Spain Mobile devices have become a new platform with many possibilities to develop studies and implement projects. The power and current capabilities of these devices besides its market penetration makes applications and services in the area of mobility particularly interesting. Mobile terminals have become small computers, they have an operating system, storage capacity so it is possible to develop applications that run on them. Today these applications are highly valued by users. Nowadays we want not only to talk or send messages by mobile terminal, but also to play games, to buy cinema tickets, to read email… We can bring these capabilities in our pocket. The University may not be aware of this fact. The students, due to their age, are the main users and purchasers. In this sense, Chapter 14 presents three applications developed for mobile devices, that are being used in Universidad Pontificia de Salamanca. All of them work on a university scenario and use different kind of services. Chapter 15 Event Detection in Wireless Sensor Networks.................................................................................... 237 Sohail Anwar, Penn State University, USA Chongming Zhang, Shanghai Normal University, China Wireless Sensor Networks (WSNs) have experienced an amazing evolution during the last decade. Compared with other wired or wireless networks, wireless sensor networks extend the range of data collection and make it possible for us to get information from every corner of the world. The chapter begins with an introduction to WSNs and their applications. Chapter 15 recognizes event detection as a key component for WSN applications. The chapter provides a structured and comprehensive overview of various techniques used for event detection in WSNs. Existing event detection techniques have

been grouped into threshold based and pattern based mechanisms. For each category of event detection mechanism, the chapter surveys some representative technical schemes. The chapter also provides some analyses on the relative strengths and weaknesses of these technical schemes. Towards the end, the trends in the research regarding the event detection in WSNs are described. Chapter 16 M-English Podcast: A Tool for Mobile Devices.................................................................................. 250 Célia Menezes, Escola Sec. Inf. D. Henrique, Portugal Fernando Moreira, Universidade Portucalense, Portugal At the beginning of the 21st century in a world dominated by technology it is essential to enhance and update the school, creating conditions for the students to succeed, consolidating the role of ICT as a key resource for learning and teaching in this new era. In Chapter 16 the authors describe a study that was carried out in a Portuguese school. As a means to overcome some of the existing logistical obstacles in the school, where the possibility of carrying out ICT activities without restrictions was still dreamlike, the podcast was implemented as an m-learning tool. Being aware of the fact that nowadays mobile phones and mp3 players are part of our students’ lives, the authors took advantage of this fact and the podcast was used as a tool to support, to enhance and to motivate students to learn English, used thus as a complement to traditional (face-to-face) learning. Chapter 17 Public Safety Networks........................................................................................................................ 267 Giuliana Iapichino, EURECOM, France Daniel Câmara, EURECOM, France Christian Bonnet, EURECOM, France Fethi Filali, EURECOM, Qatar Disaster can be defined as the onset of an extreme event causing profound damage or loss as perceived by the afflicted people. The networks built in order to detect and handle these events are called Public safety networks (PSNs). These networks have the fundamental role of providing communication and coordination for emergency operations. Many of the problems of the PSN field come from the heterogeneity of systems and agencies involved in the crisis site and from their mobility at the disaster site. The main aim of Chapter 17 is to provide a broad view of the PSN field, presenting the different emergency management phases, PSNs requirements, technologies and some of the future research directions for this field. Chapter 18 Mobile Applications as Mobile Learning and Performance Support Tools in Psychotherapy Activitites..................................................................................................................... 285 María Luisa Pérez-Guerrero, Universidad Politecnica de Catalunya, Spain Jose María Monguet-Fierro, Universidad Politecnica de Catalunya, Spain Carmina Saldaña-García, Universidad de Barcelona, Spain

The purpose of Chapter 18 is the analysis of mobile applications as performance and informal learning support tools that facilitate the development of the psychotherapy process. The “e-therapy” has become a common term to refer the delivery of mental health services, on-line or related to a computer mediated communication between a psychotherapist and the patient. Initially, a background on e-therapy is provided and after the “self-help therapy”–a kind of e-therapy where the concept of patient empowerment is important– is exposed. Then, the integration of mobile devices in the psychotherapy process will be explained considering how their technological features support patient therapeutic activities like behavior assessment and informal mobile learning. The relation of the mobile devices with psychotherapist work activities such as evidence gathering and patient monitoring will also be explained. The chapter includes a discussion on the mobile learning practices as a source of potential strategies that can be applied in the therapeutic field and finally a set of recommendations and future directions are described to explore new lines of research. Chapter 19 CampusLocator: A Mobile Location-Based Service for Learning Resources..................................... 298 Hassan Karimi, University of Pittsburgh, USA Mahsa Ghafourian, University of Pittsburgh, USA Location-based services (LBSs) are impacting different aspects of human’s life. To date, different LBSs have been emerged, each supporting a specific application or service. While some LBSs have aimed at addressing the needs of general populations, such as navigation systems, others have been focused on addressing the needs of specific populations, including kids, youths, elderly, and people with special needs. In recent years, interest in taking a LBS approach in education and learning has grown. The main purpose of such educational LBSs is facilitating a means for learners to be more efficient and effective in their learning activities using their location as the underlying information in decision making. In this chapter, we present a novel LBS, called CampusLocator, whose main goal is to assist students in locating and accessing learning resources including libraries, seminars, and tutorials that are available on a campus. Chapter 20 The Future of WiMAX......................................................................................................................... 314 Dennis Viehland, Massey University, New Zealand Sheenu Chawla, SUSH Global Solutions, New Zealand WiMAX is being promoted as a potential solution to a number of problems that have plagued the wired and wireless broadband industry since it originated. Can WiMAX fulfill this promise in a crowded and competitive market? If so, what factors are critical to its success? Who will use WiMAX and for what purposes? This chapter identifies both the critical success factors that will give WiMAX an edge over other existing wireless technologies and the key applications that will contribute to its success. The top three critical success factors for WiMAX are availability of handset devices and consumer premise equipment, bandwidth speed, and interoperability and standardization. A panel of WiMAX experts concludes that broadband on demand, wireless services provider access, and Voice over IP are the top three killer applications for WiMAX.

Chapter 21 Determinants of Loyalty Intention in Portuguese Mobile Market....................................................... 327 Maria Borges Tiago, University of the Azores, Portugal Francisco Amaral, University of the Azores, Portugal Chapter 21 conceptualizes and highlights the determinants of customers’ loyalty in the Portuguese mobile market. The authors raise questions about the interrelationships of the cost and values dimensions and the consequences of these relationships on customer satisfaction and trust and consequently loyalty among different operators, addressing some recent models. By organizing and synthesizing the major research streams and tests empirically a conceptual framework through a SEM, with data gather in a survey of Portuguese clients, the present study advances knowledge on the nature of the relative importance of different components of loyalty to mobile communications operators. Some useful preliminary insights were produced related to customers’ retention process in primary mobile operator, which appears strongly related to price/quality, followed by the emotional connection to the operator staff and others clients. Nonetheless, a considerable number of issues were left for future research, including the possibility of extending the investigation to other countries. Chapter 22 Mobile Agents: Concepts and Technologies........................................................................................ 343 Agostino Poggi, Università degli Studi di Parma, Italy Michele Tomaiuolo, Università degli Studi di Parma, Italy Current technological advances and the increasing diffusion of its use for scientific, financial and social activities, make Internet the de facto platform for providing worldwide distributed data storage, distributed computing and communication. It creates new opportunities for the development of new kinds of applications, but it will also create several challenges in managing the information distributed on the Internet and in guaranteeing its “on-time” access through the network infrastructures that realize the Internet. Many researchers believed and still believe that the mobile agents could propose several attractive solutions to deal with such challenges and problems. Chapter 22 presents the core concepts of mobile agents, and attempts to provide a clear idea of the possibility of their use by introducing the problems they cope with, the application areas where they provide advantages with respect to other technologies and the available mobile agent technologies. Chapter 23 Vehicular Delay Tolerant Networks..................................................................................................... 356 Daniel Câmara, EURECOM Sophia Antipolis, France Nikolaos Frangiadakis, University of Maryland, USA Fethi Filali, QU Wireless Innovations Center, Qatar Christian Bonnet, EURECOM Sophia Antipolis, France Traditional networks suppose the existence of some path between endpoints, small end to end roundtrip Delay time, and loss ratio. Today, however, new applications, environments and types of devices are challenging these assumptions. In Delay Tolerant Networks (DTNs), an end-to-end path from source to destination may not exist. Nodes may connect and exchange information in an opportunistic

way. Chapter 23 presents a broad overview of DTNs, particularly focusing on Vehicular DTNs, their main characteristics, challenges, and research projects on this field. In the near future, cars are expected to be equipped with devices that will allow them to communicate wirelessly. However, there will be strict restrictions to the duration of their connections with other vehicles, whereas the conditions of their links will greatly vary; DTNs present an attractive solution. Therefore, VDTNs constitute an attractive research field. Chapter 24 Monitoring the Learning Process through the Use of Mobile Devices............................................... 368 Francisco Rodríguez-Díaz, University of Granada, Spain Natalia Padilla Zea, University of Granada, Spain Marcelino Cabrera, University of Granada, Spain It has been substantially proven that computer operation can be learnt at an early age, and that the use of new technologies can improve a child’s learning process. However, the main problem for the teacher continues to be that he/she cannot pay attention to all children at the same time. Sometimes it is necessary to decide which child must be first attended to. It is in this context that we believe our system has the ability to greatly help teachers: we have developed a learning process control system that allows teachers to determine which students have problems, how many times a child has failed, which activities they are working on and other such useful information, in order to decide how to distribute his/ her time. Furthermore, bearing in mind the attention required by kindergarten students, we propose the provision of mobile devices (PDA - Personal Digital Assistant) for teachers, permitting free movement in the classroom and allowing the teacher to continue to help children while information about other students is being received. Therefore if a new problem arises the teacher is immediately notified and can act accordingly. Chapter 25 The Making of Nomadic Work: Understanding the Mediational Role of ICTs................................... 381 Aparecido Fabiano Pinatti de Carvalho, University of Limerick, Republic of Ireland Luigina Ciolfi, University of Limerick, Republic of Ireland Breda Gray, University of Limerick, Republic of Ireland Computer technologies, especially ICT have become ubiquitous in people’s lives. Nowadays, mobile phones, PDAs, laptops and a constellation of software tools are more and more used for a variety of activities carried out in both personal and professional lives. Given the features that these technologies provide and are provided with, for example, connectivity and portability, it can be said that ICTs have the potential to support nomadic work practices which are seen as increasingly characteristic of the knowledge economy. Chapter 25 presents a review of the concept of nomadic work and, based on a broad literature analysis, discusses the ways in which ICTs may empower people who are involved with nomadic work practices. It aims to give a starting point for those who intend to develop further research on technologically-mediated nomadic work practices in the knowledge economy.

Chapter 26 I-Gate: Interperception – Get all the Environments............................................................................. 397 Rummenigge Dantas, Universidade Federal do Rio Grande do Norte, Brasil Luiz Marcos Gonçalves, Universidade Federal do Rio Grande do Norte, Brasil Claudio Schneider, Universidade Federal do Rio Grande do Norte, Brasil Aquiles Burlamaqui, Universidade Federal do Rio Grande do Norte, Brasil Ricardo Dias, Universidade Federal do Rio Grande do Norte, Brasil Hugo Sena, Universidade Federal do Rio Grande do Norte, Brasil Julio cesar Melo, Universidade Federal do Rio Grande do Norte, Brasil Chapter 26 introduces the I-GATE architecture, a new approach, which includes a set of rules and software architecture, to connect users from different interfaces and devices in the same virtual environment, transparently, even with low capacity of resources. The system detects the user resources and provides transformations on the data in order for its visualization in 3D, 2D and textual-only (1D) interfaces. This allows users from any interface to get a connection in the system using any device and to access and exchange information with other users (including ones with other interface types) in a straightforward way, without need to changing hardware or software. The authors formalize the problem, including modeling, implementation, and usage of the system, also introducing some applications that they have created and implemented in order to evaluate the proposal. Authors have used these applications in cell phones, PDAs, Digital Television, and heterogeneous computers, using the same architecture, with success. Chapter 27 CONNECTOR: A Geolocated Mobile Social Service......................................................................... 414 Pedro Almeida, University of Aveiro, Portugal Jorge Abreu, University of Aveiro, Portugal Margarida Almeida, University of Aveiro, Portugal Maria Antunes, University of Aveiro, Portugal Lidia Silva, University of Aveiro, Portugal Melissa Saraiva, University of Aveiro, Portugal Jorge Teixeira, University of Aveiro, Portugal Fernando Ramos, University of Aveiro, Portugal The widespread and availability of increasingly powerful mobile devices is contributing for the incorporation of new services and features on our daily communications and social relationships. In this context, geolocation of users and points of interest in mobile devices may contribute, in a natural way, to support either the mediation of remote conversations as the promotion of face-to-face meetings between users, leveraging social networks. The CONNECTOR system is based on geolocation data (people, content and activities), enabling users to create and develop their personal relations with other members of the CONNECTOR social network. Users, maps, sharing features and multimedia content are actors in this social network allowing CONNECTOR to address the promotion of geolocated social networks driven by physical proximity and common interests among users. Chapter 27 discusses the work undertaken for the conceptualization and development of the CONNECTOR system. Preliminary evaluation results along with usage contexts are also presented. The chapter concludes with a discussion about future developments in geolocation and personalization in mobile communication services.

Chapter 28 Providing VoIP and IPTV Services in WLANs................................................................................... 426 Miguel Edo, Polytechnic University of Valencia, Spain Alejandro Canovas, Polytechnic University of Valencia, Spain Miguel Garcia, Polytechnic University of Valencia, Spain Jaime Lloret, Polytechnic University of Valencia, Spain Nowadays, triple-play services are offered in both wireless and wired networks. The network convergence and the new services such as VoIP and IPTV are a reality. However, the future of these networks will have a different concept, the ubiquity. The solutions must be based on the structures and current environments to carry out those challenges in a correct way. In order to reach this ubiquity, the science community has to take into account that its implementation should not assume a high cost to the user and that the system must comply with the quality of service measurements to satisfy the user. Chapter 28 introduces the main VoIP and IPTV (IP Television) transmission protocols and the compression formats most used as also the bandwidth needed. The goal is to provide ubiquity into multimedia scenarios in WLANs. Authors will carry out tests to guarantee the appropriate values in some network parameters such jitter, delay, number of lost packets and enough effective bandwidth that should be satisfied. They demonstrate the measurements taken from several test benches. They show the parameter values that the devices should perform in order to stay connected from anywhere at any time to these services. Chapter 29 SIe-Health, e-Health Information System............................................................................................ 445 Juan Carlos González Moreno, University of Vigo, Spain Loxo Lueiro Astray, University of Vigo, Spain Rubén Romero González, University of Vigo, Spain Cesar Parguiñas Portas, University of Vigo, Spain Castor Sánchez Chao, University of Vigo, Spain In recent years, the incessant development of new communication technologies has provided a better way for accessing information and a countless amount of useful opportunities. One of the sectors that have a great potential to use and exploit this kind of technologies is the healthcare sector. Nowadays, the application of all these new technologies to support the clinical procedures has taken part in the definition of a new concept known as e-Health. This concept involves a lot of different services related with the medicine/health terms and the information technologies. However, to provide emergency transportation with better care capabilities to the patient is something that still has a lot to improve. Within this context SIe-Health comes into being a software platform oriented for developing Telemedicine solutions. The solution model proposed in this chapter allows remote assistance for a mobile health emergency, integrating in this service electro-medical devices and videoconference services. Chapter 30 Combining Location Tracking and RFID Tagging toward an Improved Research Infrastructure...........459 Greg Wilson, Virginia Tech, USA Scott McCrickard, Virginia Tech, USA

The popularity of mobile computing creates new opportunities for information sharing and collaboration through technologies like radio frequency identification (RFID) tags and location awareness technologies. Chapter 30 discusses how these technologies, which provide subtly different information, can be used together toward increased benefit to users. This work introduces technologies for RFID and location awareness, including a survey of projects. The authors describe advantages of combining these technologies, illustrated through our system, TagIt, that uses these technologies in a traditional research poster environment to provide a rich multimedia experience and encourage ongoing feedback from poster viewers. An overview of TagIt is provided, including user commenting and information sharing capabilities that make use of RFID and location information. User feedback and an expert review highlights how TagIt could benefit authors, information consumers, and the research community, leading to future directions for the research community. Chapter 31 Model and Infrastructure for Communications in Context-Aware Services........................................ 472 Cristina Rodriguez-Sanchez, Universidad Rey Juan Carlos, Spain Susana Borromeo, Universidad Rey Juan Carlos, Spain Juan Hernandez-Tamames, Universidad Rey Juan Carlos, Spain The appearance of concepts such as “Ambient Intelligent”, “Ubiquitous Computing” and “ContextAwareness” is causing the development of a new type of services called “Context-Aware Services” that in turn may affect users of mobile communications. This technology revolution is a complex process because of the heterogeneity of contents, devices, objects, technologies, resources and users that can be coexist at the same local environment. The novel approach in Chapter 31 is the development of a ”Local Infrastructure” in order to provide intelligent, transparent and adaptable services to the user and solves the problem of local context control. Authors present a conceptual model for development of local infrastructure, an architecture design to control the service offered by the local infrastructure. This infrastructure proposed consists of an intelligent device network to link the personal portable device with the contextual services. The device design is modular, flexible, scalable, adaptable and reconfigurable remotely in order to tolerate new demand services whenever is needed. Finally, the result suggests that authors will be able to develop a wide range of new and useful applications, not conceived at origin. Chapter 32 Network Mobility and Mobile Applications Development................................................................. 487 Rui Rijo, IPLeiria, Portugal Nuno Veiga, IPLeiria, Portugal Silvio Bernardes, IPLeiria, Portugal The use of mobile devices with possible connection to the Internet is increasing tremendously. This mobility poses new challenges at various levels, including hardware, network services, and the development of applications. The user searches small and lightweight devices, easy to use, and with vast autonomy in terms of energy. She/He seeks also to connect the Internet “every time, everywhere”, possibly using different access technologies. Given the interface limitations, and processing capabilities of small mobile devices, the software and the operating system used must be necessarily adapted. Chapter 32 overviews the mobility area, provides deep insight in the field, and presents the main existing prob-

lems. Mobility and the development of mobile applications are closed related. The advances in network mobility lead to different approaches in the mobile applications development. The chapter proposes a model for developing mobile applications, based on our research. Chapter 33 Building Mobile Sensor Networks Using Smartphones and Web Services: Ramifications and Development Challenges.............................................................................................................. 502 Hamilton Turner, Vanderbilt University, USA Jules White, Vanderbilt University, USA Brian Dougherty, Vanderbilt University, USA Doug Schmidt, Vanderbilt University, USA Wireless sensor networks are composed of geographically dispersed sensors that work together to monitor physical or environmental conditions. In addition, wireless sensor networks are used in many industrial, social, and regulatory applications, including industrial process monitoring and control, environment and habitat monitoring, healthcare, home automation, and traffic control. Developers of wireless sensor networks face a number of programming and deployment challenges. Chapter 33 shows how smartphones can help reduce the development, operation, and maintenance costs of wireless sensor networks, while also enabling these networks to use web services, high-level programming APIs, and increased hardware capability, such as powerful microprocessors. Moreover, this chapter examines key challenges associated with developing and maintaining a large wireless sensor network and presents a novel smartphone wireless sensor network that uses smartphones as sensor nodes. The work is validated in the context of Wreck Watch, which is a smartphone-based sensor network for detecting traffic accidents that authors use to demonstrate solutions to multiple challenges in current wireless sensor networks. The authors also describe common pitfalls of using smartphones as sensor nodes in wireless sensor networks and summarize how have addressed these pitfalls in Wreck Watch. Chapter 34 Technologies to Improve the Quality of Handovers: Ontologies, Contexts and Mobility Management................................................................................................................... 522 Edson Moreira, University of São Paulo, Brazil Bruno Kimura, University of São Paulo, Brazil Renata Maria Vanni, University of São Paulo, Brazil Roberto Yokoyama, University of São Paulo, Brazil Modern life makes people internet-dependents. They want to move connected and care for always getting the best options for connectivity, hoping between providers. Freedom for choosing providers and the business options which these exchanges can offer are the motivations for Chapter 34. After pointing out some characteristics which make the basics of the current handover technologies, we describe an information infrastructure, based on context and ontologies which can be used to foster an intelligent, efficient and profitable scenario for managing handovers in the Next Generation Networks. Some experiments are described and the potential of using these technologies are evaluated.

Chapter 35 Making Location-Aware Computing Working Accurately in Smart Spaces....................................... 539 Teddy Mantoro, International Islamic University Malaysia, Malaysia Media Ayu, International Islamic University Malaysia, Malaysia Maarten Weyn, Artesis University College of Antwerpen, Belgium In smart environments, making a location-aware personal computing working accurately is a way of getting close to the pervasive computing vision. The best candidate to determine a user location in indoor environment is by using IEEE 802.11 (Wi-Fi) signals, since it is more and more widely available and installed on most mobile devices used by users. Unfortunately, the signal strength, signals quality and noise of Wi-Fi, in worst scenario, it fluctuates up to 33% because of the reflection, refraction, temperature, humidity, the dynamic environment, etc. The authors present in Chapter 35 their current development on a light-weight algorithm, which is easy, simple but robust in producing the determination of user location using WiFi signals. The algorithm is based on “multiple observers” on ηk-Nearest Neighbour. The authors extend the approach in the estimation indoor-user location by using combination of different technologies, i.e. WiFi, GPS, GSM and Accelerometer. The algorithm is based on opportunistic localization algorithm and fuse different sensor data in order to be able to use the data which is available at the user position and processable in a mobile device. Chapter 36 User Pro-Activities Based on Context History.................................................................................... 558 Teddy Mantoro, International Islamic University Malaysia, Malaysia Media Ayu, International Islamic University Malaysia, Malaysia Context-aware computing is a class of mobile computing that can sense its physical environment and adapt its behavior accordingly; it is a component of the ubiquitous or pervasive computing environment that has become apparent with innovations and challenges. Chapter 36 reviews the concept of contextaware computing, with focus on the user activities that benefit from context history. How user activities in the smart environment can make use of context histories in applications that apply the concept of context prediction integrated with user pro-activity is explored. A brief summary of areas which benefit from these technologies as well as corresponding issues are also investigated. Section 2 Emerging Technologies Chapter 37 Research Challenge of Locally Computed Ubiquitous Data Mining.................................................. 576 Aysegul Cayci, Sabanci University, Turkey João Bártolo Gomes, Universitad Politecnica, Spain Andrea Zanda, Universitad Politecnica, Spain Ernestina Menasalvas, Universitad Politecnica, Spain Santiago Eibe, Universitad Politecnica, Spain

Advances in wireless, sensor, mobile and wearable technologies present new challenges for data mining research on providing mobile applications with intelligence. Autonomy and adaptability requirements are the two most important challenges for data mining in this new environment. In this chapter, the authors analyse the challenges of designing ubiquitous data mining services by examining the issues and problems while paying special attention to context and resource awareness. The authors focused on the autonomous execution of a data mining algorithm and analyzed the situational factors that influence the quality of the result. Already existing solutions in this area and future directions of research are also covered. Chapter 38 Emerging Wireless Networks for Social Applications......................................................................... 595 Raúl Aquino, University of Colima, México Luis Villaseñor, CICESE Research Centre, México Víctor Rangel, National Autonomous University of Mexico, México Miguel García, University of Colima, México Artur Edwards, University of Colima, México Chapter 38 describes the implementation and performance evaluation of a novel routing protocol called Pandora, designed for social applications, and that can be implemented in a broad number of devices, such as commercial wireless routers and laptops. It also provides a robust backbone integrating and sharing data, voice and video between computers and mobile devices. Pandora offers great performance with both fixed and mobile devices and includes important features such as: geographic positioning, residual battery energy monitoring, and bandwidth utilization. In addition, Pandora also considers the number of devices attached to the network. Pandora is experimentally evaluated in a testbed with laptops for the first stage and commercial wireless routers for the second stage. The main goal of Pandora is to provide a reliable backbone for social applications requiring a quality of service (QoS) guarantee. With this in mind, the following evaluation of Pandora considers the following types of traffic sources: transport control protocol (TCP), voice, video and user datagram protocol (UDP) without marks. Pandora is also compared with different queuing disciplines, including: priority queuing discipline (PRIO), hierarchical token bucket (HTB) and DSMARK. Finally, an Internet radio transmission is employed to test the network re-configurability. Results show that queuing the PRIO and HTB disciplines, which prioritizes UDP traffic, performed the best. Chapter 39 An Approach to Mobile Grid Platforms for the Development and Support of Complex Ubiquitous Application........................................................................................................ 617 Carlo Bertolli, University of Pisa, Italy Daniele Buono, University of Pisa, Italy Gabriele Mencagli, University of Pisa, Italy Marco Vanneschi, University of Pisa, Italy Several complex and time-critical applications require the existence of novel distributed, heterogeneous and dynamic platforms composed of a variety of fixed and mobile processing nodes and networks. Such platforms, that can be called Pervasive Mobile Grids, aim to merge the features of Pervasive Comput-

ing and High-performance Grid Computing onto a new emerging paradigm. In this chapter we study a methodology for the design and the development of high-performance, adaptive and context-aware applications. Chapter 39 describes a programming model approach, and the authors compare it with other existing research works in the field of Pervasive Mobile Computing, discussing the rationales of the requirements and the features of a novel programming model for the target platforms and applications. In order to exemplify the proposed methodology, the authors introduce the programming framework ASSISTANT, and provide some interesting future directions in this research field. Chapter 40 Towards a Programming Model for Ubiquitous Computing............................................................... 634 Jorge Barbosa, Universidade do Vale do Rio dos Sinos (Unisinos), Brazil Fabiane Dillenburg, Universidade Federal do Rio Grande do Sul, Brazil Alex Garzão, Universidade do Vale do Rio dos Sinos (Unisinos), Brazil Gustavo Lermen, Universidade do Vale do Rio dos Sinos (Unisinos), Brazil Cristiano Costa, Universidade do Vale do Rio dos Sinos (Unisinos), Brazil Potentially, in the mobile computing scenario, the users can move in different environments and the applications can automatically explore their surroundings. This kind of context-aware application is emerging, but is not yet widely disseminated. Based on perceived context, the application can modify its behavior. This process, in which software modifies itself according to sensed data, is named Adaptation and constitutes the core of Ubiquitous Computing. The ubiquitous computing scenario brings many new problems such as coping with the limited processing power of mobile devices, frequent disconnections, the migration of code and tasks between heterogeneous devices, and others. Current practical approaches to the ubiquitous computing problem usually rely upon traditional computing paradigms conceived back when distributed applications where not a concern. Holoparadigm (Holo) was proposed as a model to support the development of distributed systems. Based on Holo concepts, a new programming language called HoloLanguage (HoloL) was created. The authors propose the use of Holo for developing and executing ubiquitous applications, explore the HoloL for ubiquitous programming and propose a full platform to develop and execute Holo programs. The execution environment is based on a virtual machine that implements the concepts proposed by Holo. Chapter 41 An Agent-Based Operational Virtual Enterprise Framework Enabled by RFID................................. 649 Özgür Ünver, TOBB-University of Economics and Technology, Turkey Bahram Lotfi Sadigh, Middle East Technical University, Turkey The Virtual Enterprise (VE) is a collaboration model between multiple business partners in a value chain that aims to cope with turbulent business environments, mainly characterized by demand unpredictability, shortening product lifecycles, and intense cost pressures. The VE model is particularly viable and applicable for SME and industry parks containing multiple SMEs that have different vertical competencies. As small firms collaborate effectively under VE model, it will be only possible to emerge products by joining their diverse competencies and mitigate the effects of market turbulence by minimizing their investment. A typical VE model has four phases; opportunity capture, formation, operation, and dissolution. The goal of this chapter is to present a conceptual VE framework, focusing

on operation phase, that incorporates Multi Agent Systems (MAS) and Radio Frequency Identification Systems (RFID) which are emerging from research to industry with a great momentum. First, state of the art for VE and the two key enabling technologies are covered in detail. After presenting conceptual view of the framework, an Information and Communication Technology (ICT) view is also given to enhance technical integration with available industry standards and solutions. Finally, process views of how a VE can operate utilizing agent based and RFID systems in order to fulfill operational requirements, are presented. Chapter 42 Ontological Dimensions of Semantic Mobile Web 2.0: First Principles............................................. 667 Gonzalo Aranda-Corral, University of Sevilla, Spain Joaquín Borrego-Díaz, University of Sevilla, Spain Chapter 42 advances, from the point of view of Knowledge Representation and Reasoning, an analysis of which ontological dimensions are needed to develop Mobile Web 2.0 on top of Semantic Web. This analysis will be particularly focused on social networks and it will try to make an outlook about the new knowledge challenges on this field. Some of these new challenges will be linked to Semantic Web context, while others will be inherent to Semantic Mobile Web 2.0. Chapter 43 Unobtrusive Interaction with Mobile and Ubiquitous Computing Systems through Kinetic User Interfaces..................................................................................................................................... 689 Vincenzo Pallotta, Webster University, Switzerland Unobtrusiveness is a key factor in usability of mobile and ubiquitous computing systems. These systems are made of several ambient and mobile devices whose goal is supporting everyday life users’ activities, hopefully without interfering with them. The chapter addresses the topic of obtrusiveness by assessing its impact in the design of interfaces for mobile and ubiquitous computing systems. Authors discuss how unobtrusive interfaces can be designed by means of Kinetic User Interfaces: an emerging interaction paradigm where input to system is provided through coordinated motion of objects and people in the physical space. Chapter 44 Impact of Advances on Computing and Communication Systems in Automotive Testing................. 703 Luís Serrano, Polithecnic Institute Leiria, Portugal José Costa, University Coimbra, Portugal Manuel Silva, University Coimbra, Portugal A huge amount of information is used nowadays by modern vehicles, and it may be accessed through an On Board Diagnosis (OBD) connection. A technique using the already installed OBD system to communicate with the vehicle together with a Global Position System (GPS) provides reliable data which allow a detailed analysis of real on road tests. Different kinds of circulation circuits (urban, extra-urban and highway) were analyzed in Chapter 44, using the capabilities of OBD II installed on the tested vehicles. OBD provides an important set of information, namely related with data on the engine, fuel

consumption, chassis and auxiliary systems and also on combustion efficiency. The use of GPS in all the road tests performed provides important information to further determine the more sustainable from all the different solutions tested, considering the different situations imposed on each circuit. It is a fact that bench tests or a chassis dynamometer allow a fine control of the operation conditions; however the simulation is not as real as on the road. So the present methodology will allow the possibility to perform tests on the road, allowing enough control on vehicles and providing complete information of the chosen route and of the trip history. This is a possibility that ensures new tools with more reliable data which can give faster answers for the development of high efficiency, economic and environmentally neutral automotive technologies. Chapter 45 RFID and NFC in the Future of Mobile Computing........................................................................... 719 Diogo Simões, Movensis, Portugal Vitor Rodrigues, Movensis, Portugal Luis Veiga, INESC ID / Technical University of Lisbon, Portugal Paulo Ferreira, INESC ID / Technical University of Lisbon, Portugal RFID (Radio Frequency Identification) technology consists of a tag that can be used to identify an animal, a person or a product, and a device responsible for transmitting, receiving and decoding the radio waves. RFID tags work in two different modes: they wake up when they receive a radio wave signal and reflect it (Passive Mode) or they emit their own signal (Active Mode). The tags store information which allows univocally identifying something or someone. That information is stored in an IC (Integrated Circuit) which is connected to an antenna, responsible for transmitting the information. An evolution of this technology is the Near Field Communication (NFC). It consists of a contactless Smart Card technology, based in short-range RFID. Currently, there are mobile phones with NFC embedded in such a way that they work both as a tag and as a NFC reader. These technologies will be widely available both in mobile phones and other devices (e.g. personal digital assistants, etc.) in the near future allowing us to get closer to a ubiquitous and pervasive world. This chapter describes the most important aspects of RFID and NFC technology, illustrating their applicative potential, and provides a vision of the future in which the virtual and real worlds merge together as if an osmosis took place. Chapter 46 A Multi-Loop Development Process for a Wearable Computing System in Autonomous Logistics......................................................................................................................... 738 Jakub Piotrowski, Bremer Institut für Produktion und Logistik, Germany Carmen Ruthenbeck, Bremer Institut für Produktion und Logistik, Germany Florian Harjes, Bremer Institut für Produktion und Logistik, Germany Bernd Scholz-Reiter, Bremer Institut für Produktion und Logistik, Germany Chapter 46 examines a multi-loop development process for a wearable computing system within a new paradigm in logistic applications. The implementation of this system will be demonstrated by an example from the field of autonomous logistics for automobile logistics. The development process is depicted from selecting and combining hardware through to the adjustment to both user and operative environment. Further, this chapter discusses critical success factors like robustness and flexibility. The objective is to present problems and challenges as well as a possible approach to cope with them.

Section 3 Critical Success Factors Chapter 47 Collaboration within Social Dimension of Computing: Theoretical Background, Empirical Findings and Practical Development.................................................................................. 760 Andreas Ahrens, University of Technology, Business and Design, Germany Jeļena Zaščerinska, University of Latvia, Latvia Olaf Bassus, University of Technology, Business and Design, Germany A proper development of computing which penetrate our society more thoroughly with the availability of broadband services is provided by varied cooperative networks. However, the success of social dimension of computing requires collaboration within a multicultural environment to be considered. Aim of Chapter 47 is to analyze collaboration within the social dimension of computing on the pedagogical discourse. The meaning of the key concepts of social dimension of computing, collaboration and its factors is studied within the search for the success of social dimension. The manuscript introduces the study conducted within the Baltic Summer School Technical Informatics and Information Technology in 2009. The conducted explorative research comprises four stages: exploration of the contexts of collaboration, analysis of the students’ needs (content analysis), data processing, analysis and data interpretation, and analysis of the results and elaboration of conclusions and hypothesis for further studies. Chapter 48 Critical Factors in Defining the Mobile Learning Model: An Innovative Process for Hybrid Learning at the Tecnologico de Monterrey, a Mexican University......................................... 774 Violeta Chirino-Barceló, Tecnologico de Monterrey Mexico, Mexico Arturo Molina, Tecnologico de Monterrey Mexico, Mexico Many factors converge when attempting to define the most adequate mobile learning model to be applied in a face-to-face university environment. As far as innovation related processes go, the implementation of mobile learning, implies defining a road map on the basis of strategic planning. It is also important to apply an action research approach in the implementation process of the model. In analyzing in depth this innovative mobile learning process, there are key factors to consider. First, there are factors related to the technology necessary for the implementation of the model—both hard and soft requirements. Second, there are cultural issues related to the use of non-native internet professors of innovative technologies. Finally, there are challenges related to defining, exactly, those educational strategies to be handled through mobile devices. Chapter 48 focuses on the critical factors involved in integrating mobile learning into a hybrid educational model at a Mexican university. Chapter 49 Critical Human Factors on Mobile Applications for Tourism and Entertainment............................... 793 Pedro Campos, University of Madeira, Portugal The purpose of Chapter 49 is to research some principles that can guide the design, development and marketing of mobile applications, with a particular focus on the tourism and entertainment applica-

tion domains. This research also fills a gap concerning impact studies of mobile applications, since the majority of the literature available today is more focused on the design and development process and results. Besides describing a set of novel mobile applications, we aim at providing an overview on innovation processed used, and conducting several experiences, gathering results from questionnaires, surveys, log data and our own observations. Regarding the mobile tourism domain, the author studied the impact of media visibility, the impact of novel interaction paradigms. Regarding the mobile entertainment applications, he focused on studying the impact brought that realism and graphics quality have on mobile games. Chapter 50 Internet Surveys: Opportunities and Challenges.................................................................................. 805 Paula Vicente, UNIDE, ISCTE – Lisbon University Institute, Portugal Elizabeth Reis, UNIDE, ISCTE – Lisbon University Institute, Portugal Internet surveys offer important advantages over traditional survey methods: they can accomplish large samples within a relatively short period of time, questionnaires may have visual and navigational functionalities impossible to implement in paper-and-pencil questionnaires, data is more efficiently processed since it already comes in electronic format and costs can be lower. But the use of the Internet for survey purposes raises important concerns related to population coverage, lack of suitable sampling frames and non-response. Despite its problems, Internet-based surveys are growing and will continue to expand presenting researchers with the challenge of finding the best way to adapt the methods and principles established in survey methodology to this new mode of data collection in order to make best use of it. Chapter 50 describes the positive features of the Internet for survey activity and examines some of the challenges of conducting surveys via the Internet by looking at methodological issues such as coverage, sample selection, non-response and data quality. Chapter 51 Evolvable Production Systems: A Coalition-Based Production Approach.......................................... 821 Marcus Bjelkemyr, The Royal Institute of Technology, Sweden Antonio Maffei, The Royal Institute of Technology, Sweden Mauro Onori, The Royal Institute of Technology, Sweden The purpose of Chapter 51 is to provide broad view of the rationale, fundamental principles, current developments and applications for Evolvable Production Systems (EPS). Special attention is given to how complexity is handled, the use of agent based and wireless technology, and how economical issues are affected by having an evolvable system. The rationale for EPS is based on current road mapping efforts, which have clearly underlined that true industrial sustainability requires far higher levels of system autonomy and adaptivity than what can be achieved within current production system paradigms. Since its inception in 2002 as a next generation of production systems, the EPS concept has been further developed and tested to emerge as a production system paradigm with technological solutions and mechanisms that support sustainability. Technically, EPS is based on the idea of using several re-configurable, process-oriented, agent-based and wireless intelligent modules of low granularity. This allows for a continuous adaption and evolution of the production system and the ability to explore emergent behavior of the system, which are imperative to remain fit with regards to the system environment.

Section 4 New Business Models Chapter 52 Viable Business Models for M-Commerce: The Key Components..................................................... 837 Jiaxiang Gan, University of Auckland, New Zealand Jairo Gutierrez, Universidad Tecnológica de Bolívar, Colombia As mobile applications increase in popularity, the issue of how to build viable business models for the m-commerce industry is becoming a clear priority for both organizations and researchers. In order to address this issue, this chapter reports on five mini cases used as a guideline, and applies the theoretical business model from Chesbrough and Rosenbloom (2002) to each of them to find out the most important components of viable business models for their m-commerce applications. The study then uses cross cases analysis as a research tool to compare and contrast each of the mini cases and to find out how the different organizations fit within the researched theoretical business model. Finally, this chapter confirms that there are 7 important components of viable business models for m-commerce which are: value proposition, market segment, value chain, profit potential, value network, competitive strategy and firm capabilities. This study also highlights the fact that the public visibility of these 7 components is uneven. Some components such as value proposition, value chain, value network and firm’s capabilities are more likely to be presented in public by organizations. However, aspects such as cost structure and profit potential, market segment and competitive strategy are more likely to be hidden from the public due to their commercial sensitivity. Chapter 53 A Service-Based Framework to Model Mobile Enterprise Architectures........................................... 853 Jose Delgado, Instituto Superior Técnico, Portugal Mobility is a relatively recent topic in the enterprise arena, but thanks to the widespread use of cell phones it has already changed much of the business landscape. It should be integrated in enterprise architectures (EAs) as an intrinsic feature and not as an add-on or as an afterthought transition. Current EA frameworks were not designed with mobility in mind and are usually based on the process paradigm, emphasizing functionality. Although the issue of establishing a systematized migration path from a non-mobile EA to a mobile one has already been tackled, the need for mobile-native EA modeling frameworks is still felt. Chapter 53 presents and discusses a resource-based and service-oriented metamodel and EA framework, in which mobility is introduced naturally from scratch, constituting the basis for some guidelines on which EA resources should be mobilized. Several simple scenarios are presented in the light of this metamodel and framework. Chapter 54 Research-Based Insights Inform Change in IBM M-Learning Strategy............................................. 871 Nabeel Ahmad, IBM Center for Advanced Learning, USA Although mobile phones have become an extension of the workplace, organizations are still exploring their effectiveness for employee training and development. A 2009 joint collaborative study between

Columbia University (New York, USA) and IBM of 400 IBM employees’ use of mobile phones revealed unexpected insights into how employees use mobile applications to improve job performance. The findings are reshaping IBM Learning’s mobile technologies strategy for networking, collaboration, and skills improvement. This chapter reveals the study’s results and IBM’s new direction for mlearning, highlighting IBM’s preparedness for a shift in its organizational learning model potentiated by ubiquitous access and mobility. Chapter 55 Location Based E-Commerce System: An Architecture...................................................................... 881 Nuno André Osório Liberato, UTAD, Portugal João Eduardo Quintela Alves de Sousa Varajão, UTAD, Portugal Emanuel Soares Peres Correia, UTAD, Portugal Maximino Esteves Correia Bessa, UTAD, Portugal Location-based mobile services (LBMS) are at present an ever growing trend, as found in the latest and most popular mobile applications launched. They are, indeed, supported by the hasty evolution of mobile devices capabilities; by users demand; and, lastly, by market drive. With e-commerce, products and services started arriving to potential customers through desktop computers, where they can be bought and fast delivered to a given address. However, expressions such as “being mobile”, “always connected”, “anytime anywhere” that already characterize life in the present will certainly continue to do so in the near future. Meanwhile, mobile devices centred commerce services seem to be the next step. Therefore, this chapter presents a system architecture designed for location-based e-commerce systems. These systems, where location plays the most important role, enable a remote products/services search, based in user parameters: after a product search, shops with that products are returned in the search results and are displayed in a map, around the user present location; and services like obtaining more information, reserving and purchasing are made available as well. This concept represents a mix between traditional client-oriented commerce and faceless mass-oriented e-commerce, enabling a proximity-based user-contextualized system, being well capable of conveying significant advantages and facilities to both service-providers/retailers and users. Section 5 Security Chapter 56 Overview of Security Issues in Vehicular Ad-Hoc Networks.............................................................. 894 José María De Fuentes, Carlos III University of Madrid, Spain Ana Isabel González-Tablas, Carlos III University of Madrid, Spain Arturo Ribagorda, Carlos III University of Madrid, Spain Vehicular ad-hoc networks (VANETs) are a promising communication scenario. Several new applications are envisioned, which will improve traffic management and safety. Nevertheless, those applications have stringent security requirements, as they affect road traffic safety. Moreover, VANETs

face several security threats. As VANETs present some unique features (e.g. high mobility of nodes, geographic extension, etc.) traditional security mechanisms are not always suitable. Because of that, a plethora of research contributions have been presented so far. This chapter aims to describe and analyze the most representative VANET security developments. Chapter 57 Modelling of Location-Aware Access Control Rules.......................................................................... 912 Michael Decker, Karlsruhe Institute of Technology, Germany Access control in the domain of information system security refers to the process of deciding whether a particular request made by a user to perform a particular operation on a particular object under the control of the system should be allowed or denied. For example, the access control component of a file server might have to decide whether user “Alice” is allowed to perform the operation “delete” on the object “document.txt”. For traditional access control this decision is based on the evaluation of the identity of the user and attributes of the object. The novel idea of location-aware access control is also to consider the user’s current location which is determined by a location system like GPS. The main purpose of Chapter 57 is to present several approaches for the modeling of location-aware access control rules. Chapter 58 Secure Techniques for Remote Reconfiguration of Wireless Embedded Systems.............................. 930 Abdellah Touhafi, Vrije Universiteit Brussel, Belgium An Braeken, Erasmushogeschool Brussel, Belgium Gianluca Cornetta, Universidad San Pablo-CEU, Spain Nele Mentens, Katholieke Universiteit Leuven, Belgium Kris Steenhaut, Vrije Universiteit Brussel, Belgium The aim of Chapter 58 is to give a thorough overview of secure remote reconfiguration technologies for wireless embedded systems, and of the communication standard commonly used in those systems. In particular, authors focus on basic security mechanisms both at hardware and protocol level. We will discuss the possible threats and their corresponding impact level. Different countermeasures for avoiding these security issues are explained. Finally, the chapter presents a complete and compact solution for a service-oriented architecture enabling secure remote reconfiguration of wireless embedded systems, called the STRES system. Chapter 59 Secure Routing and Mobility in Future IP Networks.......................................................................... 952 Kaj Grahn, Arcada University of Applied Sciences, Finland Jonny Karlsson, Arcada University of Applied Sciences, Finland Göran Pulkkis, Arcada University of Applied Sciences, Finland The evolution of computer networking is moving from static wired networking towards wireless, mobile, infrastructureless, and ubiquitous networking. In next-generation computer networks, new mobility features such as, seamless roaming, vertical handover, and moving networks are introduced. Secu-

rity is a major challenge in developing mobile and infrastructureless networks. Specific security threats in next-generation networks are related to the wireless access mediums, routing, and mobility features. Chapter 59 identifies these threats, and discuss the state of the art of security research and standardization within the area, proposing security architectures for mobile networking. A survey of security in routing is provided with special focus on mobile ad hoc networks (MANETs). The security of currently relevant protocols for management or node and network mobility, Mobile IP (MIP), Network Mobility (NEMO), Mobile Internet Key Exchange (MOBIKE), Host Identity Protocol (HIP), Mobile Stream Control Transmission Protocol (mSCTP), Datagram Congestion Control Protocol (DCCP), and Session Initiation Protocol (SIP), is described. Section 6 Applications, Surveys and Case Studies Chapter 60 Evaluation of a Mobile Platform to Support Collaborative Learning: Case Study............................. 974 Carlos Quental, Polytechnic Institute of Viseu, Portugal Luis Gouveia, University Fernando Pessoa, Portugal In an educational context, technological applications and their supporting infrastructures have been evolved in a way that the use of learning objects is no longer limited to a personal computer, but has been extended to a number of mobile devices. Such evolution leads to the creation of a technological model called m-learning that offers great benefits to education. This educational model has been developed over the recent years, which resulted in several research projects and some commercial products. This chapter describes the (re)use of an adapted platform from an API of MLE (Mobile Learning Engine), to create tests, quizzes, forums, SMS, audio, video, mobile learning objects, in combination with a learning platform in a particular setting. MLE is a special m-learning application for mobile phones (a J2ME application) that can access a LMS (Learning Management System) and use most of its activities and resources, and add new, even innovative, activities. With J2ME one can store, use content and learn without the need of further network access and even use interactive questions that can be directly solved on mobile devices. The MLE enables one to use the mobile phone as a constant way of learning. As a consequence it is possible to use every spare time to learn, no matter where we are, providing new opportunities to enhance learning. Chapter 61 Power Issues and Energy Scavenging in Mobile Wireless Ad-Hoc and Sensor Networks................. 994 Gianluca Cornetta, Universidad San Pablo-CEU, Spain Abdellah Touhafi, Erasmushogheschool Brussel, Belgium David J. Santos, Universidad San Pablo-CEU, Spain José Manuel Vázquez, Universidad San Pablo-CEU, Spain Wireless ad-hoc and sensor networks are experiencing a widespread diffusion due to their flexibility and broad range of potential uses. Nowadays they are the underlying core technology of many industrial and remote sensing applications. Such networks rely on battery-operated nodes with a limited

lifetime. Although, in the last decade, a significant research effort has been carried out to improve the energy efficiency and the power consumption of the sensor nodes, new power sources have to be considered to improve node lifetime and to guarantee a high network reliability and availability. Energy scavenging is the process by which the energy derived from external sources is captured, translated into an electric charge and stored internally to a node. At the moment, these new power sources are not intended to replace the batteries, since they cannot generate enough energy; however, working together with the conventional power sources they can significantly improve node lifetime. Low-power operation is the result of a complex cross-layer optimization process, for this reason, this chapter thoroughly reviews all the traditional methods aimed at reducing power consumption at network, MAC and PHY levels of the TCP stack, to understand advantages and limitations of such techniques, and to justify the need of alternative power sources that may allow, in the future, the design of completely self-sustained and autonomous sensor nodes. Chapter 62 A Low Cost Wireless Sensors Network with Low-Complexity and Fast-Prototyping...................... 1021 João Carmo, University of Minho, Portugal José Correia, University of Minho, Portugal This chapter presents a low cost/fast prototyping wireless sensors network that was designed for a huge range of applications and making use of low cost commercial of the shelf components. Such applications includes industrial measurements, biomedical, domestic monitoring, remote sensing, among others. The concept of the wireless sensor network is presented and simultaneously, hot topics and their implementation are discussed. Such topics are valuable tools and can’t be discarded when a wireless sensors network is planed. By the contrary, such tools must be taken in account to make the communications between the nodes and the base station the best possible reliable. The architecture, protocols and the reasons that were behind the selection of the components are also discussed. The chapter also presents performance metrics that are related to with the physical characteristics of sensors and with the radio specificity. Microcontrollers with a RISC architecture are used by the network nodes to control the communication and the data acquisition and operate in the 433 MHz ISM band with ASK modulation. Also, in order to improve the communication and to minimise the lost of data, it is predicted to put the wireless nodes to handle line and source coding schemes. Chapter 63 Unreliable Failure Detectors for Mobile Ad-hoc Networks.............................................................. 1039 Luciana Arantes, University Paris 6, France Fabiola Greve, Federal University of Bahia, Brazil Pierre Sens, University Paris 6, France Failure detection is an important abstraction for the development of fault-tolerant middleware, such as group communication toolkits, replication and transaction services. Unreliable failure detector (namely, FD) can be seen as an oracle which provides information about process failures. The dynamics and self-organization of Mobile Ad-hoc Networks (MANETs) introduce new restrictions and challenges for the implementation of FDs with which static traditional networks do not have to cope. It is worth mentioning that in some way, fault tolerance is more critical for MANETs than for the latter, since wireless

network can present high error rates and mobile nodes are more prone to failures, physical damages or transient disconnections. The aim of this chapter is thus to discuss the impact of all these characteristics, intrinsic to MANET, in the implementation of FDs. It presents a survey of the few works about FD implementations for wireless networks, including the different possible assumptions to overcome the dynamics and lack of both global view and synchrony of MANETs. Chapter 64 Mission-Aware Adaptive Communication for Collaborative Mobile Entities................................... 1056 Jérôme Lacouture, Université de Toulouse, France Ismael Bouassida Rodriguez, Université de Toulouse, France Jean-Paul Arcangeli, Paul Sabatier University, France Christophe Chassot, Université de Toulouse, France Thierry Desprats, Paul Sabatier University, France Khalil Drira, Université de Toulouse, France Francisco Garijo, Université de Toulouse, France Victor Noel, Paul Sabatier University, France Michelle Sibilla, Paul Sabatier University, France Catherine Tessier, ONERA Centre de Toulouse – DCSD, France Adaptation of communication is needed to maintain the connectivity and quality of communication in group-wide collaborative activities. This becomes quite a challenge to handle when mobile entities are part of a wireless environment, in which responsiveness and availability of the communication system are required. In this chapter, these challenges are addressed within the context of the ROSACE project where mobile ground and flying robots have to collaborate either between themselves or with remote artificial and human actors during save and rescue missions in the event of disasters such as forest fires. This chapter presents the first results. The final goal is to propose new concepts, models and architectures that supports cooperative adaptation which is aware of the mission being executed. Thus, the communication system can be adequately adapted in response to predictable or unpredictable evolutions of the activity requirements and to the unpredictable changes in the communication resource constraints. Chapter 65 OntoHealth: An Ontology Applied to Pervasive Hospital Environments.......................................... 1077 Giovani Librelotto, Federal University of Santa Maria, Brazil Iara Augustin, Federal University of Santa Maria, Brazil Jonas Gassen, Federal University of Santa Maria, Brazil Guilherme Kurtz, Federal University of Santa Maria, Brazil Leandro Freitas, Federal University of Santa Maria, Brazil Ricardo Martini, Federal University of Santa Maria, Brazil Renato Azevedo, Federal University of Santa Maria, Brazil In the last years ontologies are being used in the development of pervasive computing applications. It is habitual their use for facilitating the interoperability among context-aware applications and the entities that may enter in the context at any time. This chapter presents OntoHealth: an ontology applied to health pervasive environment and a tool to its processing. The main idea is that a hospital could be

seen as this pervasive environment, where someone, through ubiquitous computing, engages a range of computational devices and systems simultaneously, in the course of ordinary activities, and may not necessarily even be aware that they are doing so. With the proposed ontology and the tool for its processing, the medical tasks can be shared by all components of this pervasive environment. Chapter 66 Adoption of Mobile and Information Technology in an Energy Utility in Brazil............................. 1091 Osvaldo Garcia, Pontificia Universidade Católica do PR, Brazil Maria Cunha, Pontificia Universidade Católica do PR, Brazil This chapter deals with the adoption of mobile technology. The case illustrated here is the implementation of mobile and wireless technology – MIT and smartphones – at an energy utility. The objective was to understand the human and social aspects of the adoption of this technology. This paper makes use of the metaphor of hospitality proposed by Ciborra in the late 1990s. The hospitality metaphor was a useful alternative for describing the process of adopting a new technology. It touches on technical aspects and notes human reactions that become evident when a technician comes across an unknown ‘guest’, the new technology: the doubtful character of the guest, the reinterpretation of the identities of guest and host during the process, learning through trial and error, the technology’s ‘drift’, the participants’ emotions and state of mind, and the appropriation of, and the care for, the new technology. Chapter 67 Infrastructures for Development of Context-Aware Mobile Applications......................................... 1104 Hugo Feitosa de Figueirêdo, University of Campina Grande, Brazil Tiago Eduardo da Silva, University of Campina Grande, Brazil Anselmo Cardoso de Paiva, University of Maranhão, Brazil José Eustáquio Rangel de Queiroz, University of Campina Grande, Brazil Cláudio De Souza Baptista, University of Campina Grande, Brazil Context-aware mobile applications are becoming popular, as a consequence of the technological advances in mobile devices, sensors and wireless networking. Nevertheless, developing a context-aware system involves several challenges. For example, what will be the contextual information, how to represent, acquire and process this information and how it will be used by the system. Some frameworks and middleware have been proposed in the literature to help programmers to overcome these challenges. Most of the proposed solutions, however, neither have an extensible ontology-based context model nor uses a communication method that allows a better use of the potentialities of the models of this kind. Chapter 68 A Practice Perspective on Transforming Mobile Work..................................................................... 1119 Riikka Vuokko, Åbo Akademi University, Finland Chapter 68 introduces a study that explores users’ experiences during an organizational implementation of a new mobile information technology in public home care environment. The home care case illus-

trates differences between implementation project goals and expectations, and on the other hand, the daily organizing and carrying out care work, where previously, no information technology was utilized. While implementing mobile technology was expected to enhance the efficiency of care working, the project outcomes include resistance due to surveillance aspect of the new technology as well as technological problems during the implementation. Successful outcomes of the implementation include better planning of working hours and more even distribution of work resources. Chapter 69 Data Replication Support for Collaboration in Mobile and Ubiquitous Computing Environments.......... 1132 João Barreto, INESC-ID/Technical University Lisbon, Portugal Paulo Ferreira, INESC-ID/Technical University Lisbon, Portugal In Chapter 69, the authors address techniques to improve the productivity of collaborative users by supporting highly available data sharing in poorly connected environments such as ubiquitous and mobile computing environments. Authors focus on optimistic replication, a well known technique to attain such a goal. However, the poor connectivity of such environments and the resource limitations of the equipments used are crucial obstacles to useful and effective optimistic replication. The authors analyze state-of-the art solutions, discussing their strengths and weaknesses along three main effectiveness dimensions: (i) faster strong consistency, (ii) with less aborted work, while (iii) minimizing both the amount of data exchanged between and stored at replicas; and identify open research issues Chapter 70 Providing Outdoor and Indoor Ubiquity with WLANs..................................................................... 1155 Diana Bri, Polytechnic University of Valencia, Spain Hugo Coll, Polytechnic University of Valencia, Spain Sandra Sendra, Polytechnic University of Valencia, Spain Jaime Lloret, Polytechnic University of Valencia, Spain Wireless Local Area Networks are very useful for the most applications based on network. It can be developed in almost all environments and products are cheap and robust. Moreover, these networks can be formed by different devices with wireless interfaces like IP cameras, laptops, PDAs, sensors, etc. WLANs provide high bandwidth at large coverage areas, which it is necessary in many applications at different research areas. All these characteristics let WLANs be a useful technology to provide ubiquity for any type of service. If they are deployed from a good and exhaustive design, they can provide connection to any device, everywhere at anytime. In this paper we present a complete guideline about how to design and deploy WLANs and to get their best performance. Authors start from an analytical point of view and use mathematical expressions to design WLANs in both indoor and outdoor environments. Then, the authors introduce a method proposed by some authors of this chapter some years ago and how it can be used to design WLANs in indoor environments. The chapter also presents WLANs design in outdoor environments and describe two projects developed in order to provide ubiquity in real indoor and outdoor environments.

Chapter 71 INTIC for Mobile Devices: Support System for Communication with Mobile Devices for the Disabled.................................................................................................................................. 1169 Cristina Diaz Busch, University of A Coruna, Spain Alberto Moreiras Lorenzo, University of A Coruna, Spain Iván Mourelos Sánchez, University of A Coruna, Spain Betania Groba González, University of A Coruna, Spain Thais Pousada García, University of A Coruna, Spain Laura Nieto Riveiro, University of A Coruna, Spain Javier Pereira Loureiro, University of A Coruna, Spain The In-TIC system for mobile devices (in Spanish: Integration with Information and Communication Technologies system for mobile devices) represents an approach towards the area of technical aids for mobile devices. The mobile telephone is a device that makes our lives easier, allowing us to be permanently accessible and in contact, to save relevant information, and also for entertainment purposes. However, people with visual, auditory or motor impairment or the elderly still find these devices difficult to use. They have to overcome a range of difficulties when using mobile telephones: the screens are difficult to read, the buttons are too small to use, and the technical features are too complicated to understand. At present, the main advances in mobile technology have been aimed at improving multimedia-messaging services and reproducing videos and music. This new support system adds accessibility to mobile telephones, making them easier to use for the people who need them the most, people with reduced physical or mental capacities who cannot use a conventional mobile. Chapter 72 New Ways to Buy and Sell: An Information Management Web System for the Commercialization of Agricultural Products from Family Farms without Intermediaries................ 1182 Carlos Ferrás, University of Santiago Compostela, Spain Yolanda García, University of Santiago Compostela, Spain Mariña Pose, University of Santiago Compostela, Spain Granxafamiliar.com is a project for developing the Galician rural milieu both socio-economically and culturally in order to appreciate the quality of life and rural culture, to create communication links between the rural and urban world, to emphasize the importance of the traditional self-supply production market of Galician family farms, and to promote the spread of new technologies as a social intervention tool against the phenomenon of social and territorial exclusion known as the “Digital Divide”. The authors are planning the architecture of www.granxafamiliar.com, which is developing the creation of a virtual community based on boosting commercial transactions and the possibilities for buying and selling traditional self-supply products that exist in the rural environment. The authors expect to promote it globally across the Internet by promoting the use and spread of ICTs as tools and commercial channels for agricultural products and intend to carry out an in-depth empirical and theoretical study of the territorial and social effects linked to the development of the information and communication society in rural communities. Their aim is to assist the progress of public decision-making and administrative efficiency for when the time comes to invest in suitable services and activities related to the information society in the rural environment.

Chapter 73 Broadcast Quality Video Contribution in Mobility........................................................................... 1199 José Ramón Cerquides Bueno, University of Seville, Spain Antonio Foncubierta Rodriguez, University of Seville, Spain The continuous growth of the available throughput, especially in the uplink of mobile phone networks is opening the doors to new services and business opportunities without references in the past. In more concrete, new standards HSDPA/HSUPA, introduced to complement and enhance 3G networks, together with the advances in audio and especially video coding, like those adopted by the standard H.264 AVC have boosted the appearance of a new service: exploiting the mobile telephony networks for contributing broadcast quality videos. This new service is offering just now a low cost, high flexibility alternative that, in a brief period of time, will substitute the current Electronic News Gathering (ENG) Units, giving rise to what is being to be called Wireless Journalism (WENG or WiNG). This chapter discusses both the technologies involved and the business opportunities offered by this sector. Once reviewed the state of the art, different solutions will be compared, some of them recently appeared as commercial solutions, like QuickLink 3.5G Live Encoder or AirNow! and others still in research and development processes. Chapter 74 Mobile Device Selection in Higher Education: iPhone vs. iPod Touch............................................ 1213 C. Brad Crisp, Abilene Christian University, USA Michael Williams, Pepperdine University, USA Mobile devices are rapidly becoming the most common interface for accessing network resources (Hall 2008). By 2015 the average 18-year old will spend the majority of their computing time on mobile devices (Basso 2009). These trends directly affect institutions of higher learning. Many universities are offering learning initiatives and m-services designed to distribute content and services to mobile devices. Chapter 74 reports findings from an exploratory, longitudinal study at Abilene Christian University, where incoming freshmen received their choice of an Apple iPhone or iPod touch. The findings indicate that users’ device selections were affected by their perceptions of the costs of the devices, the devices’ relative characteristics, and the social influence of parents. We also found that users’ attitude, satisfaction, and confidence about their device selection varied across devices, with iPhone users having more favorable perceptions. The chapter concludes with recommendations for mobile learning initiatives and directions for future research. Chapter 75 Design of Wearable Computing Systems for Future Industrial Environments.................................. 1226 Pierre Kirisci, Bremer Institut für Produktion und Logistik GmbH, Germany Ernesto Morales Kluge, Bremer Institut für Produktion und Logistik GmbH, Germany Emanuel Angelescu, Bremer Institut für Produktion und Logistik GmbH, Germany Klaus-Dieter Thoben, Bremer Institut für Produktion und Logistik GmbH, Germany This chapter investigates the role of context, particularly in future industrial environments, and elaborates how context can be incorporated in a design method in order to support the design process of

wearable computing systems. The chapter is initiated by an overview of basic research in the area of context-aware mobile computing. The aim is to identify the main context elements which have an impact upon the technical properties of a wearable computing system. Therefore the authors describe a systematic and quantitative study of the advantages of context recognition, specifically task tracking, for a wearable maintenance assistance system. Based upon the experiences from this study, a context reference model is proposed, which can be considered supportive for the design of wearable computing systems in industrial settings, thus goes beyond existing context models, e.g. for context-aware mobile computing. The final part of this chapter discusses the benefits of applying model-based approaches during the early design stages of wearable computing systems. Existing design methods in the area of wearable computing are critically examined and their shortcomings highlighted. Based upon the context reference model, a design approach is proposed through the realization of a model-driven software tool which supports the design process of a wearable computing system while taking advantage of concise experience manifested in a well-defined context model. Chapter 76 Extending the Scope of eID Technology: Threats and Opportunities in a Commercial Setting........ 1246 Vincent Naessens, Katholieke Hogeschool Sint-Lieven, Belgium Bart De Decker, Katholieke Universiteit Leuven, Belgium In 2002, Belgium has adopted an electronic identity card as one of the first countries in Europe. By the end of 2009, the roll-out of the eID card will be completed, and this means that each Belgian citizen will possess an eID card. The card enables her to digitally prove her identity and to legally sign electronic documents. The Belgian eID card opens up new opportunities for the government, its citizens, service providers and application developers. The Belgian eID technology originally aimed at facilitating transactions between Belgian citizens and the government and although many eID applications have been developed, the success of the Belgian eID technology has not been what was expected. Therefore, the Belgian government encourages developers to build commercial applications that use the eID card (for authentication or e-signatures). However, extending the scope of the Belgian eID technology from egovernment to the commercial sector is no sinecure and not without risks. Chapter 77 Mobility and Connectivity: On the Character of Mobile Information Work..................................... 1262 Victor Gonzalez, University of Manchester, UK Antonis Demetriou, University of Manchester, UK Mobile information work, an extreme type of information work, is progressively becoming commonplace in various corporations. The availability of cheap and portable information technologies as well as the development of pervasive communication infrastructure in some parts of the world is creating scenarios where people can work from almost anyplace. Nevertheless up to now there has not been sufficient research on the particular work practices and strategies these professional workers use to be productive as they face the particular challenges of being mobile. Based on an ethnographic investigation of the experiences of mobile professional workers in a multi-national accountancy company, this chapter discusses some characteristics defining the character of modern information work with regards

mobility and connectivity while operating outside the workplace. The study highlights the importance of: location in terms of providing an adequate atmosphere and infrastructure to conduct work; regularity in terms of giving workers flexibility to connect and reconnect whenever it was more convenient for them; space in terms of letting people preserve and reconstruct their information workspaces; and balance while juggling between personal and work related commitments. The findings presented can be useful for defining the processes and technological tools supporting mobile professional workers.

lx

Foreword

Mobility has become the next big challenge in our inter-connected world. Indeed, after the Internet widespread dissemination, users are eager for information anytime anywhere. Although ubiquity brought the power of computing anywhere, it is the access to global information that enriches the usefulness of mobile computing. From an historical perspective, the Internet widespread dissemination enabled the access to huge amounts of free information. This information is accessible from any pc, provided it is connected to a network. Naturally, the users that were previously restricted to a desktop pc now aspire to have the same access to information from everywhere. Acessing information anytime, everywhere requires a mobile computing environment. To the development of a mobile computing environment contributed the miniaturization of devices and the improvement on wireless technologies. From the miniaturization of devices spurred Smartphones, devices that merge the mobile phone with the pocket pc concepts, and Netbooks, small laptops that favor (wireless) communications to raw computing power. On the other hand, the development of carrier technologies with increased network bandwidth popularized broadband access either at home (through cable) or on mobile devices (through wireless). We now live in a highly inter-connected world, made of several networks, available from a multitude of wireless technologies that enable a mobile behavior (consider for example WiFi, WiMAX, UMTS, HSDPA).

CHALLENGES AND OPPORTUNITIES AVAILABLE However, (hardware) technology by itself does not suffice to harness the challenges present on a mobile computing environment. It is necessary to consider how technology should be used by applications. The applications for the mobile environment share the same communication model but also the same limitations found in Internet applications. There is an overwhelming range of distributed applications available for distributed systems in general, and for the Internet in particular, that can be brought to the mobile environment. However, the mobile environment presents some specific challenges that are not common for a generic distributed system like the Internet. Device and service availability is an important property of a mobile system since some devices may come and go from the system, rather than being permanently available like the typical server device. A consequence of the absence of permanent availability is the absence of data persistency. Where should

lxi

data be stored? After storing data, possibly in more than a single place, synchronization is another important property. Is the data being accessed the most up-to-date? When looking at the Web evolution, it is possible to clearly identify two phases: the first is about accessing information; and the second is about accessing (web) services. Like the Web, the mobile environment is perfectly capable of offering remote services. Joining services with mobility opens up new market opportunities. For example, geographic based advertising is an example of merging information access (through web services) and mobile positioning. Several web-based map providers are starting to offer recommendations based on the location of the user. This can only be possible having a mobile device that can communicate through a wireless network. Nowadays, there are a variety of (mobile) devices and applications that are capable of running mobile services (consider for example a Smartphone with email/internet access and GPS positioning). Yet, little is known of the success of this new technology, and most importantly, how are people adopting it. The success of a mobile platform is not determined exclusively by technology, but also by how people use it to enrich their lives.

CONTRIBUTIONS OF THE BOOK This handbook of research contributes with state-of-the-art research on all three aspects considered previously: technology, applications and social impact on the adoption of these systems. Several technologies are covered in the book, ranging from performance improvements and routing optimizations on (generic) wireless networks; considerations for WiMAX; new designs for sensor networks; and new uses for RFID and Near Field Communication systems. In the application arena, several areas of interest have been covered, namely: health, e-government, advertisement, e-learning, university campus services, vehicular networks, Virtual and Small-Medium Enterprises and mobile grid computing. A particular attention was given to e-learning, with several contributions in providing lessons through mobile devices, supporting and monitoring tools on mobile devices and identifying success factors for mobile learning model, including specific case studies in large organizations. Device localization received several contributions for being an important operational feature for positioning mobile devices. Two wide range of localization schemes were considered: GPS based and local wireless (WiFi) based. Device localization is necessary to enhance a whole range of services, from advertisement to social ones. Vehicular networks are yet another mobile computing platform that has specific characteristics as a wireless ad-hoc network. This type of network raises unique challenges and security issues that were identified and are not present on typical (computer) networks. In the social impact of a mobile computing platform, several domains have been presented, from a specific collaborative learning environment, to the use of mobile devices in higher education institutions to enterprise institutions. Indeed, the adoption of mobile devices is a disruption of the typical stationary computing platform. Youngsters tend to adopt more easily new technologies, in this case mobile devices, than elders. The typical population of education institutions is younger than enterprise institutions, hence the interest of these case studies in different environments. However, the mobile platform is not confined to the educational environment, and hence the interest of considering how mobile applications can be applied to

lxii

new domains like tourism and entertainment. Regardless of the specific domain in use, it is shown the viability of business models for mobile e-commerce. In summary, a handbook of research that demonstrates what can be achieved within this domain would have been welcome in itself. A handbook that besides this purpose gives real insight, and that drives and inspires those that read it to contribute to the ongoing technological improvements and development of successful applications, represents a major contribution to the body of knowledge in mobility and computing.

Nuno Lopes Assistant Professor of Networks and Communication Technologies Polytechnic Institute of Cávado e Ave Portugal Nuno Lopes received his Bachelor (5 year degree) in Systems and Informatics Engineering, in 2002, from University of Minho, Braga, Portugal. During this course, he made an internship at Philips Research, Eindhoven, The Netherlands. Later on he received his PhD degree in Computer Science, from University of Minho, Portugal, in 2009. His PhD focused on the building of large-scale indexing systems through the use of structured peer-to-peer networks. He is currently an Assistant Professor at the Instituto Politécnico do Cávado e do Ave, Barcelos, Portugal, teaching Network Communications and Operating Systems courses, among others. His research interests include Distributed Systems, Decentralized Algorithms, Peer-to-peer Networks, and Large-scale Information Retrieval.

lxiii

Preface

ABOUT THE SUBJECT The advances in mobile and ubiquitous technologies have more than 35 years of research. After all these years the ubiquity vision is becoming a reality, due to the evolution and new features of the hardware and software components. The advent of Web services and the progress on wearable devices, ambient components, user-generated content, mobile communications, and many other gave origin to new applications and services. Recently advances in web services technologies allowed the integration into mobility. With these two components appeared new business models which imply improvements in technical infrastructure that enables the progress of mobile services and applications. These include dynamic and on-demand service, context-aware services, and mobile web services. While driving new business models and new online services, particular techniques must be developed for web service composition, web service-driven system design methodology, creation of web services, and on-demand web services. The technological, social and organizational dimensions of mobile and ubiquitous computing will be supported by new educational paradigm. Two trends converge to make this possible; increasingly powerful cell phones and PDAs, and improved access to wireless broadband. This handbook of research represents a collection of the most recent developments on the technological, organizational and social dimensions of this highly potential technology of mobility and computing, that dictate a paradigm shift in organizations, society and people. It covers the following dimensions: • •

•

The technological dimension, addressing emerging technologies such as Wireless data communication networks; Mobile technologies; Standards and reference models; and Security. The organizational, human and social perspectives, comprising: New business models; Collaborative work; Learning and teaching; Social networks; Studies on adoption; Studies of impact; Critical success factors; and level of preparedness of organizations and people. Applications and solutions developed or under development, and research and development results that address a broad range of sectors, from business to government services, from education to e-Health. The chapters included in the handbook,

•

discuss the importance of this technologies and their support to new applications and solutions, namely in sectors of activity such as business, new organizational and ubiquitous models, eCommerce, government, e-Health, e-Learning, etc.;

lxiv

• • • • • •

present practical solutions and recent developments; introduce the state-of-the-art technologies; discuss organizational preparedness for new organizational models potentiated by ubiquity and mobility; discuss the impact on society and on people; discuss critical success factors; introduce future generations of communications and networks.

ORGANIZATION OF THE HANDBOOK This handbook is a compilation of 77 contributions to the discussion of the main issues, challenges, opportunities and developments related with mobility and computing, in a very comprehensive way, in order to disseminate current achievements and practical solutions and applications. These 77 chapters are written by a group of near 240 authors that include many internationally renowned and experienced researchers and specialists in the e-Business field and a set of younger authors, showing a promising potential for research and development. Contributions came from the five continents, and are included contributions from academia, research institutions and industry, representing a good and comprehensive representation of the state-of-the-art approaches and developments that address the several dimensions of this fast evolutionary thematic. The “Handbook of Research on Mobility and Computing: Evolving Technologies and Ubiquitous Impacts” is organized in six sections: Section 1: Mobile Technologies Section 2: Emerging Technologies Section 3: Critical Success Factors Section 4: New Business Models Section 5: Security Section 6: Applications, Surveys and Case Studies The first section, Mobile Technologies is composed by 36 chapters that address relevant research and development contributions to the field of mobility and computing, summarized below. Chapter one, Evaluating the Context Aware Browser: A Benchmark for Proactive, Mobile, and Contextual Web Search, discusses the evaluation of highly interactive and novel context-aware system with a methodology based on a TREC-like benchmark. The authors take as a case study an application for Web content perusal by means of context-aware mobile devices, named Context-Aware Browser. In this application, starting from the representation of the user's current context, queries are automatically constructed and used to retrieve the most relevant Web contents. Since several alternatives for query construction exist, it is important to compare their effectiveness, and to this aim we developed a TREClike benchmark. The authors present their approach to an early stage evaluation, describing their aims and the techniques applied, underlining for, for the evaluation of context-aware retrieval systems, the benchmark methodology adopted could be an extensible and reliable tool. Routing is the process of finding a path from a source node to a destination node. Proposed routing schemes can be divided into topological and position based, depending on the availability of geographic

lxv

location for nodes. Topological routing may be proactive or reactive, while position based routing consists of greedy approaches applied when a neighbor closer to the destination (than the node currently holding the packet) exists, and recovery schemes otherwise. In order to preserve bandwidth and power which are critical resources in ad hoc and sensor networks, localized approaches are proposed, where each node acts based solely on the location of itself, its neighbors, and the destination. There are various measures of optimality which lead to various schemes which optimize hop count, power, network lifetime, delay, or other metrics. This second chapter, Routing in Wireless Ad Hoc and Sensor Networks, describes a uniform solution based on ratio of cost and progress. Mobile communication networks have become an integral part of our society, significantly enhancing communication capabilities. Mobile ad hoc networks (MANETs) extend this capability to any time/ anywhere communication, providing connectivity without the need of an underlying infrastructure. Chapter three, Mobile Ad Hoc Networks: Protocol Design and Implementation investigates the new coming realm of mobile ad hoc networks, focusing on research problems related to the design and development of routing protocols, both from a formal and technical point of view. Then the link stability in a high mobility environment is examined, and a route discovery mechanism is analyzed, together with a practical implementation of a routing protocol in ad hoc multi-rate environments which privileges link stability instead of traditional speed and minimum distance approaches. The convergence of wireless applications presents a greater hope for consolidating e-Government applications even in resource-constrained countries such as those in Africa. Chapter four, Convergence of Wireless Technologies in Consolidating E-Government Applications in Sub-Saharan Africa, presents an exploratory study that aims at discussing the extent as to how convergence of wireless technologies from different vendors promises to contribute to the consolidation of e-Government applications in SubSaharan-Africa (SSA). This is done by reviewing the different adoption stages of ICT and e-Government in SSA. It looks at challenges facing adoption of wireless technologies (GSMs, Wireless Internet Access, satellite transmission, etc.) across all the socio-economic value chains in SSA. The chapter looks at Botswana and South Africa as case studies by bringing out the different interventions that have been done in the realm of facilitating a conducive environment for the convergence of different wireless technologies. Out of the analysis of legal, regulatory, market and spectrum policies affecting the adoption of wireless communications in SSAs, the chapter draws out recommendations on how to consolidate wireless communications to be adopted in different socio-economic setups (e.g. e-government, e-Health, e-Banking, etc.). The fifth chapter, Process Innovation with Ambient Intelligence (AmI) Technologies: A Comparison of Australian and German Small and Medium Enterprise (SME) Manufacturers, considers the potential for absorptive capacity limitations to prevent SME manufacturers benefiting from the implementation of Ambient Intelligence (AmI) technologies. The chapter also examines the role of intermediary organisations in alleviating these absorptive capacity constraints. In order to understand the context of the research, a review of the role of SMEs in the Australian manufacturing industry, plus the impacts of government innovation policy and absorptive capacity constraints in SMEs in Australia is provided. Advances in the development of ICT industry standards, and the proliferation of software and support for the Windows/ Intel platform have brought technology to SMEs without the need for bespoke development. The results from the joint European and Australian AmI-4-SME projects suggest that SMEs can successfully use "external research sub-units" in the form of industry networks, research organisations and technology providers to offset internal absorptive capacity limitations.

lxvi

Chapter six, Providing Ubiquitous Access to Synthetic Sign Language Contents Over Multiple Platforms, presents the design of distributed sign language synthesis architecture. The main objective of this design is to adapt the synthesis process to the diversity of user devices. The synthesis process has been divided into several independent modules that can be run either in a dedicated server or in the client device. Depending on the modules assigned to the server or to the client, four different scenarios have been defined. These scenarios may vary from a heavy client design which executes the whole synthesis process, to a light client design similar to a video player. These four scenarios will provide equivalent signed message quality independently of the device’s hardware and software resources. Chapter seven, The Impact of MIMO Communication on Non-Frequency Selective Channels Performance, reviews the basic concepts of multiple-input multiple-output (MIMO) communication systems and analyses their performance within non-frequency selective channels. The MIMO system model is established and by applying the singular value decomposition (SVD) to the channel matrix, the whole MIMO system can be transformed into multiple single-input single-output (SISO) channels having unequal gains. In order to analyze the system performance, the quality criteria needed to calculate the error probability of M-ary QAM (Quadrature Amplitude Modulation) are briefly reviewed and used as reference to measure the improvements when applying different signal processing techniques. Bit and power allocation is a well-known technique that allows improvement in the bit-error rate (BER) by managing appropriately the different properties of the multiple SISO channels. It can be used to balance the BER’s in the multiple SISO channels when minimizing the overall BER. In order to compare the various results, the efficiency of fixed transmission modes is studied in this work regardless of the channel quality. It is demonstrated that only an appropriate number of MIMO layers should be activated when minimizing the overall BER under the constraints of a given fixed date rate. Chapter eight, Node Localization in Ad-hoc Networks, introduces node localization techniques in adhoc networks including received signal strength (RSS), time-of-arrival (TOA) and direction-of-arrival (DOA). Wireless channels in ad-hoc networks can be categorized as LOS and NLOS. In LOS channels, the majority of localization techniques perform properly. However, in NLOS channels, the performance of these techniques reduces. Therefore, non-line-of-sight (NLOS) identification and mitigation techniques, and localization techniques for NLOS scenarios are briefly reviewed. The technological revolution that has created a vast health problem due to a drastic change in lifestyle also holds great potential for individuals to take better care of their own health. This is the focus of the presented overview of current applications, and prospects for future research and innovations, addressed by chapter nine, Wireless and Mobile Technologies Improving Diabetes Self-Management. The presented overview and the main goals of the systems included are to utilize ICT as aids in self-management of individual health challenges, for the disease Diabetes, both for Type 1 and Type 2 diabetes. People with diabetes are generally as mobile as the rest of the population, and should have access to mobile technologies for managing their disease. Forty-seven relevant studies and prototypes of mobile, diabetesspecific self-management tools meeting our inclusion criteria have been identified; 27 publicly available products and services, nine relevant patent applications, and 31 examples of other disease-related mobile self-management systems are included to provide a broader overview of the state of the art. Finally, the reviewed systems are compared, and future research directions are suggested. Chapter ten, Adaptive Multicarrier Frequency Hopping Spread Spectrum Combined with Channel Coding, presents an adaptive Multicarrier Frequency Hopping Spread Spectrum (MCFH-SS) system employing proposed Quasi Cyclic Low Density Parity Check (QC-LDPC) codes instead of the conventional LDPC codes. A new technique for constructing the QC-LDPC codes based on row division method

lxvii

is proposed. The new codes offer more flexibility in terms of girth, code rates and codeword length. Moreover, a new scheme for channel prediction in MCFH-SS system is also proposed. The technique adaptively estimates the channel conditions and eliminates the need for the system to transmit a request message prior to transmitting the packet data. The proposed adaptive MCFH-SS system uses PN sequences to spread out frequency spectrum, reduce the power spectral density and minimize the jammer effects. Although B2B e-commerce provides healthcare organizations a wealth of new opportunities and ways of doing business, it also presents them with a series of challenges. B2B e-commerce adoption remain poorly understood and it is also a relatively under-researched area. Therefore, case studies were conducted to investigate the challenges and issues in adopting and utilizing B2B e-commerce systems in the healthcare sector. The major aims of the study presented in chapter eleven, Key Adoption Challenges and Issues of B2B E-Commerce in the Healthcare Sector are to: (a) identify and examine main B2B e-commerce adoption challenges and issues for healthcare organizations; and (b) develop a B2B e-commerce adoption challenges and issues table to assist healthcare organizations in identifying and managing them. Chapter twelve, A Pervasive Polling Secret-Sharing Based Access Control Protocol for Sensitive Information, presents a novel access control mechanism for sensitive information which requires permission from different entities or persons to be accessed. The mechanism consists of a file structure and a protocol which extend the features of the OpenPGP Message Format standard by using secret sharing techniques. Several authors are allowed to work in the same file, while access is blocked for not authorized users. Access control rules can be set indicating the minimum number of authors that need to be gathered together in order to open the file. Furthermore, these rules can be different for each section of the document, allowing collaborative work. Non-repudiation and authentication are achieved by means of a shared signature. The scheme’s features are best appreciated when using it in a mobile scenario. Deployment in such an environment is easy and straight. Mobile devices, including cell phones, capable of geo-positioning (or localization) are paving the way for new computer assisted systems called mobile location-based recommenders (MLBRs). MLBRs are systems that combine information on user’s location with information about user’s interests and requests to provide recommendations that are based on “location”. MLBR applications are numerous and emerging. One MLBR application is in advertisement where stores announce their coupons and users try to find the coupons of their interests nearby their locations through their cell phones. Chapter thirteen, Mobile Location-Based Recommenders: An Advertisement Case Study discusses the concept and characteristics of MLBRs and presents the architecture and components of a MLBR for advertisement. Mobile devices have become a new platform with many possibilities to develop studies and implement projects. The power and current capabilities of these devices besides its market penetration makes applications and services in the area of mobility particularly interesting. Mobile terminals have become small computers, they have an operating system, storage capacity so it is possible to develop applications that run on them. Today these applications are highly valued by users. Nowadays we want not only to talk or send messages by mobile terminal, but also to play games, to buy cinema tickets, to read email… We can bring these capabilities in our pocket. The University may not be aware of this fact. The students, due to their age, are the main users and purchasers. In this sense, chapter 14, Services for Mobile Devices in a Universitary Scenario, presents three applications developed for mobile devices, that are being used in Universidad Pontificia de Salamanca. All of them work on a university scenario and use different kind of services.

lxviii

Wireless Sensor Networks (WSNs) have experienced an amazing evolution during the last decade. Compared with other wired or wireless networks, wireless sensor networks extend the range of data collection and make it possible for us to get information from every corner of the world. The chapter begins with an introduction to WSNs and their applications. Chapter 15, Event Detection in Wireless Sensor Networks recognizes event detection as a key component for WSN applications. The chapter provides a structured and comprehensive overview of various techniques used for event detection in WSNs. Existing event detection techniques have been grouped into threshold based and pattern based mechanisms. For each category of event detection mechanism, the chapter surveys some representative technical schemes. The chapter also provides some analyses on the relative strengths and weaknesses of these technical schemes. Towards the end, the trends in the research regarding the event detection in WSNs are described. At the beginning of the 21st century in a world dominated by technology it is essential to enhance and update the school, creating conditions for the students to succeed, consolidating the role of ICT as a key resource for learning and teaching in this new era. In Chapter 16, M-English – Podcast: A Tool for Mobile Devices, the authors describe a study that was carried out in a Portuguese school. As a means to overcome some of the existing logistical obstacles in the school, where the possibility of carrying out ICT activities without restrictions was still dreamlike, the podcast was implemented as an m-learning tool. Being aware of the fact that nowadays mobile phones and mp3 players are part of our students’ lives, the authors took advantage of this fact and the podcast was used as a tool to support, to enhance and to motivate students to learn English, used thus as a complement to traditional (face-to-face) learning. Disaster can be defined as the onset of an extreme event causing profound damage or loss as perceived by the afflicted people. The networks built in order to detect and handle these events are called Public safety networks (PSNs). These networks have the fundamental role of providing communication and coordination for emergency operations. Many of the problems of the PSN field come from the heterogeneity of systems and agencies involved in the crisis site and from their mobility at the disaster site. The main aim of Chapter 17, Public Safety Networks, is to provide a broad view of the PSN field, presenting the different emergency management phases, PSNs requirements, technologies and some of the future research directions for this field. The purpose of Chapter 18, Mobile Applications as Mobile Learning and Performance Support Tools in Psychotherapy Activitites, is the analysis of mobile applications as performance and informal learning support tools that facilitate the development of the psychotherapy process. The “e-therapy” has become a common term to refer the delivery of mental health services, on-line or related to a computer mediated communication between a psychotherapist and the patient. Initially, a background on e-therapy is provided and after the “self-help therapy”–a kind of e-therapy where the concept of patient empowerment is important– is exposed. Then, the integration of mobile devices in the psychotherapy process will be explained considering how their technological features support patient therapeutic activities like behavior assessment and informal mobile learning. The relation of the mobile devices with psychotherapist work activities such as evidence gathering and patient monitoring will also be explained. The chapter includes a discussion on the mobile learning practices as a source of potential strategies that can be applied in the therapeutic field and finally a set of recommendations and future directions are described to explore new lines of research. Location-based services (LBSs) are impacting different aspects of human’s life. To date, different LBSs have been emerged, each supporting a specific application or service. While some LBSs have aimed at addressing the needs of general populations, such as navigation systems, others have been focused

lxix

on addressing the needs of specific populations, including kids, youths, elderly, and people with special needs. In recent years, interest in taking a LBS approach in education and learning has grown. The main purpose of such educational LBSs is facilitating a means for learners to be more efficient and effective in their learning activities using their location as the underlying information in decision making. In CampusLocator: A Mobile Location-Based Service for Learning Resources the authors present a novel LBS, called CampusLocator, whose main goal is to assist students in locating and accessing learning resources including libraries, seminars, and tutorials that are available on a campus. WiMAX is being promoted as a potential solution to a number of problems that have plagued the wired and wireless broadband industry since it originated. Can WiMAX fulfill this promise in a crowded and competitive market? If so, what factors are critical to its success? Who will use WiMAX and for what purposes? This chapter, The Future of WiMAX, identifies both the critical success factors that will give WiMAX an edge over other existing wireless technologies and the key applications that will contribute to its success. The top three critical success factors for WiMAX are availability of handset devices and consumer premise equipment, bandwidth speed, and interoperability and standardization. A panel of WiMAX experts concludes that broadband on demand, wireless services provider access, and Voice over IP are the top three killer applications for WiMAX. Chapter 21, Determinants of Loyalty Intention in Portuguese Mobile Market, conceptualizes and highlights the determinants of customers’ loyalty in the Portuguese mobile market. The authors raise questions about the interrelationships of the cost and values dimensions and the consequences of these relationships on customer satisfaction and trust and consequently loyalty among different operators, addressing some recent models. By organizing and synthesizing the major research streams and tests empirically a conceptual framework through a SEM, with data gather in a survey of Portuguese clients, the present study advances knowledge on the nature of the relative importance of different components of loyalty to mobile communications operators. Some useful preliminary insights were produced related to customers’ retention process in primary mobile operator, which appears strongly related to price/ quality, followed by the emotional connection to the operator staff and others clients. Nonetheless, a considerable number of issues were left for future research, including the possibility of extending the investigation to other countries. Current technological advances and the increasing diffusion of its use for scientific, financial and social activities, make Internet the de facto platform for providing worldwide distributed data storage, distributed computing and communication. It creates new opportunities for the development of new kinds of applications, but it will also create several challenges in managing the information distributed on the Internet and in guaranteeing its “on-time” access through the network infrastructures that realize the Internet. Many researchers believed and still believe that the mobile agents could propose several attractive solutions to deal with such challenges and problems. Chapter 22, Mobile Agents: Concepts and Technologies, presents the core concepts of mobile agents, and attempts to provide a clear idea of the possibility of their use by introducing the problems they cope with, the application areas where they provide advantages with respect to other technologies and the available mobile agent technologies. Traditional networks suppose the existence of some path between endpoints, small end to end roundtrip Delay time, and loss ratio. Today, however, new applications, environments and types of devices are challenging these assumptions. In Delay Tolerant Networks (DTNs), an end-to-end path from source to destination may not exist. Nodes may connect and exchange information in an opportunistic way. Chapter 23, Vehicular Delay Tolerant Networks, presents a broad overview of DTNs, particularly focusing on Vehicular DTNs, their main characteristics, challenges, and research projects on this field. In the near

lxx

future, cars are expected to be equipped with devices that will allow them to communicate wirelessly. However, there will be strict restrictions to the duration of their connections with other vehicles, whereas the conditions of their links will greatly vary; DTNs present an attractive solution. Therefore, VDTNs constitute an attractive research field. It has been substantially proven that the use of new technologies can improve a child's learning process. However, the main problem for the teacher continues to be that he/she cannot pay attention to all children at the same time. Sometimes it is necessary to decide which child must be first attended to. It is in this context that we believe our system has the ability to greatly help teachers: we have developed a learning process control system that allows teachers to determine which students have problems, how many times a child has failed, which activities they are working on and other such useful information, in order to decide how to distribute his/her time. Furthermore, bearing in mind the attention required by kindergarten students, Chapter 24, Monitoring the Learning Process through the use of Mobile Devices, proposes the provision of mobile devices (PDA - Personal Digital Assistant) for teachers, permitting free movement in the classroom and allowing the teacher to continue to help children while information about other students is being received. Therefore if a new problem arises the teacher is immediately notified and can act accordingly. Computer technologies, especially ICT have become ubiquitous in people’s lives. Nowadays, mobile phones, PDAs, laptops and a constellation of software tools are more and more used for a variety of activities carried out in both personal and professional lives. Given the features that these technologies provide and are provided with, for example, connectivity and portability, it can be said that ICTs have the potential to support nomadic work practices which are seen as increasingly characteristic of the knowledge economy. Chapter 25, The Making of Nomadic Work: Understanding the Mediational Role of ICTs, presents a review of the concept of nomadic work and, based on a broad literature analysis, discusses the ways in which ICTs may empower people who are involved with nomadic work practices. It aims to give a starting point for those who intend to develop further research on technologically-mediated nomadic work practices in the knowledge economy. Chapter 26, I-GATE: Interperception - Get All The Environments, introduces the I-GATE architecture, a new approach, which includes a set of rules and software architecture, to connect users from different interfaces and devices in the same virtual environment, transparently, even with low capacity of resources. The system detects the user resources and provides transformations on the data in order for its visualization in 3D, 2D and textual-only (1D) interfaces. This allows users from any interface to get a connection in the system using any device and to access and exchange information with other users (including ones with other interface types) in a straightforward way, without need to changing hardware or software. The authors formalize the problem, including modeling, implementation, and usage of the system, also introducing some applications that they have created and implemented in order to evaluate the proposal. Authors have used these applications in cell phones, PDAs, Digital Television, and heterogeneous computers, using the same architecture, with success. The widespread and availability of increasingly powerful mobile devices is contributing for the incorporation of new services and features on our daily communications and social relationships. In this context, geolocation of users and points of interest in mobile devices may contribute, in a natural way, to support either the mediation of remote conversations as the promotion of face-to-face meetings between users, leveraging social networks. The CONNECTOR system is based on geolocation data (people, content and activities), enabling users to create and develop their personal relations with other members of the CONNECTOR social network. Users, maps, sharing features and multimedia content

lxxi

are actors in this social network allowing CONNECTOR to address the promotion of geolocated social networks driven by physical proximity and common interests among users. Chapter 27, CONNECTOR: A Geolocated Mobile Social Service, discusses the work undertaken for the conceptualization and development of the CONNECTOR system. Preliminary evaluation results along with usage contexts are also presented. The chapter concludes with a discussion about future developments in geolocation and personalization in mobile communication services. Nowadays, triple-play services are offered in both wireless and wired networks. The network convergence and the new services such as VoIP and IPTV are a reality. However, the future of these networks will have a different concept, the ubiquity. The solutions must be based on the structures and current environments to carry out those challenges in a correct way. In order to reach this ubiquity, the science community has to take into account that its implementation should not assume a high cost to the user and that the system must comply with the quality of service measurements to satisfy the user. Chapter 28, Providing VoIP and IPTV Services in WLANs, introduces the main VoIP and IPTV (IP Television) transmission protocols and the compression formats most used as also the bandwidth needed. The goal is to provide ubiquity into multimedia scenarios in WLANs. Authors will carry out tests to guarantee the appropriate values in some network parameters such jitter, delay, number of lost packets and enough effective bandwidth that should be satisfied. They demonstrate the measurements taken from several test benches. They show the parameter values that the devices should perform in order to stay connected from anywhere at any time to these services. One of the sectors that have a great potential to use and exploit the new communication technologies is the healthcare sector. Nowadays, the application of all these new technologies to support the clinical procedures has taken part in the definition of a new concept known as e-Health. This concept involves a lot of different services related with the medicine/health terms and the information technologies. However, to provide emergency transportation with better care capabilities to the patient is something that still has a lot to improve. Within this context SIe-Health comes into being a software platform oriented for developing Telemedicine solutions. The solution model proposed in Chapter 29 SIe-Health, e-Health Information System, allows remote assistance for a mobile health emergency, integrating in this service electro-medical devices and videoconference services. The popularity of mobile computing creates new opportunities for information sharing and collaboration through technologies like radio frequency identification (RFID) tags and location awareness technologies. Chapter 30, Combining Location Tracking and RFID Tagging toward an Improved Research Infrastructure, discusses how these technologies, which provide subtly different information, can be used together toward increased benefit to users. This work introduces technologies for RFID and location awareness, including a survey of projects. The authors describe advantages of combining these technologies, illustrated through our system, TagIt, that uses these technologies in a traditional research poster environment to provide a rich multimedia experience and encourage ongoing feedback from poster viewers. An overview of TagIt is provided, including user commenting and information sharing capabilities that make use of RFID and location information. User feedback and an expert review highlights how TagIt could benefit authors, information consumers, and the research community, leading to future directions for the research community. The appearance of concepts such as “Ambient Intelligent”, “Ubiquitous Computing” and “ContextAwareness” is causing the development of a new type of services called “Context-Aware Services” that in turn may affect users of mobile communications. This technology revolution is a complex process because of the heterogeneity of contents, devices, objects, technologies, resources and users that can be

lxxii

coexist at the same local environment. The novel approach in Chapter 31, Model and Infrastructure for Communications in Context-Aware Services, is the development of a ”Local Infrastructure” in order to provide intelligent, transparent and adaptable services to the user and solves the problem of local context control. Authors present a conceptual model for development of local infrastructure, an architecture design to control the service offered by the local infrastructure. This infrastructure proposed consists of an intelligent device network to link the personal portable device with the contextual services. The device design is modular, flexible, scalable, adaptable and reconfigurable remotely in order to tolerate new demand services whenever is needed. Finally, the result suggests that authors will be able to develop a wide range of new and useful applications, not conceived at origin. The use of mobile devices with possible connection to the Internet is increasing tremendously. This mobility poses new challenges at various levels, including hardware, network services, and the development of applications. The user searches small and lightweight devices, easy to use, and with vast autonomy in terms of energy. She/He seeks also to connect the Internet “every time, everywhere”, possibly using different access technologies. Given the interface limitations, and processing capabilities of small mobile devices, the software and the operating system used must be necessarily adapted. Chapter 32, Network Mobility and Mobile Applications Development, overviews the mobility area, provides deep insight in the field, and presents the main existing problems. Mobility and the development of mobile applications are closed related. The advances in network mobility lead to different approaches in the mobile applications development. The chapter proposes a model for developing mobile applications, based on our research. Wireless sensor networks are composed of geographically dispersed sensors that work together to monitor physical or environmental conditions. In addition, wireless sensor networks are used in many industrial, social, and regulatory applications, including industrial process monitoring and control, environment and habitat monitoring, healthcare, home automation, and traffic control. Developers of wireless sensor networks face a number of programming and deployment challenges. Chapter 33, Building Mobile Sensor Networks Using Smartphones and Web Services: Ramifications and Development Challenges, shows how smartphones can help reduce the development, operation, and maintenance costs of wireless sensor networks, while also enabling these networks to use web services, high-level programming APIs, and increased hardware capability, such as powerful microprocessors. Moreover, this chapter examines key challenges associated with developing and maintaining a large wireless sensor network and presents a novel smartphone wireless sensor network that uses smartphones as sensor nodes. The work is validated in the context of Wreck Watch, which is a smartphone-based sensor network for detecting traffic accidents that authors use to demonstrate solutions to multiple challenges in current wireless sensor networks. The authors also describe common pitfalls of using smartphones as sensor nodes in wireless sensor networks and summarize how have addressed these pitfalls in Wreck Watch. Modern life makes people internet-dependents. They want to move connected and care for always getting the best options for connectivity, hoping between providers. Freedom for choosing providers and the business options which these exchanges can offer are the motivations for Chapter 34, Technologies to improve the quality of handovers: Ontologies, Contexts and Mobility Management. After pointing out some characteristics which make the basics of the current handover technologies, we describe an information infrastructure, based on context and ontologies which can be used to foster an intelligent, efficient and profitable scenario for managing handovers in the Next Generation Networks. Some experiments are described and the potential of using these technologies are evaluated. In smart environments, making a location-aware personal computing working accurately is a way of getting close to the pervasive computing vision. The best candidate to determine a user location in

lxxiii

indoor environment is by using IEEE 802.11 (Wi-Fi) signals, since it is more and more widely available and installed on most mobile devices used by users. Unfortunately, the signal strength, signals quality and noise of Wi-Fi, in worst scenario, it fluctuates up to 33% because of the reflection, refraction, temperature, humidity, the dynamic environment, etc. Chapter 35, Making Location-Aware Computing Working Accurately in Smart Spaces, presents the current development on a light-weight algorithm, which is easy, simple but robust in producing the determination of user location using WiFi signals. The algorithm is based on "multiple observers" on ηk-Nearest Neighbour. The authors extend the approach in the estimation indoor-user location by using combination of different technologies, i.e. WiFi, GPS, GSM and Accelerometer. The algorithm is based on opportunistic localization algorithm and fuse different sensor data in order to be able to use the data which is available at the user position and processable in a mobile device. Context-aware computing is a class of mobile computing that can sense its physical environment and adapt its behavior accordingly; it is a component of the ubiquitous or pervasive computing environment that has become apparent with innovations and challenges. Chapter 36, User Pro-Activities Based on Context History, reviews the concept of context-aware computing, with focus on the user activities that benefit from context history. How user activities in the smart environment can make use of context histories in applications that apply the concept of context prediction integrated with user pro-activity is explored. A brief summary of areas which benefit from these technologies as well as corresponding issues are also investigated. The ten chapters of Section 2, Emerging Technologies, introduce emerging technologies able to dictate a complete shift of applications and services towards ubiquity. Advances in wireless, sensor, mobile and wearable technologies present new challenges for data mining research on providing mobile applications with intelligence. Autonomy and adaptability requirements are the two most important challenges for data mining in this new environment. In Chapter 37, Research Challenge of Locally Computed Ubiquitous Data Mining, the authors analyse the challenges of designing ubiquitous data mining services by examining the issues and problems while paying special attention to context and resource awareness. The authors focused on the autonomous execution of a data mining algorithm and analyzed the situational factors that influence the quality of the result. Already existing solutions in this area and future directions of research are also covered. Chapter 38, Emerging Wireless Networks for Social Applications, describes the implementation and performance evaluation of a novel routing protocol called Pandora, designed for social applications, and that can be implemented in a broad number of devices, such as commercial wireless routers and laptops. It also provides a robust backbone integrating and sharing data, voice and video between computers and mobile devices. Pandora offers great performance with both fixed and mobile devices and includes important features such as: geographic positioning, residual battery energy monitoring, and bandwidth utilization. In addition, Pandora also considers the number of devices attached to the network. Pandora is experimentally evaluated in a testbed with laptops for the first stage and commercial wireless routers for the second stage. The main goal of Pandora is to provide a reliable backbone for social applications requiring a quality of service (QoS) guarantee. With this in mind, the following evaluation of Pandora considers the following types of traffic sources: transport control protocol (TCP), voice, video and user datagram protocol (UDP) without marks. Pandora is also compared with different queuing disciplines, including: priority queuing discipline (PRIO), hierarchical token bucket (HTB) and DSMARK. Finally, an Internet radio transmission is employed to test the network re-configurability. Results show that queuing the PRIO and HTB disciplines, which prioritizes UDP traffic, performed the best.

lxxiv

Several complex and time-critical applications require the existence of novel distributed, heterogeneous and dynamic platforms composed of a variety of fixed and mobile processing nodes and networks. Such platforms, that can be called Pervasive Mobile Grids, aim to merge the features of Pervasive Computing and High-performance Grid Computing onto a new emerging paradigm. In this chapter we study a methodology for the design and the development of high-performance, adaptive and context-aware applications. Chapter 39, “An Approach to Mobile Grid Platforms for the Development and Support of Complex Ubiquitous Applications”, describes a programming model approach, and the authors compare it with other existing research works in the field of Pervasive Mobile Computing, discussing the rationales of the requirements and the features of a novel programming model for the target platforms and applications. In order to exemplify the proposed methodology, the authors introduce the programming framework ASSISTANT, and provide some interesting future directions in this research field. Potentially, in the mobile computing scenario, the users can move in different environments and the applications can automatically explore their surroundings. This kind of context-aware application is emerging, but is not yet widely disseminated. Based on perceived context, the application can modify its behavior. This process, in which software modifies itself according to sensed data, is named Adaptation and constitutes the core of Ubiquitous Computing. The ubiquitous computing scenario brings many new problems such as coping with the limited processing power of mobile devices, frequent disconnections, the migration of code and tasks between heterogeneous devices, and others. Current practical approaches to the ubiquitous computing problem usually rely upon traditional computing paradigms conceived back when distributed applications where not a concern. Holoparadigm (Holo) was proposed as a model to support the development of distributed systems. Based on Holo concepts, a new programming language called HoloLanguage (HoloL) was created. Chapter 40, Towards a Programming Model for Ubiquitous Computing, propose the use of Holo for developing and executing ubiquitous applications, explore the HoloL for ubiquitous programming and propose a full platform to develop and execute Holo programs. The execution environment is based on a virtual machine that implements the concepts proposed by Holo. The Virtual Enterprise (VE) is a collaboration model between multiple business partners in a value chain that aims to cope with turbulent business environments, mainly characterized by demand unpredictability, shortening product lifecycles, and intense cost pressures. The VE model is particularly viable and applicable for SME and industry parks containing multiple SMEs that have different vertical competencies. As small firms collaborate effectively under VE model, it will be only possible to emerge products by joining their diverse competencies and mitigate the effects of market turbulence by minimizing their investment. A typical VE model has four phases; opportunity capture, formation, operation, and dissolution. The goal of Chapter 41, An Agent-Based Operational Virtual Enterprise Framework Enabled by RFID, is to present a conceptual VE framework, focusing on operation phase, that incorporates Multi Agent Systems (MAS) and Radio Frequency Identification Systems (RFID) which are emerging from research to industry with a great momentum. First, state of the art for VE and the two key enabling technologies are covered in detail. After presenting conceptual view of the framework, an Information and Communication Technology (ICT) view is also given to enhance technical integration with available industry standards and solutions. Finally, process views of how a VE can operate utilizing agent based and RFID systems in order to fulfill operational requirements, are presented. Chapter 42, Ontological Dimensions of Semantic Mobile Web 2.0. First Principles, advances, from the point of view of Knowledge Representation and Reasoning, an analysis of which ontological dimensions are needed to develop Mobile Web 2.0 on top of Semantic Web. This analysis will be particularly focused on social networks and it will try to make an outlook about the new knowledge challenges on

lxxv

this field. Some of these new challenges will be linked to Semantic Web context, while others will be inherent to Semantic Mobile Web 2.0. Unobtrusiveness is a key factor in usability of mobile and ubiquitous computing systems. These systems are made of several ambient and mobile devices whose goal is supporting everyday life users’ activities, hopefully without interfering with them. We intend to address the topic of obtrusiveness by assessing its impact in the design of interfaces for mobile and ubiquitous computing systems. The authors of Chapter 43, Unobtrusive Interaction with Mobile and Ubiquitous Systems, will make the case of how unobtrusive interfaces can be designed by means of Kinetic User Interfaces: an emerging interaction paradigm where input to system is provided through coordinated motion of objects and people in the physical space. A huge amount of information is used nowadays by modern vehicles, and it may be accessed through an On Board Diagnosis (OBD) connection. A technique using the already installed OBD system to communicate with the vehicle together with a Global Position System (GPS) provides reliable data which allow a detailed analysis of real on road tests. Different kinds of circulation circuits (urban, extra-urban and highway) were analyzed in Chapter 44, Impact of Advances on Computing and Communication Systems in Automotive Testing, using the capabilities of OBD II installed on the tested vehicles. OBD provides an important set of information, namely related with data on the engine, fuel consumption, chassis and auxiliary systems and also on combustion efficiency. The use of GPS in all the road tests performed provides important information to further determine the more sustainable from all the different solutions tested, considering the different situations imposed on each circuit. It is a fact that bench tests or a chassis dynamometer allow a fine control of the operation conditions; however the simulation is not as real as on the road. So the present methodology will allow the possibility to perform tests on the road, allowing enough control on vehicles and providing complete information of the chosen route and of the trip history. This is a possibility that ensures new tools with more reliable data which can give faster answers for the development of high efficiency, economic and environmentally neutral automotive technologies. RFID (Radio Frequency Identification) technology consists of a tag that can be used to identify an animal, a person or a product, and a device responsible for transmitting, receiving and decoding the radio waves. RFID tags work in two different modes: they wake up when they receive a radio wave signal and reflect it (Passive Mode) or they emit their own signal (Active Mode). The tags store information which allows univocally identifying something or someone. That information is stored in an IC (Integrated Circuit) which is connected to an antenna, responsible for transmitting the information. An evolution of this technology is the Near Field Communication (NFC). It consists of a contactless Smart Card technology, based in short-range RFID. Currently, there are mobile phones with NFC embedded in such a way that they work both as a tag and as a NFC reader. These technologies will be widely available both in mobile phones and other devices (e.g. personal digital assistants, etc.) in the near future allowing us to get closer to a ubiquitous and pervasive world. Chapter 45, RFID and NFC in the Future of Mobile Computing describe the most important aspects of RFID and NFC technology, illustrating their applicative potential, and provides a vision of the future in which the virtual and real worlds merge together as if an osmosis took place. Chapter 46, A Multi-Loop Development Process for a Wearable Computing System in Autonomous Logistics, examines a multi-loop development process for a wearable computing system within a new paradigm in logistic applications. The implementation of this system will be demonstrated by an example from the field of autonomous logistics for automobile logistics. The development process is depicted from

lxxvi

selecting and combining hardware through to the adjustment to both user and operative environment. Further, this chapter discusses critical success factors like robustness and flexibility. The objective is to present problems and challenges as well as a possible approach to cope with them. The next chapters, that form Section 3 Critical Success Factors, address the critical success factors affecting the full exploitation of the potential of the services and applications of mobility and computing. A proper development of computing which penetrate our society more thoroughly with the availability of broadband services is provided by varied cooperative networks. However, the success of social dimension of computing requires collaboration within a multicultural environment to be considered. Aim of Chapter 47, Collaboration within Social Dimension of Computing: Theoretical Background, Empirical Findings and Practical Development, is to analyze collaboration within the social dimension of computing on the pedagogical discourse. The meaning of the key concepts of social dimension of computing, collaboration and its factors is studied within the search for the success of social dimension. The manuscript introduces the study conducted within the Baltic Summer School Technical Informatics and Information Technology in 2009. The conducted explorative research comprises four stages: exploration of the contexts of collaboration, analysis of the students’ needs (content analysis), data processing, analysis and data interpretation, and analysis of the results and elaboration of conclusions and hypothesis for further studies. Many factors converge when attempting to define the most adequate mobile learning model to be applied in a face-to-face university environment. As far as innovation related processes go, the implementation of mobile learning, implies defining a road map on the basis of strategic planning. It is also important to apply an action research approach in the implementation process of the model. In analyzing in depth this innovative mobile learning process, there are key factors to consider. First, there are factors related to the technology necessary for the implementation of the model—both hard and soft requirements. Second, there are cultural issues related to the use of non-native internet professors of innovative technologies. Finally, there are challenges related to defining, exactly, those educational strategies to be handled through mobile devices. Chapter 48, Critical Factors In Defining The Mobile Learning Model. An Innovative Process For Hybrid Learning At The Tecnologico De Monterrey, A Mexican University, focuses on the critical factors involved in integrating mobile learning into a hybrid educational model at a Mexican university. The purpose of Chapter 49, Critical Human Factors on Mobile Applications for Tourism and Entertainment, is to research some principles that can guide the design, development and marketing of mobile applications, with a particular focus on the tourism and entertainment application domains. This research also fills a gap concerning impact studies of mobile applications, since the majority of the literature available today is more focused on the design and development process and results. Besides describing a set of novel mobile applications, we aim at providing an overview on innovation processed used, and conducting several experiences, gathering results from questionnaires, surveys, log data and our own observations. Regarding the mobile tourism domain, the author studied the impact of media visibility, the impact of novel interaction paradigms. Regarding the mobile entertainment applications, he focused on studying the impact brought that realism and graphics quality have on mobile games. Internet surveys offer important advantages over traditional survey methods: they can accomplish large samples within a relatively short period of time, questionnaires may have visual and navigational functionalities impossible to implement in paper-and-pencil questionnaires, data is more efficiently processed since it already comes in electronic format and costs can be lower. But the use of the Internet for survey purposes raises important concerns related to population coverage, lack of suitable sampling

lxxvii

frames and non-response. Despite its problems, Internet-based surveys are growing and will continue to expand presenting researchers with the challenge of finding the best way to adapt the methods and principles established in survey methodology to this new mode of data collection in order to make best use of it. Chapter 50, Internet Surveys: Opportunities and Challenges, describes the positive features of the Internet for survey activity and examines some of the challenges of conducting surveys via the Internet by looking at methodological issues such as coverage, sample selection, non-response and data quality. The purpose of Chapter 51, EPS A Coalition-Based Production Approach, is to provide broad view of the rationale, fundamental principles, current developments and applications for Evolvable Production Systems (EPS). Special attention is given to how complexity is handled, the use of agent based and wireless technology, and how economical issues are affected by having an evolvable system. The rationale for EPS is based on current road mapping efforts, which have clearly underlined that true industrial sustainability requires far higher levels of system autonomy and adaptivity than what can be achieved within current production system paradigms. Since its inception in 2002 as a next generation of production systems, the EPS concept has been further developed and tested to emerge as a production system paradigm with technological solutions and mechanisms that support sustainability. Technically, EPS is based on the idea of using several re-configurable, process-oriented, agent-based and wireless intelligent modules of low granularity. This allows for a continuous adaption and evolution of the production system and the ability to explore emergent behavior of the system, which are imperative to remain fit with regards to the system environment. The proposal of new business models, based on the emerging novelty and huge opportunities of mobility and computing is the theme of the next four chapters, which compose Section 4 New Business Models. As mobile applications increase in popularity, the issue of how to build viable business models for the m-commerce industry is becoming a clear priority for both organizations and researchers. In order to address this issue, this chapter, Viable Business Models for M-Commerce: The Key Components, reports on five mini cases used as a guideline, and applies the theoretical business model from Chesbrough and Rosenbloom (2002) to each of them to find out the most important components of viable business models for their m-commerce applications. The study then uses cross cases analysis as a research tool to compare and contrast each of the mini cases and to find out how the different organizations fit within the researched theoretical business model. Finally, this chapter confirms that there are 7 important components of viable business models for m-commerce which are: value proposition, market segment, value chain, profit potential, value network, competitive strategy and firm capabilities. This study also highlights the fact that the public visibility of these 7 components is uneven. Some components such as value proposition, value chain, value network and firm’s capabilities are more likely to be presented in public by organizations. However, aspects such as cost structure and profit potential, market segment and competitive strategy are more likely to be hidden from the public due to their commercial sensitivity. Mobility is a relatively recent topic in the enterprise arena, but thanks to the widespread use of cell phones it has already changed much of the business landscape. It should be integrated in enterprise architectures (EAs) as an intrinsic feature and not as an add-on or as an afterthought transition. Current EA frameworks were not designed with mobility in mind and are usually based on the process paradigm, emphasizing functionality. Although the issue of establishing a systematized migration path from a nonmobile EA to a mobile one has already been tackled, the need for mobile-native EA modeling frameworks is still felt. Chapter 53, A Service-Based Framework to Model Mobile Enterprise Architectures, presents and discusses a resource-based and service-oriented metamodel and EA framework, in which mobility is introduced naturally from scratch, constituting the basis for some guidelines on which EA resources

lxxviii

should be mobilized. Several simple scenarios are presented in the light of this metamodel and framework. Although mobile phones have become an extension of the workplace, organizations are still exploring their effectiveness for employee training and development. A 2009 joint collaborative study between Columbia University (New York, USA) and IBM of 400 IBM employees’ use of mobile phones revealed unexpected insights into how employees use mobile applications to improve job performance. The findings are reshaping IBM Learning's mobile technologies strategy for networking, collaboration, and skills improvement. Research-Based Insights Inform Change in IBM M-Learning Strategy reveals the study's results and IBM's new direction for m-learning, highlighting IBM’s preparedness for a shift in its organizational learning model potentiated by ubiquitous access and mobility. Location-based mobile services (LBMS) are at present an ever growing trend, as found in the latest and most popular mobile applications launched. They are, indeed, supported by the hasty evolution of mobile devices capabilities; by users demand; and, lastly, by market drive. With e-commerce, products and services started arriving to potential customers through desktop computers, where they can be bought and fast delivered to a given address. However, expressions such as “being mobile”, “always connected”, “anytime anywhere” that already characterize life in the present will certainly continue to do so in the near future. Meanwhile, mobile devices centred commerce services seem to be the next step. Therefore, Chapter 55, Location Based E-commerce System: An Architecture, presents a system architecture designed for location-based e-commerce systems. These systems, where location plays the most important role, enable a remote products/services search, based in user parameters: after a product search, shops with that products are returned in the search results and are displayed in a map, around the user present location; and services like obtaining more information, reserving and purchasing are made available as well. This concept represents a mix between traditional client-oriented commerce and faceless mass-oriented e-commerce, enabling a proximity-based user-contextualized system, being well capable of conveying significant advantages and facilities to both service-providers/retailers and users. The four chapters of Section 5, Security, address the topic of security in mobile systems and applications. Vehicular ad-hoc networks (VANETs) are a promising communication scenario. Several new applications are envisioned, which will improve traffic management and safety. Nevertheless, those applications have stringent security requirements, as they affect road traffic safety. Moreover, VANETs face several security threats. As VANETs present some unique features (e.g. high mobility of nodes, geographic extension, etc.) traditional security mechanisms are not always suitable. Because of that, a plethora of research contributions have been presented so far. Chapter 56, Overview of Security Issues in Vehicular Ad-hoc Networks, aims to describe and analyze the most representative VANET security developments. Access control in the domain of information system security refers to the process of deciding whether a particular request made by a user to perform a particular operation on a particular object under the control of the system should be allowed or denied. For example, the access control component of a file server might have to decide whether user “Alice” is allowed to perform the operation “delete” on the object “document.txt”. For traditional access control this decision is based on the evaluation of the identity of the user and attributes of the object. The novel idea of location-aware access control is also to consider the user’s current location which is determined by a location system like GPS. The main purpose of Chapter 57, Modelling of Location-Aware Access Control Rules, is to present several approaches for the modeling of location-aware access control rules. The aim of Chapter 58, Secure Techniques for Remote Reconfiguration of Wireless Embedded Systems, is to give a thorough overview of secure remote reconfiguration technologies for wireless embedded systems, and of the communication standard commonly used in those systems. In particular, we focus

lxxix

on basic security mechanisms both at hardware and protocol level. We will discuss the possible threats and their corresponding impact level. Different countermeasures for avoiding these security issues are explained. Finally, we present a complete and compact solution for a service-oriented architecture enabling secure remote reconfiguration of wireless embedded systems, called the STRES system. The evolution of computer networking is moving from static wired networking towards wireless, mobile, infrastructureless, and ubiquitous networking. In next-generation computer networks, new mobility features such as, seamless roaming, vertical handover, and moving networks are introduced. Security is a major challenge in developing mobile and infrastructureless networks. Specific security threats in next-generation networks are related to the wireless access mediums, routing, and mobility features. Chapter 59, Secure Routing and Mobility in Future IP Networks, identifies these threats, and discuss the state of the art of security research and standardization within the area, proposing security architectures for mobile networking. A survey of security in routing is provided with special focus on mobile ad hoc networks (MANETs). The security of currently relevant protocols for management or node and network mobility, Mobile IP (MIP), Network Mobility (NEMO), Mobile Internet Key Exchange (MOBIKE), Host Identity Protocol (HIP), Mobile Stream Control Transmission Protocol (mSCTP), Datagram Congestion Control Protocol (DCCP), and Session Initiation Protocol (SIP), is described. Section 6 Applications, Surveys and Case Studies is focused on the presentation of a wide variety of new applications and solutions, surveys related with adoption of technologies and case studies, along its 18 chapters. In an educational context, technological applications and their supporting infrastructures have been evolved in a way that the use of learning objects is no longer limited to a personal computer, but has been extended to a number of mobile devices. Such evolution leads to the creation of a technological model called m-learning that offers great benefits to education. This educational model resulted in several research projects and some commercial products. Chapter 60, Evaluation of a Mobile Platform to Support Collaborative Learning: Case Study, describes the (re)use of an adapted platform from an API of MLE (Mobile Learning Engine), to create tests, quizzes, forums, SMS, audio, video, mobile learning objects, in combination with a learning platform in a particular setting. MLE is a special m-learning application for mobile phones (a J2ME application) that can access a LMS (Learning Management System) and use most of its activities and resources, and add new, even innovative, activities. With J2ME one can store, use content and learn without the need of further network access and even use interactive questions that can be directly solved on mobile devices. The MLE enables one to use the mobile phone as a constant way of learning. As a consequence it is possible to use every spare time to learn, no matter where we are, providing new opportunities to enhance learning. Wireless ad-hoc and sensor networks are experiencing a widespread diffusion due to their flexibility and broad range of potential uses. Nowadays they are the underlying core technology of many industrial and remote sensing applications. Such networks rely on battery-operated nodes with a limited lifetime. Although, in the last decade, a significant research effort has been carried out to improve the energy efficiency and the power consumption of the sensor nodes, new power sources have to be considered to improve node lifetime and to guarantee a high network reliability and availability. Energy scavenging is the process by which the energy derived from external sources is captured, translated into an electric charge and stored internally to a node. At the moment, these new power sources are not intended to replace the batteries, since they cannot generate enough energy; however, working together with the conventional power sources they can significantly improve node lifetime. Low-power operation is the result of a complex cross-layer optimization process, for this reason, this chapter “Power Issues and

lxxx

Energy Scavenging in Mobile Wireless ad-hoc and Sensor Networks”, thoroughly reviews all the traditional methods aimed at reducing power consumption at network, MAC and PHY levels of the TCP stack, to understand advantages and limitations of such techniques, and to justify the need of alternative power sources that may allow, in the future, the design of completely self-sustained and autonomous sensor nodes. Chapter 62, A Low Cost Wireless Sensors Network with Low-Complexity and Fast-Prototyping, presents a low cost/fast prototyping wireless sensors network that was designed for a huge range of applications and making use of low cost commercial of the shelf components. Such applications includes industrial measurements, biomedical, domestic monitoring, remote sensing, among others. The concept of the wireless sensor network is presented and simultaneously, hot topics and their implementation are discussed. Such topics are valuable tools and can't be discarded when a wireless sensors network is planed. By the contrary, such tools must be taken in account to make the communications between the nodes and the base station the best possible reliable. The architecture, protocols and the reasons that were behind the selection of the components are also discussed. The chapter also presents performance metrics that are related to with the physical characteristics of sensors and with the radio specificity. Microcontrollers with a RISC architecture are used by the network nodes to control the communication and the data acquisition and operate in the 433 MHz ISM band with ASK modulation. Also, in order to improve the communication and to minimise the lost of data, it is predicted to put the wireless nodes to handle line and source coding schemes. Failure detection is an important abstraction for the development of fault-tolerant middleware, such as group communication toolkits, replication and transaction services. Unreliable failure detector (namely, FD) can be seen as an oracle which provides information about process failures. The dynamics and selforganization of Mobile Ad-hoc Networks (MANETs) introduce new restrictions and challenges for the implementation of FDs with which static traditional networks do not have to cope. It is worth mentioning that in some way, fault tolerance is more critical for MANETs than for the latter, since wireless network can present high error rates and mobile nodes are more prone to failures, physical damages or transient disconnections. The aim of Chapter 63 Unreliable Failure Detectors for Mobile Ad-hoc Networks is thus to discuss the impact of all these characteristics, intrinsic to MANET, in the implementation of FDs. It presents a survey of the few works about FD implementations for wireless networks, including the different possible assumptions to overcome the dynamics and lack of both global view and synchrony of MANETs. Adaptation of communication is needed to maintain the connectivity and quality of communication in group-wide collaborative activities. This becomes quite a challenge to handle when mobile entities are part of a wireless environment, in which responsiveness and availability of the communication system are required. In Chapter 64, Mission-Aware Adaptive Communication for Collaborative Mobile Entities, these challenges are addressed within the context of the ROSACE project where mobile ground and flying robots have to collaborate either between themselves or with remote artificial and human actors during save and rescue missions in the event of disasters such as forest fires. This chapter presents our first results. The final goal is to propose new concepts, models and architectures that supports cooperative adaptation which is aware of the mission being executed. Thus, the communication system can be adequately adapted in response to predictable or unpredictable evolutions of the activity requirements and to the unpredictable changes in the communication resource constraints. In the last years ontologies are being used in the development of pervasive computing applications. It is habitual their use for facilitating the interoperability among context-aware applications and the entities

lxxxi

that may enter in the context at any time. Chapter 65 OntoHealth: An Ontology Applied to Pervasive Hospital Environments presents OntoHealth: an ontology applied to health pervasive environment and a tool to its processing. The main idea is that a hospital could be seen as this pervasive environment, where someone, through ubiquitous computing, engages a range of computational devices and systems simultaneously, in the course of ordinary activities, and may not necessarily even be aware that they are doing so. With the proposed ontology and the tool for its processing, the medical tasks can be shared by all components of this pervasive environment. Chapter 66 Adoption of Mobile and Information Technology in an Energy Utility in Brazil deals with the adoption of mobile technology. The case illustrated here is the implementation of mobile and wireless technology – MIT and smartphones – at an energy utility. The objective was to understand the human and social aspects of the adoption of this technology. This paper makes use of the metaphor of hospitality proposed by Ciborra in the late 1990s. The hospitality metaphor was a useful alternative for describing the process of adopting a new technology. It touches on technical aspects and notes human reactions that become evident when a technician comes across an unknown ‘guest’, the new technology: the doubtful character of the guest, the reinterpretation of the identities of guest and host during the process, learning through trial and error, the technology’s ‘drift’, the participants’ emotions and state of mind, and the appropriation of, and the care for, the new technology. Context-aware mobile applications are becoming popular, as a consequence of the technological advances in mobile devices, sensors and wireless networking. Nevertheless, developing a context-aware system involves several challenges. For example, what will be the contextual information, how to represent, acquire and process this information and how it will be used by the system. Some frameworks and middleware have been proposed in the literature to help programmers to overcome these challenges. Most of the proposed solutions, however, neither have an extensible ontology-based context model nor uses a communication method that allows a better use of the potentialities of the models of this kind, as discussed in Chapter 67 Infrastructures for Development of Context-Aware Mobile Applications. Chapter 68 A Practice Perspective on Transforming Mobile Work introduces a study that explores users’ experiences during an organizational implementation of a new mobile information technology in public home care environment. The home care case illustrates differences between implementation project goals and expectations, and on the other hand, the daily organizing and carrying out care work, where previously, no information technology was utilized. While implementing mobile technology was expected to enhance the efficiency of care working, the project outcomes include resistance due to surveillance aspect of the new technology as well as technological problems during the implementation. Successful outcomes of the implementation include better planning of working hours and more even distribution of work resources. In Chapter 69, Data Replication Support for Collaboration in Mobile and Ubiquitous Computing Environments, the authors address techniques to improve the productivity of collaborative users by supporting highly available data sharing in poorly connected environments such as ubiquitous and mobile computing environments. We focus on optimistic replication, a well known technique to attain such a goal. However, the poor connectivity of such environments and the resource limitations of the equipments used are crucial obstacles to useful and effective optimistic replication. The authors analyze state-of-the art solutions, discussing their strengths and weaknesses along three main effectiveness dimensions: (i) faster strong consistency, (ii) with less aborted work, while (iii) minimizing both the amount of data exchanged between and stored at replicas; and identify open research issues.

lxxxii

Wireless Local Area Networks are very useful for the most applications based on network. It can be developed in almost all environments and products are cheap and robust. Moreover, these networks can be formed by different devices with wireless interfaces like IP cameras, laptops, PDAs, sensors, etc. WLANs provide high bandwidth at large coverage areas, which it is necessary in many applications at different research areas. All these characteristics let WLANs be a useful technology to provide ubiquity for any type of service. If they are deployed from a good and exhaustive design, they can provide connection to any device, everywhere at anytime. Chapter 70, Providing Outdoor and Indoor Ubiquity with WLANs, presents a complete guideline about how to design and deploy WLANs and to get their best performance. The authors start from an analytical point of view and we use mathematical expressions to design WLANs in both indoor and outdoor environments. Then, the authors introduce a method proposed by some authors of this chapter some years ago and how it can be used to design WLANs in indoor environments. The authors also show WLANs design in outdoor environments and describe two projects developed in order to provide ubiquity in real indoor and outdoor environments. Chapter 71, INTIC for Mobile Devices: Support System for Communication with Mobile Devices for the Disabled introduces the In-TIC system for mobile devices Integration with ICT system for mobile devices) that represents an approach towards the area of technical aids for mobile devices. The mobile telephone is a device that makes our lives easier, allowing us to be permanently accessible and in contact, to save relevant information, and also for entertainment purposes. However, people with visual, auditory or motor impairment or the elderly still find these devices difficult to use. They have to overcome a range of difficulties when using mobile telephones: the screens are difficult to read, the buttons are too small to use, and the technical features are too complicated to understand. At present, the main advances in mobile technology have been aimed at improving multimedia-messaging services and reproducing videos and music. This new support system adds accessibility to mobile telephones, making them easier to use for the people who need them the most, people with reduced physical or mental capacities who cannot use a conventional mobile. Granxafamiliar.com is a project for developing the Galician rural milieu both socio-economically and culturally in order to appreciate the quality of life and rural culture, to create communication links between the rural and urban world, to emphasize the importance of the traditional self-supply production market of Galician family farms, and to promote the spread of new technologies as a social intervention tool against the phenomenon of social and territorial exclusion known as the “Digital Divide". The authors of New Ways to Buy and Sell: An Information Management Web System for the Commercialization of Agricultural Products from Family Farms without Intermediaries are planning the architecture of www.granxafamiliar.com, which is developing the creation of a virtual community based on boosting commercial transactions and the possibilities for buying and selling traditional self-supply products that exist in the rural environment. The authors expect to promote it globally across the Internet by promoting the use and spread of ICTs as tools and commercial channels for agricultural products and intend to carry out an in-depth empirical and theoretical study of the territorial and social effects linked to the development of the information and communication society in rural communities. Their aim is to assist the progress of public decision-making and administrative efficiency for when the time comes to invest in suitable services and activities related to the information society in the rural environment. The continuous growth of the available throughput, especially in the uplink of mobile phone networks is opening the doors to new services and business opportunities without references in the past. In more concrete, new standards HSDPA/HSUPA, introduced to complement and enhance 3G networks, together with the advances in audio and especially video coding, like those adopted by the standard H.264 AVC

lxxxiii

have boosted the appearance of a new service: exploiting the mobile telephony networks for contributing broadcast quality videos. This new service is offering just now a low cost, high flexibility alternative that, in a brief period of time, will substitute the current Electronic News Gathering (ENG) Units, giving rise to what is being to be called Wireless Journalism (WENG or WiNG). Chapter 73, Broadcast Quality Video Contribution in Mobility, discusses both the technologies involved and the business opportunities offered by this sector. Once reviewed the state of the art, different solutions will be compared, some of them recently appeared as commercial solutions, like QuickLink 3.5G Live Encoder or AirNow! and others still in research and development processes. Mobile devices are rapidly becoming the most common interface for accessing network resources (Hall 2008). By 2015 the average 18-year old will spend the majority of their computing time on mobile devices (Basso 2009). These trends directly affect institutions of higher learning. Many universities are offering learning initiatives and m-services designed to distribute content and services to mobile devices. Chapter 74, Mobile Device Selection in Higher Education: iPhone vs. iPod Touch, reports findings from an exploratory, longitudinal study at Abilene Christian University, where incoming freshmen received their choice of an Apple iPhone or iPod touch. The findings indicate that users' device selections were affected by their perceptions of the costs of the devices, the devices' relative characteristics, and the social influence of parents. We also found that users' attitude, satisfaction, and confidence about their device selection varied across devices, with iPhone users having more favorable perceptions. The chapter concludes with recommendations for mobile learning initiatives and directions for future research. Chapter 75, Design of Wearable Computing Systems for Future Industrial Environments, investigates the role of context, particularly in future industrial environments, and elaborates how context can be incorporated in a design method in order to support the design process of wearable computing systems. The chapter is initiated by an overview of basic research in the area of context-aware mobile computing. The aim is to identify the main context elements which have an impact upon the technical properties of a wearable computing system. Therefore the authors describe a systematic and quantitative study of the advantages of context recognition, specifically task tracking, for a wearable maintenance assistance system. Based upon the experiences from this study, a context reference model is proposed, which can be considered supportive for the design of wearable computing systems in industrial settings, thus goes beyond existing context models, e.g. for context-aware mobile computing. The final part of this chapter discusses the benefits of applying model-based approaches during the early design stages of wearable computing systems. Existing design methods in the area of wearable computing are critically examined and their shortcomings highlighted. Based upon the context reference model, a design approach is proposed through the realization of a model-driven software tool which supports the design process of a wearable computing system while taking advantage of concise experience manifested in a well-defined context model. In 2002, Belgium has adopted an electronic identity card as one of the first countries in Europe. By the end of 2009, the roll-out of the eID card will be completed, and this means that each Belgian citizen will possess an eID card. The card enables her to digitally prove her identity and to legally sign electronic documents. The Belgian eID card opens up new opportunities for the government, its citizens, service providers and application developers. The Belgian eID technology originally aimed at facilitating transactions between Belgian citizens and the government and although many eID applications have been developed, the success of the Belgian eID technology has not been what was expected. Therefore, the Belgian government encourages developers to build commercial applications that use the eID card (for authentication or e-signatures). However, extending the scope of the Belgian eID technology from

lxxxiv

e-government to the commercial sector is no sinecure and not without risks. These issues are analysed in Extending the Scope of eID Technology: Threats and Opportunities in a Commercial Setting. Mobile information work, an extreme type of information work, is progressively becoming commonplace in various corporations. The availability of cheap and portable information technologies as well as the development of pervasive communication infrastructure in some parts of the world is creating scenarios where people can work from almost anyplace. Nevertheless up to now there has not been sufficient research on the particular work practices and strategies these professional workers use to be productive as they face the particular challenges of being mobile. Based on an ethnographic investigation of the experiences of mobile professional workers in a multi-national accountancy company, Chapter 77 Mobility and Connectivity: On the Character of Mobile Information Work, discusses some characteristics defining the character of modern information work with regards mobility and connectivity while operating outside the workplace. The study highlights the importance of: location in terms of providing an adequate atmosphere and infrastructure to conduct work; regularity in terms of giving workers flexibility to connect and reconnect whenever it was more convenient for them; space in terms of letting people preserve and reconstruct their information workspaces; and balance while juggling between personal and work related commitments. The findings presented can be useful for defining the processes and technological tools supporting mobile professional workers.

EXPECTATIONS Along these 77 chapters, the reader is faced with the discussions and confirmation of the relevance and impact of this hot topic, providing professionals, researchers and scholars, with some of the most advanced research developments, solutions, state-of-the-art enabling technologies, discussions and case studies on mobility and computing. The handbook is expected to support a professional audience of top managers, IT professionals, technology solution providers and an academic audience (teachers, researchers and students, mainly of post-graduate studies). As an academic tool, it can be a support to disciplines of post-graduate studies on IT/IS. We hope you find it useful. Enjoy your reading and study! Maria Manuela Cruz-Cunha Polytechnic Institute of Cávado e Ave, Portugal Fernando Moreira Universidade Portucalense, Portugal

lxxxv

Acknowledgment

Editing a book is a quite hard but compensating and enriching task, as it involves a set of different activities like contacts with authors and reviewers, discussion and exchange of ideas and experiences, process management, organization and integration of contents, and many other, with the permanent objective of creating a book that meets the public expectations. And this task cannot be accomplished without a great help and support from many sources. As editors we would like to acknowledge the help, support and believe of all who made possible this creation. First of all, the edition of this book would not have been possible without the ongoing professional support of the team of professionals of IGI Global. We are grateful to Dr. Mehdi Khosrow-Pour and to Mrs. Jan Travers, Managing Director, for the opportunity and belief in this project. A very very special mention of gratitude is due to Mrs. Christine Bufton, Assistant Development Editor, and to Mr. Dave DeRicco and Mr. Michael Killian, Editorial Assistants, for their highly professional support and friendly words of advisory, encouragement and prompt guidance. We also address our recognition and appreciation to all the staff at IGI Global, whose contributions throughout the process of production and making this book available all over the world was invaluable. We are grateful to all the authors, for their insights and excellent contributions to this book. Also we are grateful the authors who simultaneously served as referees for chapters written by other authors, as well as to the external referees and to the members of the editorial advisory board, for their insights, valuable contributions, prompt collaboration and constructive comments. Thank you all, authors and reviewers, you made this book! The communication and exchange of views within this truly global group of recognized individualities from the scientific domain and from industry was an enriching and exciting experience! We are also grateful to all who accede to contribute to this project, some of them with high quality chapter proposals, but unfortunately, due to several constraints could not have seen their work published. Thank you. Maria Manuela Cruz-Cunha Polytechnic Institute of Cávado e Ave, Portugal Fernando Moreira Universidade Portucalense, Portugal

Section 1

Mobile Technologies

1

Chapter 1

Evaluating the Context Aware Browser:

A Benchmark for Proactive, Mobile, and Contextual Web Search Davide Menegon University of Udine, Italy Stefano Mizzaro University of Udine, Italy Elena Nazzi University of Copenhagen, Denmark Luca Vassena University of Udine, Italy

ABSTRACT The authors discuss the evaluation of highly interactive and novel context-aware system with a methodology based on a TREC-like benchmark. We take as a case study an application for Web content perusal by means of context-aware mobile devices, named Context-Aware Browser. In this application, starting from the representation of the user’s current context, queries are automatically constructed and used to retrieve the most relevant Web contents. Since several alternatives for query construction exist, it is important to compare their effectiveness, and to this aim we developed a TREC-like benchmark. We present our approach to early stage evaluation, describing our aims and the techniques we apply. The authors underline how, for the evaluation of context-aware retrieval systems, the benchmark methodology adopted can be an extensible and reliable tool. DOI: 10.4018/978-1-60960-042-6.ch001 Copyright © 2011, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited.

Evaluating the Context Aware Browser

INTRODUCTION The diffusion of mobile devices and real-world mobile users have moved the static world of classical and Web Information Retrieval (IR) towards a dynamic and evolving context-based world. The notion of context (roughly described as the situation the user is in), and the information it conveys, is gaining increasing importance for the development of new IR systems. The concepts of context and awareness convey to the dynamic nature of the user needs, to the complexity of the information available, and of the relevance of this information. When combined with context-awareness, IR has been named Context-Aware Retrieval (CAR) (Brown, 2001). Starting from considering only a low number of contextual features (location and time), current CAR systems entail such an amount of data that a new challenge for IR is how those data can enhance user satisfaction. How to evaluate the strategies and techniques that CAR systems use for this purpose is another challenge. CAR systems imply a high amount of interactivity with the user, and a user study seems the most sensible approach. But our approach is to evaluate highly interactive and novel contextaware systems on the basis of a TREC-like benchmark methodology. This paper describes our methodology, presents the results obtained, and discusses the general approach. The present work is related to the ContextAware Browser (CAB for short), a novel contextaware retrieval application. CAB allows a proactive context-aware Web content perusal by means of mobile devices. Since several alternatives for the retrieval process exist, it is important to compare their effectiveness to find the best approach. With this in mind, we propose an evaluation benchmark, discuss its limits, and test it. Although we focus on the needs of CAB, the problems to solve are typical of any proactive context-aware retrieval system.

2

We first briefly survey evaluation methodologies in IR and in CAR systems (Sect. Related Work), introducing our case study application. We then present our early evaluation approach (Sect. Experimental Evaluation) describing aims, techniques, and results. We discuss the reliability and usefulness of our methodology in Sect. Discussion, while in Sect. Lessons Learned we present the lessons learned. Finally in Sec. Conclusions we draw some conclusion and present future work.

RELATED WORK Context-Aware Retrieval With the spread of the concepts related to context-aware computing, Information Retrieval has gained new and increasing importance. The newborn field of CAR, instead of concentrating only on topicality, incorporates contextual information into the retrieval process, aiming at discovering “the query behind the context”: to retrieve what the users need, even if they did not issue any query (Mizzaro, 2008). CAR systems are concerned with the acquisition and understanding of context, and with a behavior based on the recognized context. Thus the CAR model includes, among the elements of the classical IR model, the user’s context. This context is both used in the query formulation process and associated with the documents candidate for retrieval. Typical CAR applications present the following characteristics (Jones, 2004): a mobile user, i.e., a user whose context is changing; interactive or automatic actions, if there is no need to consult the user; time dependency, since the context may change; appropriateness and safety to disturb the user. Although CAR applications can be both interactive and proactive in their communication with the user, we concentrate on the proactive aspects, since they are more relevant to our proposal. Besides, we concentrate on the association between CAR and mobile application, as they

Evaluating the Context Aware Browser

can be considered as the prime field for CAR (Jones, 2004).

The Context-Aware Browser Idea & Architecture The Context Aware Browser (CAB) is a generalpurpose solution to Web content fruition by means of context-aware mobile devices. It allows a “physical browsing”: browsing the digital world based on the situations in the real world. The main idea behind CAB is to empower a generic mobile device with a browser able to automatically and dynamically load web pages, services and applications selected according to the current context the user is in. Despite the apparent simplicity of this approach, a more thorough definition has to take into account several features; whence, we can say that the Context-Aware Browser is best described by the sum of the following parts: •

•

•

•

a Web browser: CAB is able to interpret and render in a sound way (X)HTML code, to interpret client side (e.g., JavaScript) code, and to fully exploit AJaX-based applications; a context-aware application: CAB is able to automatically retrieve and constantly update the contextual information gathered from the surrounding environment (this feature allows CAB to provide contents varying in dependence of the current context of the user); a search engine: CAB is able to search both for “traditional” web pages and applications on the Web and for specifically tailored applications, as we will see in the following; a proactive application able to automatically search for and download contents: the resources retrieved by the search engine are automatically filtered against user’s

•

preferences and several other parameters, in order to reduce the cognitive load imposed on the user by automatically selecting the most appropriate contents; an application working on any kind of Web content, i.e., both Web pages and applications: CAB is able to manage both static resources (e.g., plain (X)HTML pages) and dynamic ones requiring user interaction (e.g., web applications).

The CAB acquires information related to the user and the surrounding environment, by means of sensors installed on the device or through external servers. This information, combined with the user’s personal history and the community behavior, is exploited to infer the user’s current context (and its likelihood) that is represented by a list of terms. In the subsequent retrieval process, starting from this list of terms, a query is automatically built and sent to an external search engine, in order to find the most suitable Web pages for the sensed context. In this paper we study different strategies to build automatically the query sent to the external search engine. We provide some more details on CAB implementation. Figure 1 shows CAB overall three-tiered architecture, where the topmost layer manages the interaction with the user, the medium layer bridges the topmost and the inner layer, which in turn is responsible for sensors, inferential network, and filtering engine management. The main modules of CAB architecture are: •

Context server and Sensors. They are responsible for collecting and sending to the CAB Core the contextual information gathered from the surrounding environment, in the form of an inferential network that represents (part of) the knowledge base of the Context manager. Then, all the sensors specified are instantiated by the Context manager. Such sensors communicate with the Context server in order to re-

3

Evaluating the Context Aware Browser

Figure 1. CAB architecture

•

•

4

ceive the contextual information (location, date/time, temperature, etc.), in either push or pull mode. Context manager, and its internal inferential system are the core component responsible for the synthesis of the current context, starting from the contextual information gathered by sensors. Filter module and the Descriptor search engine are in charge to search for and filter the most suitable contents according to the

inferred context. The contextual information, obtained from the Context manager, is represented through a Context Descriptor that holds the most important information related to the user’s current context. CAB allows the user to specify which filtering information is public and which one is private to achieve a reasonable privacy level. Public information is used to formulate the query to be sent to the Descriptor search engine out of the device, while private in-

Evaluating the Context Aware Browser

•

•

formation is used in “internal” filtering activities only. Connector is an infrastructure component used to separate the user interface from the context management part. This abstraction allows to easily change the way users interact with the system, or to modify the techniques exploited for the current context inference and management. Currently, as the user interface is Web-browser based, the Connector is implemented as a local web server, being able to open HTTP connections and to manage browser’s requests. Browser module manages the presentation of the web contents to the user and it also automatically starts the most relevant ones. Such features are implemented in two submodules: the AJaX engine and the User interface.

Scenario To better understand the CAB functioning we present a simple example scenario, where we consider a tourist visiting a museum. As she enters the museum main hall with her own mobile device, a Web page with the list of the collections available in the museum is presented to her (e.g., the official museum home page, or a mobile adapted version). While the user action has simply been to enter the museum, the CAB has received, analyzed, and collected contextual information and then retrieved the associated resources. Moving around the museum the tourist is guided by the system that contextually retrieves the most interesting information (Figure 2): moving towards an artwork the tourist will automatically obtain the detailed description of that artwork. Because of the huge amount of information that is relevant to each artwork, the ubiquitous search has to be precise and has to provide only the information interesting and important for the tourist, based on her preferences, activities, interests and based on the community behavior

and on the accuracy of the same information. In fact, when the tourist comes near an artwork, the detailed description automatically provided, is accompanied with pictures, comments by experts or other tourists, etc. As usual, the most interesting artworks are surrounded by lots of people. In this case, since the user has to wait in the queue, she cannot yet see the painting, so probably she is not particularly interested in the detailed description. In this situation the system detects the crowd situation and retrieves different resources, like for example Web pages with podcasts about that artwork. If the tourist takes into consideration those resources the system annotates them with the tourist’s context; hence, other users arriving near the painting in crowded situations, are more likely to get the podcast or the text description otherwise.

Evaluation In the development of a CAR system for mobile environments, evaluation plays an important role, as it allows to measure the effectiveness of the system and to better understand problems from both the system and the user interaction point of view. The challenges for context-aware computing evaluation have been studied deeply (Scholtz, 2001; Carter, 2004). In particular these challenges are: the need to identify meaningful metrics (each single application has peculiar aspects to be evaluated that differ from other types of applications); the difficulty to evaluate in small scale a system that is meant to be adopted by many users; the explicit testing of a novel system, meant to be integrated in everyday life and thus invisible, as the basic vision of Ubiquitous computing (Weiser, 1991). Depending on resources and aims, different evaluation metrics are chosen and different evaluation approaches are adopted (e.g., user studies and benchmark; lightweight and heavyweight (Harper, 1997); etc.). As CAR applications are strictly related to users, the user-centered evaluation (live or in laboratory) seems the most natural

5

Evaluating the Context Aware Browser

Figure 2. CAB employed in a museum (iPhone and Android version) (a) List of available information and services;(b) Detailed contextual information

one. In (Scholtz, 2004), for example, a framework for user evaluations of ubiquitous computing has been proposed. A detailed work about metrics for evaluating systems for information access has been done in (Scholtz, 2006). More recently, in AmbieSense (Göker, 2008), a user centered, iterative, and progressive evaluation has been adopted combining IR evaluation methods with humancomputer interaction development techniques. Another largely adopted evaluation methodology is benchmark evaluation. Taking example from TREC initiative for large scale evaluation of IR effectiveness (http://trec.nist.gov/), benchmarks are defined by a collection of topics (expressions of user’s information needs), a collection of resources to retrieve, and a set of relevance judgments about those resources for each topic. GeoCLEF (http:// www.uni-hildesheim.de/geoclef/) is an example of

6

a benchmark for Geographical IR (GIR) system evaluation. It is based on CLEF (http://www. clef-campaign.org/) and it has the aim to provide a framework for evaluation of systems with search tasks involving both spatial and multilingual aspects. Its importance is given by the large collection of topics and documents, but its focus is on users with geographical related tasks, not really on mobile user, in mobile environments. Within the CAR field, a benchmark evaluation named MREC (MoBe Retrieval Evaluation Collection) (Mizzaro, 2008) has been proposed to evaluate MoBe (Bulfoni, 2008) a framework for CAR of mobile applications. Although it is usually considered the most appropriate for CAR applications, the user evaluation approach has also some, not negligible, drawbacks, when compared with benchmarks. First it is an

Evaluating the Context Aware Browser

Figure 3. A (part of a) context descriptor

high level evaluation, where the primary interest is to study how the system satisfies the user’s needs, rather than on how the system serves the information needs of its users. Benchmark evaluations, on the contrary, are system centered: they directly focus on the evaluation of implementation details, and they can evaluate different aspect of the retrieval process. Second a mature prototype to test has to be available, complete of an effective user interface. This contrasts with the purposes of developers, and it forces significant implementation decisions uncovered by evaluations. Moreover user evaluation is more complex to perform, requires much more time than benchmarks, and it is more dependent on users’ subjectivity.

EXPERIMENTAL EVALUATION Although CAB application development is in an advanced phase, its retrieval mechanism needs a more accurate implementation. An early evaluation of the strategies we would like to implement is needed, so, considering the successful approach of MREC and the useful insights obtained, we adopted again a TREC-like benchmark evaluation, named CREC (CAB Retrieval Evaluation Collec-

tion). CREC focuses on the retrieval of Web pages and starts from the results obtained with MREC. We were interested in the following questions: how to build in an automatic way the queries to be sent to the external engine? Which is the best strategy in terms of effectiveness? How does the retrieval effectiveness change on the basis of the increase/decrease of the number of terms in the query, or of different kinds of terms exploited in the query? How effective is an automatic query formulation compared to a user manual search?

CREC: An Incremental Benchmark CREC is constituted by the usual TREC three components: the statements of information needs, a collection of documents, and a set of relevance judgments. The statements of information needs are defined by context descriptors, which represent different examples of user’s contexts in different domains (Figure 3). CREC includes 10 context descriptors, which differ for user activities, location, time, etc.; they have been designed similarly to TREC topics. A unique judge using a four level relevance scale made the relevance judgments: relevant, partially relevant, not classified, or not relevant.

7

Evaluating the Context Aware Browser

The and fields in the context descriptor helped the judgments. The documents collection consisted of Web pages. Due to the evolving behavior of the Web (pages are dynamically added, removed, or modified), we opted for a dynamic collection evolving during the tests. Moreover, and more importantly, if a new implementation of the CAB external search engine needs to be evaluated, CREC will not contain, in general, all the retrieved pages. Since this would make the evaluation not reliable, in our approach the collection is not static: newly retrieved documents, and their judgments extend the collection. We have built two CREC versions so far. The first version was constructed performing 5 manual queries for each context descriptor, and judging the first 150 single retrieved documents for each context descriptor. Starting from this version of the collection, we adopted an “interactive search and judge” (Cormack, 1998) approach to add more relevant documents. In particular, we ran some queries, automatically built from context descriptors and, for each query, the Web pages that were not already in the collection and were retrieved in the first 10 ranks were added to the collection and judged. This second version of the collection had 3634 total pages: 494 relevant, 596 partially relevant, 34 not classified, and 2510 not relevant. Contrary to a real setting where the number of relevant documents is much lower than the number of documents in the collection, CREC collection had a high number of relevant documents because we searched for relevant documents. This is not a problem since we assume that unjudged documents are not relevant. Moreover we are not interested in all the documents, but just in the first retrieved ones.

Strategies We used CREC to compare four automatic query construction strategies. All of them work on lists of terms extracted automatically from the field

8

Figure 4. The four strategies and a representation of how they combine the term indexes

<description> of the context descriptor (conversely, the human relevance judge uses the whole context descriptor): for instance, the strategies see the context descriptor in Figure 3 as: “user just landed london heathrow international airport looking flight timetable timetable connections london lunch time”. The strategies are based on two main indexes: tf.idf and geoterms (i.e., terms that refer to geographical information - a term is a geoterm if its Wikipedia page contains geographical coordinates). We chose tf.idf, because it is a classical and largely used IR technique, and geoterms because location is probably the contextual dimension that is more informative of user’s current context. These indexes, differently combined, are used to rank the lists of terms according to their importance in each strategy. The four strategies are (Figure 4): 1. tf.idf: all the significant terms in a context descriptor are taken into account for query

Evaluating the Context Aware Browser

Figure 5. nDCG@10 for the five strategies on all contexts

formulation based on the evaluation of their tf.idf value, in descending order. 2. inverse tf.idf: the previous strategy with terms taken in ascending order. This strategy represents a lower bound to use in the comparisons. 3. geoterms + tf.idf: the query exploits first the geoterms and then the remaining terms ordered by their tf.idf value, in descending order. 4. tf.idf + geoterms: similar to the previous strategy, but at first two terms with the highest tf.idf are introduced, then the geoterms, and then again the following tf.idf terms in descending order. The strategies have been implemented in Java using Yahoo! as external search engine through the API provided (http://developer.yahoo.com/ search/web/).

Results For each strategy and context descriptor, 10 queries of different lengths (from 1 to 10 terms) were automatically formulated, incrementally selecting

the first 10 terms of the ranked lists. Thus our query construction is incremental: once a term is in a query, it will remain in longer queries as well. We measured strategies’ effectiveness by means of a standard IR metric, [email protected] emphasizes quality at the top of the ranked list, and takes into consideration the first 10 retrieved items, which is reasonable for CAB, since the user is unlikely to scroll long lists of retrieved items. The manual approach, where users autonomously define their queries, was our upper reference strategy. In particular, for each context descriptor, a human operator has defined 5 queries. Figure 5 compares the four strategies, with the manual one, showing their effectiveness (nDCG@10, on the Y axis) averaged on all 10 contexts, for different query lengths (on the X axis). We notice four main aspects: 1. all the proposed strategies have lower performances than the manual one; 2. apart from the manual one, the most effective strategy is the geo + tf.idf; 3. in general, the performance of long queries is very low. This is probably because CREC contexts are made up by more than one

9

Evaluating the Context Aware Browser

Figure 6. Comparison of manual and geo + tf.idf approach on four contexts

facet (for example different activities), and their descriptors contain terms that relate to these facets. Thus increasing the number of terms in the query, it is more probable to have terms related to different facets, and this could decrease the performance; 4. the manual strategy has a different behavior: its performance tends to increase with query length. If performed by a human, a search session could improve with query refinements that add or change terms on the basis of the knowledge acquired by visualizing results. This is not performed by automatic strategies that construct the query incrementally. This is probably one of the reasons why manual strategy has a performance improvement the more terms are used. Figure 6 shows the comparison of the best automatic strategy and the manual one for four

10

different contexts. We learnt that there was not a stable relationship between the two strategies: in some cases the manual had better effectiveness, in some other had worst effectiveness. In general however there is a clear margin to improve automatic approaches to query construction. Different context descriptors involved a different number of geoterms. Thus, to have a uniform view based on the number of geoterms adopted in the queries, we performed more analysis on the geoterms based strategies. The average results among all contexts are shown in Figure 7; in the graphics both the average, min, max values and the variance are shown. For the geoterms + tf.idf approach, the results increased adding the term with the highest tf.idf value after the geoterms, both with only 1 or 2 geoterms. The same is for the tf.idf + geoterms approach. The difference was that in this strategy, we had low performances using only the first 2 terms with highest

Evaluating the Context Aware Browser

Figure 7. Detailed results for the strategies with different number of geoterms

tf.idf value. Thus, from the results presented, we have understood that terms related to locations were more significant in a query, than the terms ordered based on their tf.idf value. In particular, the maximum performance was obtained adding one tf.idf term after all the geoterms. Moreover we noticed that there was no relationship between geoterms and their value in the tf.idf ordered list: in some cases geoterms have a high tf.idf value, in some other a very low one.

DISCUSSION Benchmark Reliability As a unique judge made the relevance judgments, to verify that subjectivity is not an issue for our benchmark, we performed an additional experiment, involving two more judges. We measured inter-judge agreement on a pool of retrieved pages,

and judgments agreed on average on 65% pages (which became 92% after a discussion between judges). Moreover we used the ADM measure (Della Mea, 2004) to better understand the different judgments. The values in the relevance scale, relevant, partially relevant, not classified, and not relevant, correspond respectively to the values 1, 0.5, 0.25, and 0. We obtained ADM values of 0.757 and 0.846 before the discussion, and 0.950 and 0.981 after the discussion. Figure 8 offers a graphic representation: the primary judge is presented on the X axis and a secondary judge (J1) is on the Y axis; the figure on the left is before the discussion between judges, the one on the right is after the discussion. To understand how our approach to collection construction copes with Web dynamics, we performed another experiment. The two secondary judges manually formulated 3 queries, in relation to a given context descriptor. Then we counted

11

Evaluating the Context Aware Browser

Figure 8. ADM values for the experiment involving the first secondary judge

the pages retrieved by the secondary judges but not retrieved by the automatically constructed queries, and among them the relevant pages. J1 retrieved 63 new never judged pages (21 relevant), while J2 retrieved 56 (34 relevant). This shows that new pages have appeared on the Web at a fast pace or, more probably, the initial collection was not complete. The new retrieved pages were also added to the collection, and the effectiveness evaluation was performed again. The dashed line in Figure 5 shows the effectiveness of geo + tf.idf computed considering the new pages and judgments: it does not change significantly. From the results CREC seems complete and reliable at least to a reasonable extent, and for its aims.

Benchmark Extensibility Like the Web, also our collection is in continuous evolution, thus we performed an additional experiment to understand the effort necessary to extend it. On the basis of a context descriptor not used in the benchmark reliability experiments, two judges

12

(J and a new J3) are taken to build a manual query and then evaluate the retrieved pages (36 pages for person were evaluated). While J3 had never done this experiment before, J is an expert of both the domain, as he was the designer of the contexts descriptors, and of the evaluation procedure, as he already judged pages and he is used to the 4 level relevance scale adopted. The average time needed for the evaluation of one page is 48.9 seconds for J and 62.6s for J3, with a standard deviation of 14.3s for J and 32.3s for J3. The expert then used about 75% of the time needed by the not expert to perform the new judgments. Since the average time to evaluate a page is 55.7s, and since there are 2405 different pages from the first to the second version of our collection, the time to increase the collection was about 37 hours.

LESSONS LEARNED From this research we learned several kinds of lessons. Concerning the CAB development process, the CREC benchmark gave good insights

Evaluating the Context Aware Browser

(e.g., in the best strategy, the best performance appears just adding a term after the “geoterms”) and underlined weak points (e.g., adding more and more terms in the query does not increase performance). A more general outcome concerns the evaluation procedure. We understood that our approach is useful as it can simplify the user testing evaluation. For example, knowing which is the best strategy allows us to give to users one prototype, instead of different prototypes, one for each strategy. Moreover, once the benchmark is configured, it can be reused to test new strategies or related features, in a semi-automatic way (new judgments are needed but the update time is reasonable). Our approach seems interesting and valid for IR applications that need high precision, like CAB (for which nDCG@10 is an adequate metric). However, with high recall, the effort for the new judgments would probably be too high. A third general consideration concerns the relationship between relevance and usefulness in the context-aware retrieval field. In this work the information need gathered from the user’s context is adopted as reference point to judge the Web pages in the collection. In classical IR, we considered a page relevant if it provides an answer to the user’s information need. However, contextual information compounds the overall situation: taking into consideration the context, a page could be relevant for the user, but at the same time not useful. A page with too much information, and thus difficult to read on a mobile device, is not useful for the context, even if it is relevant to the user’s information need. In the same way we can consider a tourist in a town never visited before, without a means of transport, and looking for a tapas (variety of snacks in Spanish cuisine) bar in a rainy day. The system answers to the user query with the page of a tapas bar 10 minutes away, and with a map that guides the user to the destination. These results are highly relevant, but at the same time they could be useless: considering the overall situation in a rainy

day, a ten minutes walk could be unacceptable for a user. From this point of view, a page not useful has still to be considered relevant? The problem here is that in the context-aware retrieval field user’s information needs are not the only element to be considered for relevance: the contextual information represents constraints that can alter the relevance of a document. In particular, contextual values can modify both what the user needs and how the user needs it. The fourth kind of lesson we learned is related to the measures used to estimate the importance of terms related to context. For some experiment we exploited the tf.idf measure and at the beginning it seemed to be a good choice. However, we discovered that it is too much dependent on the context descriptor and, if the contexts were represented and described in a different way, we could have obtained very different tf.idf values. Thus tf.idf and all measures based on terms frequency in context descriptor could not be the proper solution. Moreover, other problems appear when we have to calculate tf.idf values, not in experimental condition, but at runtime, in a real situation. While tf values can be computed directly from the user context descriptor, which are the other contexts (documents) we exploit to compute the idf value? The history of previous contexts? Or other users’ contexts? Differently from documents, that are rigidly separated from each other, contexts cannot be easily identified as they are not “watertight compartments” but they continuously flow in each other.

CONCLUSION In this paper we have presented our approach for the evaluation of a CAR application for the mobile environment. We evaluated different strategies for automatic query building based on users’ current context. Despite user testing is the main evaluation technique adopted in this field, the early stage of development of our system and

13

Evaluating the Context Aware Browser

the need of measuring the effectiveness of different strategies guided us toward a TREC-like benchmark approach. In this way, we have made our second step to refine a methodology whose aim is to become a general early stage evaluation technique for retrieval processes in context-aware systems. On the basis of our experience, we believe that early stage benchmark evaluations followed by user studies, are an effective methodology to evaluate systems like CAB. The benchmark, in fact, does not substitute the user testing evaluation. Rather, several early stage benchmark experiments could provide more solid basis for the subsequent user testing that can thus be more focused. In the future we will work on two issues: to seek for more effective strategies that better compete with the manual one (e.g., removing the constraint of incremental query construction) and, more generally, to better understand reliability and usefulness of our incremental benchmark approach. It would be also interesting to work towards a publicly available or open source context test collection, containing a set of context descriptors (perhaps with a representation different from TREC topics, and maybe including a temporal longitudinal component as well, e.g., using Schank and Abelson’s scripts), Web resources, and relevance judgments. Finally we plan to execute the experiments in a fully mobile environment and to study how relevance judgments and results change switching from a desktop to a mobile environment.

ACKNOWLEDGMENT The authors acknowledge the financial support of the Italian Ministry of Education, University and Research (MIUR) within the FIRB project number RBIN04M8S8, and the region Friuli Venezia Giulia.

14

REFERENCES Brown, P. J., & Jones, G. J. F. (2001). Contextaware retrieval: Exploring a new environment for information retrieval and information filtering. Personal and Ubiquitous Computing, 5(4), 253–263. doi:10.1007/s007790170004 Bulfoni, A., Coppola, P., Della Mea, V., Di Gaspero, L., Mischis, D., Mizzaro, S., et al. (2008). AI on the move: Exploiting AI techniques for context inference on mobile devices. In Proc. of 5th Prestigious Applications of Intelligent Systems (PAIS 2008), co-located with ECAI08 (pp. 668–672). IOS Press. Carter, S., & Mankoff, J. (2004). Challenges for Ubicomp Evaluation. Technical Report UCBCSD-04-1331, Computer Science Division, University of California, Berkeley. Coppola, P., Lomuscio, R., Mizzaro, S., Nazzi, E., & Vassena, L. (2008). Mobile social software for cultural heritage: A reference model. In 2nd Workshop on Social Aspects of the Web (pp. 69–80). Cormack, G. V., Palmer, C. R., & Clarke, C. L. A. (1998). Efficient construction of large test collections. In SIGIR ’98: Proceedings of the 21st annual international ACM SIGIR conference on Research and development in information retrieval (pp. 282–289). ACM. Della Mea, V., & Mizzaro, S. (2004). Measuring retrieval effectiveness: A new proposal and a first experimental validation. Journal of the American Society for Information Science and Technology, 55(6), 530–543. doi:10.1002/asi.10408 Göker, A., & Myrhaug, H. (2008). Evaluation of a mobile information system in context. Information Processing & Management, 44(1), 39–65. doi:10.1016/j.ipm.2007.03.011

Evaluating the Context Aware Browser

Harper, D. J., & Hendry, D. G. (1997). Evaluation light. In M. Dunlop (ed.), Proc. of the second MIRA workshop (pp. 53–56). Technical Report TR-1997-2, Department of Computing Science, University of Glasgow, Glasgow. Jones, G. J. F., & Brown, P. J. (2004). Contextaware retrieval for ubiquitous computing environments. In: Mobile HCI Workshop on Mobile and Ubiquitous Information Access (pp. 227–243). Volume 2954. Springer LNCS. Mizzaro, S. Nazzi, & E., Vassena, L. (2008). Retrieval of context-aware applications on mobile devices: how to evaluate? In Proc. of Information Interaction in Context (IIiX ’08) (pp. 65–71). ACM. Scholtz, J. (2001). Evaluation methods for ubiquitous computing. Ubicomp Workshop. Scholtz, J. (2006). Metrics for evaluating human information interaction systems. Interacting with Computers, 18, 507–527. doi:10.1016/j. intcom.2005.10.004 Scholtz, J., & Consolvo, S. (2004). Towards a discipline for evaluating ubiquitous computing applications. Technical Report IRS-TR-04-004. Weiser, M. (1991). The computer for the 21st century. Scientific American, 265(3), 94–104. doi:10.1038/scientificamerican0991-94

KEY TERMS AND DEFINITIONS Mobile Devices: a mobile device (also known as cellphone device, handheld device, handheld computer) is a pocket-sized computing device. Information Retrieval: is the science of searching for documents, for information within documents. Context-Awareness: refers to the idea that computers can both sense, and react based on their environment, and based on the situation they are used in. Context Aware Browser: general-purpose solution to Web content fruition by means of context-aware mobile devices. Proactive: controlling a situation by causing something to happen rather than waiting to respond to it after it happens Evaluation: action of determining how a particular system behaves using criteria against a set of standards Benchmark: in IR a benchmark is composed by a collection of topics (expressions of user’s information needs), a collection of resources to retrieve, and a set of relevance judgments about those resources for each topic. It is a kind of evaluation that aims at measuring performances of systems in a controlled environment.

15

16

Chapter 2

Routing in Wireless Ad Hoc and Sensor Networks Milos Stojmenovic University of Ottawa, Canada

ABSTRACT Routing is the process of finding a path from a source node to a destination node. Since each node has a limited transmission range, the message is normally forwarded by other nodes in an ad hoc or sensor network. Therefore routes normally consist of several hops. Proposed routing schemes can be divided into topological and position based, depending on the availability of geographic location for nodes. Topological routing may be proactive or reactive. Position based routing consists of greedy approaches applied when a neighbor closer to the destination (than the node currently holding the packet) exists, and recovery schemes otherwise. In order to preserve bandwidth and power which are critical resources in ad hoc and sensor networks, localized approaches are proposed, where each node acts based solely on the location of itself, its neighbors, and the destination. There are various measures of optimality which lead to various schemes which optimize hop count, power, network lifetime, delay, or other metrics. A uniform solution based on ratio of cost and progress is described here

INTRODUCTION Routing is the process of finding a path from a source node to a destination node. Since each node DOI: 10.4018/978-1-60960-042-6.ch002

has a limited transmission range, the message is normally forwarded by other nodes in an ad hoc or sensor network. Therefore routes normally consist of several hops. Proposed routing schemes can be divided into topological and position based, depending on the availability of geographic

Copyright © 2011, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited.

Routing in Wireless Ad Hoc and Sensor Networks

location for nodes. Topological routing may be proactive, reactive or hybrid. Position based routing consists of greedy approaches applied when a neighbor closer to the destination (than the node currently holding the packet) exists, and recovery schemes otherwise. In order to preserve bandwidth and power which are critical resources in ad hoc and sensor networks, localized approaches are proposed, where each node acts based solely on the location of itself, its neighbors, and the destination. There are various measures of optimality which lead to various schemes which optimize hop count, power, network lifetime, delay, or other metrics. A uniform solution based on ratio of cost and progress is described here. An ad hoc network is a set of interconnected nodes that are deterministically or randomly dispersed in a given area and communicate over a wireless medium. Each of these nodes has the ability to send and receive messages to and from other nodes. Connections are possible over multiple nodes (multihop ad hoc network). Typical examples of ad hoc networks are conference and disaster relief networks. Sensor networks are similar to ad hoc networks in the sense that nodes have the ability to communicate over a wireless medium, but they differ in purpose and communication patterns. Each node in a sensor network is a small, battery powered device that has the capability of measuring or tracking the environment. Sensors can measure distance, speed, humidity, temperature, light, motion, seismic data, torque, and a host of other quantitatively measurable attributes of the environment that they are located in. They differ from ad hoc networks in purpose since each sensor has a specialized purpose of measuring and reporting the measured data. The two types of networks also differ in communication patterns. The nodes in ad hoc networks usually communicate between each other where the source and destination can be any two nodes. In sensor networks, nodes typically only communicate with a base station or sink. They usually only need to report data,

and are not able to perform a certain action based on the collected data, other than transmitting or receiving messages. Therefore, communication in a sensor network is much more structured than in an ad hoc network. Nodes in ad hoc or sensor networks may or may not know their geographic positions. The availability of positional information depends on the hardware of the nodes in the network. Typical devices used for positional information are GPS locators which are relatively small. Positional information is essential in sensor networks. Sensors need to know where they are located in order for whatever they are measuring to make sense to the base station. If a sensor detects fire, for instance, it must report the location of the fire, or else an appropriate response is not possible. However, sensors can be easily equipped with GPS locators, and the sensors themselves are relatively small devices. Positional information of sensors is not an easy problem (Bachrach & Taylor, 2005), but it is definitely an important part of a sensor’s makeup. It is assumed that nodes have equal transmission radii R, where two nodes are neighbors if they are located at most R units away from each other. That is, two nodes can directly communicate if and only if the distance between them is ≤R. This is referred to as the unit disk graph model. The wireless communication medium is different from the wired one. The wired medium normally provides only one-to-one communication, meaning that a message sent to a neighbor is only received by that neighbor. However, the wireless medium enables one-to-all communication, where a message sent by one node is received by all of its neighbors. This provides both advantages and limitations. A single message reaches more nodes; however, at the same time, bandwidth is limited because normally all nodes use the same frequency for communication. Sensors also have a sensing radius, which is normally less than the communication radius.

17

Routing in Wireless Ad Hoc and Sensor Networks

Presented protocols are applicable to both ad hoc and sensor networks, with different efficiencies, depending on the network assumptions. The availability of position information and mobility of the destination determines the choice of the corresponding class of protocols. Relative mobility impacts the suitability of protocols. For static nodes, proactive approaches are efficient, while for mobile nodes, reactive ones work better. Changes from active to sleep status at individual nodes also make certain protocols more efficient compared to others.

Routing without Position Information Routing without positional information is more difficult than routing with positional information, and requires more overhead in wireless networks. There are two general methodologies that could be used to route messages if position information is not available: proactive routing, and reactive routing. Proactive routing means that routes must be established before messages are sent through the network. This involves storing routing tables at each node in the network that act as roadmaps for routing. Routing tables are very cumbersome from a memory point of view and costly in the case of mobile nodes. Proactive routing with highly mobile nodes is very impractical since frequent routing table updates are necessary for the scheme to work. It is practical in situations where nodes are mostly static, such as in the case of a conference. Reactive routing is a more practical solution to networks with mobile nodes. Using this routing framework, nodes that need to send a message first need to find the destination node by flooding the network with a route discovery message. The destination node would then reply to the source using the memorized hops through the network. This type of routing is used in ad hoc networks where the network composition is variable. It could also be used in certain scenarios in sensor networks where nodes could turn themselves off

18

to conserve energy. New routes would then have to be found.

Routing with Position Information Positional information is helpful for routing in ad hoc networks because it is possible to find a route between two nodes without flooding and without routing tables. In both the cases of sensor networks and ad hoc networks, knowing the positions of nodes facilitates localized routing. Localized routing is a phrase that means routing just by knowing local information. A node that needs to send a message to another node knows its own position, the positions of all of its neighbors, and the position of the destination. Assuming this information is available, simple greedy localized routing protocols exist, which are competitive to the shortest path algorithm for dense networks. There also exist intelligent localized routing algorithms (Bose & Al, 2001), which guaranty delivery of the message to the destination in connected unit disk graphs, without using any memorization at nodes. Localized routing has considerably lower communication overhead compared to approaches using global network information, in networks with mobile nodes, or nodes that change their status between active and sleeping. The main advantage is in the cost of updating the needed network information. Updating information for localized routing is simple: each node periodically transmits a ‘hello’ message, received by all its neighbors. Other approaches require propagating topological changes throughout the network or the exchange of large messages containing routing tables. There must exist node location service mechanisms in ad hoc networks since destination nodes may move, or simply turn themselves off, whereas the location of a sink or base station is normally fixed and known to sensors in sensor networks. This may affect the route of a potential message, and more severely impact the power and bandwidth consumption of nodes due to excess

Routing in Wireless Ad Hoc and Sensor Networks

communications. The topic of location updates in mobile ad hoc networks is covered in greater detail in (Stojmenovic, 2006). Estimating hop counts between nodes is discussed in (Contla & Stojmenovic, 2003).

TOPOLOGICAL ROUTING Topological routing refers to routing where positional information is not used. We now describe two basic types of topological routing: proactive and reactive.

Proactive Routing A typical example of a proactive routing algorithm is the Bellman-Ford scheme (Jubin & Tornow, 1987). In this framework, each node exchanges its routing table with all of its neighbors and updates its own routing table if better routes to some destinations are available based on the new data. The entries in the routing table for one node are the costs of sending messages from that node to all other nodes using the best know routes. These entries also contain the next forwarding node to send the message to for all known nodes. Routing table exchanges of information are done periodically, so new routes become known gradually, as several iterations of the table exchange step are performed. This protocol is a popular one on the Internet, since most of the nodes are not mobile. In a highly mobile environment, table exchanges would have to be more frequent, which results in high overhead, and an impractical solution. To illustrate the Bellman-Ford technique, we imagine that a source node S has 3 neighbors, and intends to send a message to destination D. We also assume that the table exchanging step is about to be performed. When node S exchanges its routing table with its neighbors, it analyzes which neighbor to send the message to. In the received routing tables, information is listed about the costs of sending a message from that node to node D.

Node S then forwards the message to the node F that minimizes the cost of link S to F plus the cost of sending from F to the destination. Another way to obtain the shortest possible routes between any two pairs of nodes is to run Dijkstra’s shortest path algorithm for every pair of nodes in the network (McQuillan & Al, 1980). However, all link and node (topological) changes need to be broadcasted in the network. All nodes then maintain global network information, and are able to make optimal decisions when routing is in progress. Some optimizations are proposed in (Jacquet & Al, 2001), in the form of OLSR (optimized link state routing) protocol. Topological changes are normally broadcasted to all nodes via flooding, where each receiving node will retransmit the message exactly once. OLSR (Jacquet & Al, 2001) limits the retransmissions to certain nodes, which together create a backbone (any node not in the backbone is a direct neighbor to at least one node from the backbone). The specific protocol used, which reduces message flooding, is called Multi Point Relay MPR. Each transmitting node includes the list of neighbors that are asked to retransmit the message in the packet. The selection is made so that all 2-hop neighbors (neighbors of neighbors) are ‘covered’, that is, receive the message only if the selected neighbors retransmit. When a message is received at a node, it simply checks whether it is ‘asked’ to retransmit it or not, and behaves accordingly (Jacquet & Al, 2001). This reduces the flooding driven effects of wasted retransmissions of the original message.

Reactive Routing This type of routing is analogous to tourists trying to find an obscure address in a new city: they get closer and closer to the destination by asking people along the way. There are two variants of reactive routing strategies that we will examine here. They work very similarly, and can almost be considered as extensions of one another. One

19

Routing in Wireless Ad Hoc and Sensor Networks

Figure 1. Route discovery

Figure 2. Destination reports back

variant of route discovery is called the Advanced On-demand Distance Vector AODV (Perkins & Royer, 2002), and the other is Dynamic Source Routing DSR (Johnson & Maltz, 1996). The main idea here is that a source node first has to find the destination node before it can start sending messages. Therefore, it floods a route discovery message throughout the entire network. Each node retransmits the packet exactly once, upon the first receipt of it. We turn our attention to Figure 1 to illustrate this procedure. Node S starts flooding the network by passing the route discovery message to all of its neighboring nodes. The flooding is continued as outlined by the green arrows in the same figure. Note that messages are not retransmitted more than once since subsequent copies of the same message are ignored. Flooding is continued as depicted by the dark red and dark blue arrows until the destination is reached. For clarity, retransmissions by nodes that do not reach any node for the first time are not drawn. The route that is uncovered is transmitted back to the source in the reverse order of its discovery, as seen in Figure 2. In the AODV scheme, each node in the path would memorize the next hop back to the source. Thus A records S, B records A, C records B and D records C, which enables D to report to S via C, B and A (see Figure 2). In the DSR variant, the route discovery packet carries the whole paths from the source; therefore, destination D receives the memorized path SABCD. The source would be notified of the entire path to the destination.

After discovering and memorizing the route, S uses it to send actual traffic to D. Messages could now be sent using either hop by hop or total path variants. The described reactive routing algorithm is a basis for several existing protocols applied in specific scenarios. For example, ‘directed diffusion’ (Intanagonwiwat & Al, 2000) is a highly cited routing protocol for sensor networks, applying a ‘data centric’ approach. The data sink identifies a set of attributes and propagates an interest message throughout the network. The interest is flooded blindly. Each receiving node records the interest and establishes a so called gradient, which is simply a link toward a neighbor on a route back to the sink. The protocol is thus algorithmically equivalent to route discovery where the search for an IP address has been replaced with search for data of interest located in some sensors, which have no identities. AODV (Perkins & Royer, 2002) and DSR (Johnson & Maltz, 1996) use the hop count metric (the number of transmissions) as the measure of optimality of a route. The route discovery message might contain several other cost measures which may be taken into consideration while choosing the best path. Depending on the application, cost measures might include delay, congestion or power consumption. Several paths might be available, but the best one is chosen at the destination. Applications in which delay is paramount, such as QoS, may require the usage of several paths to the destination to increase bandwidth. The opti-

20

Routing in Wireless Ad Hoc and Sensor Networks

mal route from source to destination, measured by a particular metric, may not be the first one that reaches the destination. The route discovery process is therefore modified. When a route discovery packet originating from S is received by node A for the first time, the total cost (on a partial route from S to A) is recorded. Each new route request packet for the same source-destination pair arrives with its new total cost. If that cost is lower than the previously recorded cost, the route request is retransmitted, with that new total cost being appended to the message. Otherwise this message is ignored. A technique for reducing the cost of flooding is an expanding ring search. In this procedure, the source node would send out a route discovery message to its neighboring nodes with an attached time to live TTL variable. This TTL variable is used in determining how many links the route discovery message is permitted to cross from the source node. For instance, the TTL can be set at 2 hops initially to limit the search radius of the initial route discovery message. The idea is that the destination might be near or far away from the source. Instead of flooding the network right away, we try to find the destination among the nodes that are close to the source. If the destination node is not found, then the TTL variable is increased (e.g. doubled), and the radius of the search expands. These expanding ring search phases are repeated until the destination is found. Figure 3 shows a sample network with source and destination nodes labeled as S and D respectively. The purple ellipse marks the area that the route discovery message with TTL 2 would search. The destination is not found in this area, and the TTL is doubled. The aqua green area is searched with a TTL of 4. The destination would be discovered in the next search phase where the TTL is 8.

Hybrid Routing Proactive and reactive routing approaches can be combined into a single approach. Proactive rout-

Figure 3. Expanded ring search strategy

ing is applied if the destination is nearby, while reactive routing is applied when it is far away. For instance, in the ZRP (Zone Routing Protocol) (Beijar, 2002), a zone for a source includes all nodes up to k hops away. Routing information inside its zone is constantly updated. Proactively, whenever a link state is changed, a notice is sent k hops away by flooding. Therefore, the source always knows how to reach a destination that is at most k hops away. Routing to a destination that is not nearby (more than k hops away) is done by the reactive approach. A route discovery message is flooded in the network until a zone containing the destination is found. The border node of that zone then can report back to the source.

POSITION BASED ROUTING Advance Mode Routing In this section, we consider the situation where nodes are equipped with positional information. The positions of each node’s neighbors may be obtained by occasional exchanges of ‘hello’ messages.

21

Routing in Wireless Ad Hoc and Sensor Networks

Cost to Progress Routing Framework We will now describe a general routing framework (Stojmenovic-2, 2006) which is applicable for a variety of cost metrics. It is understood here that each message sent between two nodes has an inherent cost associated with it, and a progress value. The cost of transmitting a message between two nodes can be defined in several ways. Typically, it falls under hop count, power consumption and delay. Progress measures the advancement of the message toward the destination. One way to measure progress is by considering the distance of the message to the destination. In Figure 4, we see a network with labeled nodes, and cost values at each edge. The red sequence of arrows is the shortest path from source S to destination D. We require global information about the network (locations of all nodes) in order to be able to find the shortest path. However, we are confined to local information. Therefore, node S has the option of forwarding the message to one of nodes F, I and K. It is proposed here that the cost of the next link should be compared to the progress made by forwarding it to that link. Therefore, values: 18/ (|SD|-|FD|), 15/(|SD|-|ID|), and 7/(|SD|-|KD|) are compared and the minimal value determines the next forwarding node in the chain. If such logic is followed, the purple sequence of arrows marks the resulting path the message takes from source to destination. Cost to progress ratio based routing was recently enhanced. Instead of sending directly to neighbor A with the optimal cost to progress ratio, source S constructs the shortest weighted path (with the considered cost as the weight) toward A, and the message is sent along that path until a node B that is closer to D than S is found. Note that intermediate nodes may recompute the path, so that the message remains of constant length. Node B then resumes the same protocol.

22

Figure 4. Shortest weighted path and cost to progress ratio protocols

Greedy Routing and MFR Greedy routing in wireless networks was first proposed by Finn (Finn, 1987). In his work, it is understood that the cost metric of sending one message to a neighboring node is one hop. This is a special case of the cost to progress routing framework since it assumes that message sending costs between neighboring nodes are equal regardless of the distance between them. In order to illustrate this point, we observe the network in Figure 4 again. Since the cost of traversing an edge is now one, we can disregard the weights of the edges. Source S still needs to pick one of its neighbors (F, I, or K) to forward the message to. The greedy algorithm now only considers the progress made by forwarding the message to each of these nodes since transmission costs are now constant. By progress, we mean minimizing the distance from the next node to the destination. The algorithm therefore takes the minimum distance of FD, ID, and KD. It chooses I as the next node to forward the message to. The purple arrows again mark the path that the greedy algorithm would choose. A similar progress measure based on distance was proposed by (Takagi & Kleinrock, 1984) called Most Forward within Radius (MFR). Instead of choosing the next forwarding node as the one

Routing in Wireless Ad Hoc and Sensor Networks

which is minimally distanced from the destination, we take the one whose projection onto the line CD is closest to destination D, as seen in Figure 5. In this figure, node C is deciding where to forward the message intended for destination D. The projections of potential forwarding nodes F, A and G are seen as F’, A’ and G’ respectively. A’ is closest to the destination, and therefore, the message is forwarded to node A. This is loop free algorithm, although by no means a solution that guarantees delivery. Even though loops are not possible, reaching a dead end is, and recovery protocols will be discussed in section 3.2.

Power Aware Routing This form of localized routing was first studied in (Stojmenovic & Lin, 2001). It takes into consideration the power needed to transmit a message to the next node. The power needed to send a packet from C to A is proportional to rα+c, where α is a power attenuation factor (2≤α≤6), r=|CA| is the distance from node C to node A, while c is a constant greater than 0. Constant c accounts for the fixed energy costs of sending and receiving a message and running such an electronic circuit. Power aware routing is another example of localized routing. The algorithm will forward the message to the node A that minimizes the ratio of transmission energy to message progress (rα+c)/ (|CD|-|AD|) (Kuruvila & Al, 2006). One of the disadvantages of power aware routing (as with all of the other greedy algorithms we have seen so far) is that certain routes become popular, and soon the nodes along these routes become drained. It is therefore beneficial to consider a different routing strategy that takes into account the remaining energy of a possible forwarding node before nominating it to continue the routing task. In such cases, nodes have to announce their remaining energy status via periodic ‘hello’ messages. The details of localized power and remaining energy aware routing protocols

Figure 5. Selecting best neighbor in MFR routing protocol

are in (Stojmenovic & Lin, 2001), (Kuruvila & Al, 2006).

QoS Routing Qos routing is highly applicable to multimedia traffic such as on-demand video. In such cases, both time and bandwidth are critical aspects of routing. Routing protocols of this nature will become more popular as handheld portables become more advanced with the ability to access TV programming and other high bandwidth consumption applications. In this type of routing (He & Al, 2005), (Huang, Dai & Wu, 2004), (Stojmenovic-2, 2006), the factor we are interested in limiting is the time it takes to send a message from a source to a destination. We therefore have to consider a broader spectrum of factors that influence message passing. Specifically, we mean the delays experienced if a message is sent to a popular node. By delays, we mean queuing and the time it takes to forward the message. The criteria for selecting the next forwarding node A from current node C is based on minimizing the delay/(|CD|-|AD|).

Beaconless Routing So far, we have seen routing approaches where nodes have local knowledge about their neighbors. Local knowledge is made available by periodically

23

Routing in Wireless Ad Hoc and Sensor Networks

Figure 6. Gabriel graph

sending ‘hello’ packets. However, this causes significant overhead, especially if the amount of actual traffic is not high. Furthermore, since nodes may move, or turn themselves off at any time, the information (position or even availability) assumed about neighbors may not be accurate. This leads to suboptimal routing procedures or even unnecessary failures. In such cases, Zorzi (Zorzi, 2004) proposes to have a dynamic neighbor discovery phase where RTS/CTS packets are sent before sending actual routing packets. A node which wishes to send a message first sends out an RTS (ready to send) packet. All receiving neighbors calculate a timeout waiting period, which corresponds to their distance from the destination. Neighbors closer to the destination select shorter timeouts. The neighbor closest to the destination is therefore the first one to respond with a CTS (clear to send) packet. Upon receiving the first response from a neighbor A, current node S transmits the actual message to A. Other neighbors overhear that message which cancels their own possible responses at the end of their timeouts.

Recovery Mode Routing It is possible for a routing algorithm to guarantee delivery of a message while only dealing with local information, and not applying any memorization at nodes (all necessary information arrives with the packet). This face algorithm (Bose & Al, 2001) only assumes that there exists a path from a source

24

Figure 7. Marked source and destination

to a destination, and a unit disk graph. The face algorithm is a recovery scheme and is normally applied when the greedy algorithm fails. The details of face and the combined GFG (greedyface-greedy) algorithms are seen in this section.

Face Routing Face routing (Bose & Al, 2001) is an algorithm that guarantees delivery of a message, but does not always have the best performance when it comes to efficiency. This type of routing algorithm has two phases: the construction of a Gabriel graph from the network, and the actual routing phase. Gabriel graphs are planar, which means that no two edges intersect. Planar graphs are a composition of faces, along the sides of which messages can be passed. Figure 6 shows an example of a Gabriel graph of a network. There exists an edge between two nodes (a, b) in a Gabriel graph if and only if there are no nodes in the disk with diameter ab. In Figure 6, we see that there is an edge in the green circle, since there are no nodes in that area. Conversely, we see no edge in the purple circle since there are nodes in that area. In Figure 7, we have the Gabriel graph of the network in Figure 6, with marked source and destination nodes, S and D. There is also a purple imaginary dashed line from the source to the destination in Figure 7. In the face routing algorithm, this line is used to determine the points at which a message would change faces on its way to the destination. To illustrate this point, we

Routing in Wireless Ad Hoc and Sensor Networks

Figure 8. Face routing

Figure 9. Greedy mode in GFG routing

Figure 10. Face mode in GFG routing

Figure 11. Result of GFG routing

focus on Figure 8. Here, we see that the message follows the contours of each face until it hits the purple dashed line. At such points, it switches faces, and follows the contours of the next face. This procedure is repeated until the message reaches the destination. We can see from this illustration that face routing is not the most efficient algorithm, but nevertheless a useful one in cases where other greedy algorithms fail. We see such a case in the next section.

purple arrow. This node cannot forward the message to any node that is closer to the destination than itself. At this point, the greedy algorithm has failed, so face mode takes over since we know that face mode guarantees delivery of the message as long as there exists a path to the destination. The dashed line is drawn from that node to the destination, and face mode begins. It traverses the edges of the outer face until it finds a node that is closer to the destination than the node at which face mode began. When it finds such a node, as seen is Figure 10, we switch back to greedy mode and continue the routing process. Figure 11 shows the result of the GFG routing protocol, where the message was successfully delivered by the greedy phase in the last step. It was shown recently in (Frey & Stojmenovic, 2006) that face routing always recovers along the first traversed face if the Gabriel graph is used for planarization. However, in some cases, such as when a realistic physical layer is considered, other methods need to be applied to planarize the

GFG Routing Greedy Face Greedy (GFG) routing is a combination of greedy routing and face routing. The idea behind this algorithm is to proceed with the greedy algorithm until it becomes stuck, then to use the face mode to recover from such instances, and to again proceed with greedy mode. Figure 9 illustrates the GFG algorithm on a sample network. In greedy mode, node S forwards the message to the next node which is singled out by the

25

Routing in Wireless Ad Hoc and Sensor Networks

graph, and face routing may have to traverse several faces. To guarantee delivery, GFG only needs the graph to be somehow planarized, independently of the method applied and the particular planar graph outcome.

Beaconless GFG In beaconless GFG, the nodes do not have knowledge of their neighbors. Therefore, at each step in forwarding a message, some local knowledge must be obtained. The node S currently holding the message would send out a single RTS packet as a single hop broadcast in which a bit would be added to signify whether the node wishes to send its message in greedy or face mode. In greedy mode, the neighboring node which is closest to the destination responds with a CTS packet after a time delay. The duration of the time delay is inversely proportional to the distance of that node to the destination. Therefore, the node that is closest to the destination transmits its CTS packets after a shorter time delay. This transmission cancels other CTS packets that are to be sent by other nodes. In face mode, an RTS packet is sent out, and neighbors that are closer to the node sending the RTS respond first with CTS packets. If a node A hears another closer neighbor B responding which is inside the circle with diameter AS, it indicates that AS is not in the Gabriel graph, and A then does not respond. Thus only the neighbors in Gabriel graph will respond. The Gabriel graph is then constructed from the received responses, and face routing continues. The details of this algorithm are found in (Chawla & Al, 2006).

ROUTING WITH A REALISTIC PHYSICAL LAYER The routing assumptions we have made so far state that two nodes are neighbors if they are at most distance R away from each other. If that condition is satisfied, then it was assumed until now that

26

a message passed from two neighboring nodes had a 100% chance of arriving. Realistically, a message is received if the signal strength at the receiver is greater than a certain threshold, but signal strength has random variations. It is still true under realistic physical layer assumptions that neighbors can pass messages, but it is not always true that each message is received the first time it is sent. This is now a probabilistic function of distance. Figure 12 shows typical reception probabilities based on distance.

Route Discovery Process The route discovery process considering a realistic physical layer is not the same as when considering just the regular unit disk model. In a basic disk model of routing, one discovery message is enough for neighbors to correctly receive it and promptly respond. Therefore, neighbor and route discovery is relatively simple. In a realistic physical layer setting, messages requesting routes may not be received by other nodes, and might have to be rebroadcast. If only one discovery message is sent out, it might not be received by a node that is a relatively good connection to the intended destination. In this case, sub-optimal paths might be taken to the destination. On the contrary, it might also happen that a very weak connection to a distant node might be discovered by just one route discovery message. Under such circumstances, when actual messages try to use this link, they will fail most of the time. It is hence much better to send out multiple instances of route finding messages, and based on the available responses, construct a path that seems most reliable (Stojmenovic & Al, 2005). Route discovery packets are generally shorter than full message packets, and are received with a higher probability. Thus, more research is needed to design a route discovery process with low overhead, but which produces good routes in practice.

Routing in Wireless Ad Hoc and Sensor Networks

Figure 12. Reception probability vs. distance

Position Based Routing with Realistic Physical Layer In any routing algorithm with a realistic physical layer, whether it be greedy or whether it has global information, the cost of passing a message from one node to the next needs to reflect the capability of efficient transmission. What we mean by this is that the cost of passing a message directly between two distant nodes is not the same as passing a message between two nodes that are relatively close to one another. Under realistic conditions, nodes have to assume that their message did not reach the intended receiver unless that receiver acknowledges receiving the message. In circumstances where neither the message reception nor the acknowledgement reception is certain, messages need to be retransmitted u times until an acknowledgement is received. Here, u is approximately 1/p(x), where p(x) is the packet reception probability for two nodes at distance

x. The cost of transmitting a message between two nodes can now be measured in expected hops, where a number of hops might be necessary to send a message to a neighboring node. This new cost function is called the expected hop count (EHC). Essentially, all of the connections, or edges, in a network become associated with weights. Greedy and other routing schemes now must consider the quality of links, and account for them by minimizing EHC(CA)/(|CD|-|AD|), where C and A are neighboring nodes (Kuruvila & Al, 2006). A localized routing algorithm with a realistic physical layer that guarantees delivery is described in (Nayak & Stojmenovic, 2006). It is based on the GFG algorithm where greedy mode is as described above, while face mode uses the same path as in the face routing scheme (Bose & Al, 2001)

27

Routing in Wireless Ad Hoc and Sensor Networks

CONCLUSION We have seen here a number of approaches of routing in ad hoc and sensor networks. Certain routing methodologies are better suited for sensor networks, while others are better suited for ad hoc networks. There are a number of factors that influence the way routing algorithms should be designed and yet still even more that influence the way that the algorithms actually work in both scenarios. We have to take into consideration that most of this work is highly theoretical, and that a fully functioning sensor network that solves a particular problem is yet to be assembled. Routing algorithms that consider local information yet guarantee delivery of messages seem most promising for large scale networks. Power aware routing seems like a good cost function for wireless networks where power consumption is the main issue. QoS routing is a logical approach for high bandwidth demand applications such as on demand video. Each routing protocol that we have seen here has its application in certain scenarios. Further development of this field, and incorporation and testing of these algorithms may result in even more efficient and productive routing strategies.

REFERENCES Bachrach, J., & Taylor, C. (2005). Localization in Sensor Networks. Handbook of Sensor Networks. New York: Wiley-Interscience. Beijar, N. (2002). Zone Routing Protocol (ZRP), Licentiate course on Telecommunications Technology. Ad Hoc Networking. Bose, P., Morin, P., Stojmenovic, I., & Urrutia, J. (2001). Routing with guaranteed delivery in ad hoc wireless networks. ACM Wireless Networks, 7(6), 609–616. doi:10.1023/A:1012319418150

28

Chawla, M., Goel, N., Kalaichelvan, K., Nayak, A., & Stojmenovic, I. (2006). Beaconless positionbased routing with guaranteed delivery for wireless ad hoc and sensor networks. International Journal of Sensor Networks, 1(1), 61–70. Contla, P., & Stojmenovic, M. (2003) Estimating Hop Counts in Position Based Routing Schemes for Ad Hoc Networks. Telecommunications Systems (Kluwer/Springer), 22(1-4), 109-118. Finn, G. G. (1987). Routing and addressing problems in large metropolitan-scale internetworks. ISI Research Report ISU/RR-87-180. Frey, H., & Stojmenovic, I. (2005). Geographic and energy aware routing in sensor networks. Handbook of Sensor Networks: Algorithms and Architectures. New York: Wiley. Frey, H., & Stojmenovic, I. (2006). On Delivery Guarantees of Face and Combined Greedy-Face Routing Algorithms in Ad Hoc and Sensor Networks. In Proceedings of the The Twelfth ACM Annual International Conference on Mobile Computing and Networking MOBICOM, Los Angeles. He, T., Stankovic, J., Lu, C., & Abdelzaher, T. (2005). A Spatiotemporal Communication Protocol for Wireless Sensor Networks, IEEE ICDCS, May 2003. IEEE Transactions on Parallel and Distributed Systems, 16(10), 995–1006. doi:10.1109/TPDS.2005.116 Huang, C. Dai, F., & Wu, J. (2004). On-demand location-aided QoS routing in ad hoc networks. In Proc. Int. Conf. Parallel Processing ICPP, Montreal, 502–509. Intanagonwiwat, C., Govindan, R., & Estrin, D. (2000). Directed diffusion: a scalable and robust communication paradigm for sensor networks. In Proceedings of the ACM/IEEE International Conference on Mobile Computing and Networking, Boston, MA, pp. 56-67.

Routing in Wireless Ad Hoc and Sensor Networks

Jacquet, P., & Muhlethaler, P. Clausen, T., Laouiti, A., Qayyum, A., & Viennot, L. (2001). Optimized link state routing protocols for ad hoc networks. In Proc. IEEE Int. Multi-Topic Conf. INMIC, Lahore, Pakistan, pp. 62-68. Johnson, D. B., & Maltz, D. A. (1996). Dynamic source routing in ad hoc wireless networks. In Imielinski, T., & Korth, H. F. (Eds.), Mobile Computing (pp. 153–181). Kluwer. doi:10.1007/9780-585-29603-6_5 Jubin, J., & Tornow, J. D. (1987). The DARPA packet radio network protocols. Proceedings of the IEEE, 75(1), 21–32. doi:10.1109/ PROC.1987.13702 Kuruvila, J., Nayak, A., & Stojmenovic, I. (2006). Progress and location based localized power aware routing for ad hoc and sensor wireless networks. International Journal of Distributed Sensor Networks, 2(2), 147–159. doi:10.1080/15501320500259159 Mauve, M., Widmer, J., & Hartenstein, H. (2001). A Survey on Position-Based Routing in Mobile Ad Hoc Networks. IEEE Network, 15(6), 30–39. doi:10.1109/65.967595 McQuillan, J. M., Richer, I., & Rosen, E. C. (1980). The new routing algorithm for ARPANET. IEEE Transactions on Communications, 28(5), 711–719. doi:10.1109/TCOM.1980.1094721 Nayak, A., & Stojmenovic, M. (2006). Localized routing with guaranteed delivery and a realistic physical layer in wireless sensor networks. Computer Communications (Elsevier), 29(13-14), 2550–2555. doi:10.1016/j.comcom.2006.02.013 Perkins, C., & Royer, E. (2002). Ad hoc on demand distance vector (AODV) routing, Proceedings of the 2nd IEEE Workshop on Mobile Computing Systems and Applications, pp. 90-100.

Royer, E. M., & Toh, C. K. (1999). A Review of Current Routing Protocols for Ad Hoc Mobile Wireless Networks. IEEE Personal Communications, 6(2), 46–55. doi:10.1109/98.760423 Ruiz, P., & Stojmenovic, I. (2007). Cost-efficient multicast routing in ad hoc and sensor networks. In Handbook on Approximation Algorithms and Metaheuristics, Chapman & Hall/CRC (Teofilo Gonzalez, ed.), Vol. 30, No. 18, pp. 3746-3756. Stojmenovic, I. (2002). Location updates for efficient routing in wireless networks. In Handbook of Wireless Networks and Mobile Computing. New York: John Wiley & Sons. doi:10.1002/0471224561.ch21 Stojmenovic, I. (2004). Geocasting with guaranteed delivery in sensor networks. IEEE Wireless Communications, 11(6), 29–37. doi:10.1109/ MWC.2004.1368894 Stojmenovic, I. (2006). Localized network layer protocols in sensor networks based on optimizing cost over progress ratio and avoiding parameters. IEEE Network, 20(1), 21–27. doi:10.1109/ MNET.2006.1580915 Stojmenovic, I., & Lin, X. (2001). Power aware localized routing in ad hoc networks. IEEE Transactions on Parallel and Distributed Systems, 12(10), 1023–1032. doi:10.1109/71.963415 Stojmenovic, I., Nayak, A., & Kuruvila, J. (2005). Design guidelines for routing protocols in ad hoc and sensor networks with a realistic physical layer. [Ad Hoc and Sensor Networks Series]. IEEE Communications Magazine, 43(3), 101–106. doi:10.1109/MCOM.2005.1404603 Stojmenovic, M. (2005). Swarm intelligence for routing in ad hoc wireless networks, Security and Routing in Wireless Networks (Y. Xiao, J. Li, and Yi Pan Yi, eds), Nova Science Publishers, pp. 163-184.

29

Routing in Wireless Ad Hoc and Sensor Networks

Takagi, H., & Kleinrock, L. (1984). Optimal transmission ranges for randomly distributed packet radio terminals. IEEE Transactions on Communications, 32(3), 246–257. doi:10.1109/ TCOM.1984.1096061 Tseng, Y. C. Liao, W.H., & Wu, S.L. (2002). Mobile ad hoc networks and routing protocols. Handbook of Wireless Networks and Mobile Computing. New York: John Wiley & Sons. Wu, J. (2002). Dominating set based routing in ad hoc wireless networks. Handbook of Wireless Networks and Mobile Computing (pp. 425–450). New York: John Wiley & Sons. Zorzi, M. (2004). A new contention-based MAC protocol for geographic forwarding in ad hoc and sensor networks. IEEE Int. Conf. on Communications (ICC), Paris, Vol. 6, pp. 3481–3485.

KEY TERMS AND DEFINITIONS Ad Hoc Network: is a self-configuring network of mobile devices connected by wireless links. Routing: Sending a packet from a source node to a destination node.

30

Proactive Routing: In this framework, each node exchanges its routing table with all of its neighbors and updates its own routing table if better routes to some destinations are available based on the new data. The entries in the routing table for one node are the costs of sending messages from that node to all other nodes using the best know routes. Reactive Routing: This type of routing is analogous to tourists trying to find an obscure address in a new city: they get closer and closer to the destination by asking people along the way. Hybrid Routing: Proactive and reactive routing approaches can be combined into a single approach. Proactive routing is applied if the destination is nearby, while reactive routing is applied when it is far away. Topological Routing: Routing where nodes do not have positional information. Localized Routing: Routing where nodes have positional information about themselves, their neighbors, and the destination, and no information about the rest of the network. Backbone: Subset of nodes so that each node not in the subset, has a neighbor from the subset

31

Chapter 3

Mobile Ad Hoc Networks:

Protocol Design and Implementation Crescenzio Gallo University of Foggia, Italy Michele Perilli University of Foggia, Italy Michelangelo De Bonis University of Foggia, Italy

ABSTRACT Mobile communication networks have become an integral part of our society, significantly enhancing communication capabilities. Mobile ad hoc networks (MANETs) extend this capability to any time/anywhere communication, providing connectivity without the need of an underlying infrastructure. The new coming realm of mobile ad hoc networks is first investigated, focusing on research problems related to the design and development of routing protocols, both from a formal and technical point of view. Then link stability in a high mobility environment is examined, and a route discovery mechanism is analyzed, together with a practical implementation of a routing protocol in ad hoc multi-rate environments which privileges link stability instead of traditional speed and minimum distance approaches.

INTRODUCTION Mobile ad hoc networks consist of interconnected mobile hosts with routing capabilities. Considerable work has been done in the development of routing protocols for ad hoc networks, starting from Internet protocols developed in the seventies. In recent years, the interest in ad hoc networks has grown due to the availability of wireless communication devices.

New research directions in theoretical computer science and in particular in protocol design make use of game-theory concepts and tools. From this “perspective”, protocols are viewed as games with players represented by network nodes which “play” (participate in) the game; each node (agent) has its own utility function, such as network flow (to be maximized) or energy consumption (to be minimized.) This approach thus is a natural pointof-view of a distributed computing architecture,

DOI: 10.4018/978-1-60960-042-6.ch003 Copyright © 2011, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited.

Mobile Ad Hoc Networks

the most interesting paradigm in actual computer science. This is especially true since the advent of wireless networks based on IEEE 802.11 Protocol (IEEE, 1999, 2003) (and in particular with the definition of the new draft “n”) where it is possible to deal with a variable-speed link going from 1 to about 300 Mbps (Fan, 2004). Besides, considering that mobile networks (see Mobile landscape, 2009) have the peculiarity of movement (which makes the link speed highly variable and therefore very unstable), stability of routes becomes a difficult undertaking. Furthermore, as for its intrinsic nature, the same protocol IEEE 802.11 introduces a considerable network overhead to control the transmission at the expense of throughput. So we think that choosing a stable routing – mainly considering stable links – is preferable to taking into account only the link speed and/or length.

BACKGROUND Distributed Computing Environments In a distributed computing environment, such as a network, different pieces of software interact by following one or more well-defined protocols. As an example, the request to access a Web page (which is routinely issued by a simple mouse click) is served by a number of such interactions that involve, besides the browser and the remote Web server hosting the page, also a number of intermediate “agents” (called routers) that make the request/response message delivery possible. There is a plethora of protocols, which are primarily classified according to the ISO/OSI layered model of networking (see Figure 1). At the one end of the stack we have the physical layer protocols, whose goal is to make possible the delivery of raw sequences of bits between directly connected computers. At the

32

Figure 1. ISO/OSI networking layered model

other end there are the application protocols supporting well-known high-level services, such as the already mentioned Web service (through the http protocol.) The most fundamental protocols are probably those that fall under the general term of TCP/IP: these are the protocols upon which the Internet is based, since they implement the key functions of routing and congestion/flow control. One common assumption in the distributed computing and networking literature is that the agents participating in a protocol execution follow the guidelines specified by the protocol itself (say, a router that is expected to route a data packet closer to its destination point will always do so, unless it is temporarily out of service.) There are clearly a number of settings in which this assumption makes sense. An obvious case, for instance, is that of a private network owned by a single Corporation. In the field of mobile networks (the one which we are most interested in here) there are application settings in which the above assumption applies

Mobile Ad Hoc Networks

Figure 2. Mobile ad hoc networks

as well. Consider, for instance, the case of sensor networks deployed with the goal of monitoring a given region. Another case is that of military networks. However, the assumption does not seem reasonable when, as in case of the Internet, the network is operated and used by different economic subjects, with possibly conflicting interests. Stated in other words, the assumption is not reasonable when the agents participating in the protocol may act selfishly with the aim at improving some personal utility. The obvious example here is a set of routers in a given domain that decide not to forward Internet traffic headed to domains belonging to competing organizations, or a military scenario where moving devices are not interested in routing enemy’s traffic. In the presence of selfish agents, the design of network protocols becomes a much harder task. Actually, this is not the whole story, as the influence of socio-economic factors calls for a completely new approach to the understanding and the design of network protocols. Quoting an influential author in the theoretical computer science community “the mathematical tools and insights most appropriate for understanding the Internet” and hence (mobile) networks “may come from a fusion of algorithmic ideas with concept and techniques

from Mathematical Economics and Game Theory.” (Papadimitriou, 2001) This new approach can be briefly synthesized as follows. First of all, the protocol and all the agents that possibly participate in the protocol define a game; this implies that each agent (player) has its own utility function. According to this scenario, a first goal is the understanding of the social cost of selfishness. In other words, given a measure of the overall performance of the protocol (this may be some global function, such as the average end-toend delay incurred by any message traveling the network), you may want to understand how bad, with respect to the ideal optimum performance, the network behaves due to the selfishness of some participants (as in Roughgarden & Tardos, 2000). A complementary goal is the design of protocols that aim at achieving some social objectives through the properties of rationality and truthfulness (Nisan & Ronen, 1999).

Mobile Ad Hoc Networks Mobile ad hoc networks are wireless networks characterized by the absence of any centralized coordination entity (see Figure 2). Each (mobile) device of the network, say a laptop or a PDA, can

33

Mobile Ad Hoc Networks

Figure 3. (a)Transmission range (b) Communication graph

be only directly connected to the devices within its transmitting range, called neighbor devices; as a consequence, traffic destined to (or originated from) a non neighbor device has to be routed through one or more intermediate devices (multihop traffic.) This characteristic represents a major difference from other wireless networks (such as cellular phone networks or wireless LAN) and is the source of a number of challenging problems.

Topology Control Given a transmitting range for each device (see Figure 3a), an ad hoc network can be modeled as a (directed) graph whose nodes are the devices and there is an edge from node u to node v if and only if v is within u’s transmitting range. Such graph is often termed as the communication graph (see Figure 3b). Topology control is the problem of establishing some desired connectivity property for the communication graph acting on the devices’ transmitting range. Depending on the application scenario and the traffic model, the required connectivity properties may vary; the most obvious is full-connectivity (i.e., the graph must be strongly connected), but in some cases you may require stronger properties, such as bi- or tri-connectivity. Also, for reasons that are related with the protocols at the other levels, it is often required that the graph

34

be symmetric, i.e., there is an edge (u, v) if and only if there is the edge (v, u). Clearly, the goal of topology control is to achieve the desired connectivity properties using the minimum possible transmitting powers (hence ranges.) Because of the potential reduction in the nodes’ transmitting power, topology control may have positive effects on both network lifetime and channel capacity, since a reduced range implies less contention on the wireless medium. Unfortunately, determining the minimum energy consumption topology is a hard problem (Blough et al., 2002; Clementi et al., 1999). Practical (suboptimal) protocols for enforcing connectivity are presented in Blough et al. (2003), Liu & Li (2002), Ramanathan & Rosales-Hain (2000).

Routing Routing is the problem of deciding the path in the communication graph for traffic destined to non neighboring nodes. Logically, the routing level is above topology control, but is strictly related to that. Some authors do not consider a separate topology control level (in particular, no topology control is needed if all the nodes have the same, maximum, transmitting range) and the network topology is “determined” by the route discovery

Mobile Ad Hoc Networks

stage itself. In consideration of the power required to transmit a packet to distance d (which grows at least quadratically with d), different sourcedestination paths can be characterized by very different energy consumption, and an obvious goal of routing is to choose the most efficient among such paths.

Medium Access Control (MAC) The wireless channel is shared among network devices; clearly, while one such device u is transmitting, no other device within u’s transmitting range may initiate another transmission (this resembles a non cooperative game scenario). When different transmitting powers are used, e.g., as a consequence of decisions made at the topology control level, the request-to-send/clear-to-send (RTS/CTS) mechanism needs to be tuned in order to silence all the possibly conflicting nodes. For instance, suppose that u is not within v’s range while v is within u’s. Then a RTS control message sent by v cannot silence u, and the latter may erroneously assume that the channel is available for transmission (Jung & Vaidya, 2002). Note that the communication graph of the above example is non symmetric. One may erroneously think that an obvious solution to the problem is simply to drop the (u, v) arc when building the network topology. However, the example shows that logical connectivity is in general different from physical connectivity. Logical symmetry enforced at the topology control level can be useful in the routing discovery phase, but what counts at the MAC level is the physical connectivity. This is a point that has been often overlooked in the literature on topology control.

Energy Efficiency Energy efficiency is a concern at all levels of the protocol stack. In particular, topology control has the specific goal of determining a low consumption communication graph. As for routing, it is

generally acknowledged that energy efficiency considerations must complement more traditional analysis based on packet loss rate, routing message overhead, etc. (Johansson et al., 1999) The ultimate goal is to increase the lifetime of individual devices and that of the network as a whole. Several definitions of network lifetime have been proposed: according to the simplest definition a network is not more alive when the first node dies; another definition considers the size of the largest connected component, which must contain at least 90% of the nodes. Even other definitions are possible. As a result of the above mentioned problems, the ad hoc network technology is still in its infancy, and no large-scale implementation of ad hoc networks has been reported yet. However, in spite of the above mentioned difficulties, there is an already large body of works that deal with ad hoc networks. This is due to the fact that ad hoc networks represent the most promising paradigm for the foreseen ubiquitous computing. As an example scenario of ubiquitous computing, consider a (wired) Internet access service to mobile users that many providers have already installed in a number of “public” sites (for instance, several airports and shopping centers are already offering this service for free.) The possibility, offered by ad hoc technology, of multi-hop access, where a mobile device communicates with the base station(s) via other mobile devices, would have a positive impact on such aspects as increased coverage area of the service and, as a result of the adoption of shorter transmitting ranges, increased network capacity and lifetime.

Security Another important issue concerning networks is security. Wireless networks use radio waves (i.e. electromagnetic signals) leading to an irradiation phenomenon in the interested area: signals permitting network connections exceed

35

Mobile Ad Hoc Networks

Figure 4. Trusted relationships

needed area limits, reaching surrounding places in spite of a loss proportional to distance. This involuntary “broadcasting” makes it possible to gain access to network not only to authorized personnel, but also to everyone in the neighbors (as done in “warchalking” or Node Runner by hackers searching for accessible wireless network nodes, marking unprotected zones with conventional chalk signals.) Yet, same security problems are encountered in wireless ad hoc (multi–hop) networks: in fact, how sure can be a node about its outgoing traffic not being tampered with? From a pure cryptographic point of view, ad hoc network services do not imply “new” problems: authentication, confidentiality and integrity are issues already found in many other public communication networks. But in an ad hoc wireless network, trust becomes a key problem (Frodigh et al., 2000). Because the communication (radio) medium cannot be trusted (being intrinsically unprotected), cryptography is a mandatory choice, as well as relying on the cryptographic keys used. So, the basic idea is that of creating trusted key relationships without an external issuing Certification Authority (CA). In fact, a wireless ad hoc network

36

comes out from spontaneously connecting and mobile nodes, and there is no warranty for a node to get trusted public keys from other nodes, nor they can exhibit third–party certificates. Nonetheless, if inter–node trust delegation is permitted, then nodes having already established trusted relationships may extend this privilege to other group members: let us see in detail how it works. The method is based upon a PKI (public key infrastructure) system. Suppose all nodes have connectivity with each other (e.g. through a reactive routing protocol), as shown in Figure 4. Initially, node N1 takes the role of server node in trust delegation procedure, and initiates the trusting process broadcasting a start message to the network. Every node receiving such a message forwards another one containing the trusted public keys set. N1 may so establish a trusted relationships map and identify them. In Figure 4 groups A, B and C participate to the trust chain. All nodes in B indirectly take part to the trust relationship with N1 (through node N3). Node N1 may so collect the signed keys received by group B through N3. Nodes in group C, instead, have not trusted relationships with node N1. However, a trusted relationship between nodes N7 (belonging

Mobile Ad Hoc Networks

Figure 5. Key exchange

to group C) and N1 can be “manually” created by means of trusted keys exchange. Node N1 can now collect the signed keys received from group C through node N7 (see Figure 5). Now, node N1 is ready to forward a message to the network containing all the collected signed keys. This procedure creates trusted relationships between nodes in groups A, B and C, and forms a new trusted group.

Protocol Design Issues As already pointed out, there is a growing interest in the computer science community to formulate new protocols (or, at the very least, to understand the behavior of currently used protocols) under the hypothesis of strategic behavior of the network nodes. In case of ad hoc networks, however, very few results are known, which almost always apply to routing. As an example of strategic behavior, a node may decide not to forward other nodes’ traffic, since during a message transmission a node consumes relatively more energy than during idle times. Of

course, if the majority of the nodes acts this way, no multi-hop traffic is possible. However, without cooperation, the ubiquitous computing scenario depicted in Figure 5 cannot occur since it is not economically feasible (Mas-Colell et al., 1995). The attention of a growing number of researchers is thus focusing on the design of protocols for ad hoc networks using the tools of game theory and mechanism design. Two properties are especially sought: rationality and truthfulness. According to the first, an agent is always motivated to participate in the game (i.e., the protocol), since its utility cannot decrease as a result of the participation. Truthfulness means that the best strategy (i.e., the one that maximizes its utility) for an agent is to behave according to the protocol. To this end, some forms of payment can be provided to motivate the players to act truthfully. Obviously, if the agents are rational and truthful, the protocol will achieve some target social goal (such as setting up power efficient end-to-end routes for all possible pairs of nodes.) Later on some practical ideas for the development of a routing protocol are discussed, upon which future research directions are based to make it rational and truthful.

37

Mobile Ad Hoc Networks

As already mentioned, there are very few results dealing with the problem of strategic behavior in wireless ad hoc networks. At the routing level, protocols have been proposed for the establishment of routes among pairs of nodes using incentive payments to motivate intermediate nodes to act loyally (Buchegger & Le Boudec, 2002, 2002a, 2002b; Buttyan & Hubaux, 2000, 2001, 2003; Zhong et al., 2003). Anderegg & Eidenbenz (2003) introduced the first truthful routing protocol: it has however a number of shortcomings, being characterized by a high communication overhead and, more importantly, not enjoying the rationality property.

MANET ROUTING PROTOCOLS Introduction In wireline networks the task to establish the path that data packets must follow is carried out only by a limited group of devices called routers. In wireless ad hoc networks, instead, every host acts both as a router and a packet sender, so the classical routing protocols used by wireline networks are not applicable at all to MANETs. Existing routing protocols may be classified following three criteria: Based on the logical organization through which the protocol “describes” the network: from this point of view they may be divided in “Uniform” and “Non Uniform” routing protocols. In the first case, every node generates path control messages while answering (incoming) path control requests: all nodes have the same function. In the latter, instead, the way in which nodes generates and/or answers path control messages may be different for different group of nodes. Non Uniform protocols may be in turn logically subdivided into Neighbor Selection and Hierarchical: in the first case every node selects a neighbor node subset to calculate the path data packets must follow;

38

in the second case, instead, nodes auto-organize in groups called clusters. For every cluster there exists a “controller” (master or cluster head) node, which has the task of coordinating the traffic inside it. It is obvious that Non Uniform protocols considerably reduce the signaling traffic vs. Uniform ones, thanks to the less number of nodes assigned to path calculations; but they pay an higher overhead due to the maintenance of complex high level structures. Based on the way routing information is obtained: from this point of view, protocols may be divided in Proactive (or Table-Driven), Reactive (or On-Demand) and Hybrid. The first continuously keep routing information updated through packet exchange at fixed time intervals: this allows immediate availability of routing at every request. The disadvantage is that Proactive algorithms produce signaling traffic also when no data packet is being transmitted; this may cause problems of excessive traffic load in the network, especially when nodes are rapidly moving, because the time interval between routing information transmissions has to be inversely proportional to the velocity with which nodes are moving, so as to get a correct routing. Examples of Proactive protocols are DSDV (DestinationSequenced Distance-Vector) and WRP (Wireless Routing Protocol.) In Reactive protocols a procedure is needed to establish the correct routing path only when packets are to be transmitted; in such a way signaling traffic is reduced, but with increasing delivery times (Johansson et al., 1999). Examples of Reactive protocols are AODV (Ad hoc On-Demand Distance Vector), DSR (Dynamic Source Routing), and TORA (Temporally Ordered Routing Algorithm.) finally, Hybrid protocols try to reach both previous protocols’ advantages, restricting Proactive algorithms application only to the nodes near the one doing data packet transmission. An example of Hybrid protocol is ZRP (Zone Routing Protocol.)

Mobile Ad Hoc Networks

Based on how the routing path is created: from this point of view, protocols may split in Source Routing and Non Source Routing. In the first ones the sending node determines the complete path to the destination, registering it directly into the packet; so, intermediate nodes only retransmit packets to those addressed by the previously established path. In the latter, instead, the only routing information contained in data packets is that represented by the best neighbor node to which communication has to be forwarded; consequently, every node must be able to optimize routing decisions.

Traditional Reactive Routing: Route Request Process Wireless ad hoc networks are generally made of nodes such as notebooks, PDAs, mobile phones. The characteristic of ad hoc networks is to have frequent changes in topology. In addition, to keep track of topology, there is a significant commitment of resources and a considerable overhead. Protocols of reactive type were designed for these environments. The aim is not to keep track of network topology: let us see the process (Paoliello-Guimaraes & Cerda-Alabern, 2007). If a node needs to reach a destination, it starts a discovery process to find the path. This process begins through the transmission – by the source node – of broadcast messages of Route Request (RREQ) type, with TTL set to 1 (Zou, 2005; Paoliello-Guimaraes & Cerda-Alabern, 2007). This RREQ message will only pass through a single node because of its TTL set to 1. Each message has a sequence number, so that only the first message is considered, while its subsequent copies are discarded. When a node receives the first copy of a RREQ from a source node, it stores the address, thereby establishing a return path (reverse route.) When the first RREQ reaches the destination, a reply message of type Route Reply (RREP) is sent to the source through the return path. This type

of protocol is generally efficient for a single rate network; in a multi-rate network, however, what counts is not to minimize the number of jumps to reach a destination, but the total throughput on a given routing. An existing technique taking into account, instead of the number of hops, the throughput is the MTM (Medium Time Metric) (Awerbuch et al., 2007; Paoliello-Guimaraes & Cerda-Alabern, 2007). In this technique a cost inversely proportional to the speed of the link is established; hence, the minimum cost link is chosen. This chapter fits in this area. In fact, in choosing the path, it is not enough to consider only the cost of the link: its stability should also (and perhaps especially) be considered (Dube et al., 1997).

Protocol Issues The Problem of Routing Instability in High Mobility Networks Although existing routing techniques are of indisputable validity, as a result of lengthy trials conducted in wired networks, a problem which causes the performance loss in wireless ad hoc networks impacting on the route discovery processes is the same routing instability, given that we are dealing with high mobility networks. What is “routing instability”? Let us consider a node represented by a mobile phone transmitting while in movement and think how variable is the signal received from a surrounding node as the issuer node moves in a closed or open environment. The level of the received signal, changing constantly, causes a continuously variable ratio of Signal-to-Noise (S/N), altering the bit-rate and consequently the “cost” of the link. This variability would lead to a continuous instability of routing, causing a continuous search of the “best path”. This implies an overhead’s increase impacting greatly on the performance and throughput of the entire network.

39

Mobile Ad Hoc Networks

A technique that keeps track of link stability is now presented, so as to avoid too unstable links in the route discovery process.

Table 1. Node link transitions

Keeping Track of Routing (In)Stability Keeping memory of stability means understanding how stable are connections between nodes; the idea is to have a table maintaining information associated with each link on its state transitions. With the word “transition” you can consider the link’s moving from one transmission intensity (measured in dBm and equal to the signal/noise ratio) to another: Table 1 illustrates each node’s link associated with its number of transitions.

Stability Index and Threshold Let us now define what causes the increase in the number of transitions associated with the link. In order to record the link’s stability, omit all transitions lying within a defined tolerance, i.e. those which do not make a significant loss in link’s performance. The key idea is to record a transition whenever the link gets a difference between the new Ii and the previous Ii-1 sampled transmission intensity outside a predefined absolute percentage δI. So you keep track of a transition when: Ii − Ii − 1 > δI Ii − 1 In order to correctly keep track of transition frequency, it is advisable to sum the number of transitions of a link compared to a period of observation. For example, if N is the number of transitions in the time interval ∆T, the frequency F will be: F=

40

N ∆T

Node link

No. of transitions

L1

N1

L2

N2

L3

N3

...

...

Observation’s Time Interval To establish a statistical time interval ∆T is not simple. You can guess it to be inversely proportional to the average transmission intensity of the links and directly proportional to the number of nodes. Thus, given a network of N nodes, with links’ average intensity μI, you can say that: ∆T α

N µI

After this interval the various counters (column “Number of transitions” in Table 1) are zeroed. At the end, a maximum threshold α for the number of transitions in the time interval remains to be defined. Consider, for example, a time interval ∆T = 300 msec. and a possible maximum value Fmax for the transition frequency F of one transition every 15 msec. From that, you have α = ∆T · Fmax = 20. In a nutshell, if there are more than 20 transitions within an observation period of 300 msec. (i.e. F > Fmax) you will say that the network link is unstable. In order to practice an effective implementation of the mechanisms given above, one can follow two approaches. The first is to monitor the stability of the link, the second provides for the updating of the link stability table only after a route discovery request. Given the high overhead required by the first approach, it seems preferable to implement the second as better detailed later.

Mobile Ad Hoc Networks

Influence of Best Path Choice in Route Discovery When deciding on the best path two alternative approaches called “Link Stability” and “Link Rate” are considered and described below in detail.

Coming back to the above sketched technique, the algorithm − among two links of equal speed − will choose (only under such conditions) the stablest. To define this stability the same considerations outlined in the previous technique can be done.

Link Stability

Protocol Design and Implementation

This technique, as the term shows, prefers link stability and, then, in the route choice to build the path it excludes a priori all links having a transition frequency F above a certain threshold. Returning to our example, if Fmax = 1/15 is the threshold corresponding to a transition every 15 milliseconds, all links having F > 1/15 will be excluded from the choice. Note that, albeit being true that the stablest link has to be chosen, a stable link could also be one with a zero (i.e. not working) signal intensity. Therefore a minimum threshold Imin should be set for the link intensity, below which the choice cannot be done even if the link is very stable. So, considering threshold values Fmax and Imin a network link is stable when I ≥ Imin and F ≤ Fmax.

From what said, the focus is here on the approach called “Link Stability”, in which the characteristic parameters for network monitoring are highlighted, i.e. the transition frequency of the received signal intensity (dBm) and the signal intensity itself. In the protocol design and implementation a crucial role is played by the link stability table. To optimize efficiency, the table will be updated at the beginning of every route discovery process and used in the same process to identify the route.

Link Rate In this technique stability becomes of secondary importance: the speed of the link is in any case to be preferred. So, when choosing routes for the construction of the best path, the stablest link will be chosen only within equal links’ cost (at an equal speed.) But what does equal speed mean? first, it should be noted that from a practical point of view having two links of the same speed may not correspond to reality, if not for a purely random case. Therefore two links are of “equal speed” if the difference in speed between them is no more than 20%. E.g., if the link L1 has a bit rate V1 = 100 Mbps you can say that a second link L2 has the same speed V2 if 80 Mbps ≤ V2 ≤ 120 Mbps.

The Routing Process Each node manages a routing-path table keeping track of all incoming and outgoing connections. The table contains, with respect to a classic routing table, not the mere next hop off an interface but the entire route (that is, all node addresses belonging to the route to destination) and a time-stamp field used to delete the obsolete routes not used for more than a threshold time limit, as shown below. Observing Figure 6, the upper layer (Application Layer) sends a route request to the network layer. The routing protocol (interior protocol) verifies if the destination address already exists in the device routing-path table. If the destination address doesn’t exist, the route discovery process is activated in order to enter the destination node address and its path in the routing-path table. If the destination address exists in the device routing-path table, the relative path (included in the routing-path table) will be used and updated by the real time-stamp value of the sender. In MANETs, the paths included in the routing-path table cannot be static since the network topology

41

Mobile Ad Hoc Networks

Figure 6. WLAN communication of two nodes (source CNET)

changes very frequently; the responsibility of route availability is demanded to the upper layer (because there may be an expired timeout waiting for an acknowledgment.) It’s still the upper layer to decide whether to delete a route from the table, even if its time limit has not expired: this means that the upper layer will order a route discovery every time a packet nondelivery happens.

The Stable/Unstable Link Definition To define the (un)stable link during the operation of route discovery process, the node must collect n transmission intensity values Ii, i = 1..n (expressed in dBm) during a statistical time interval ∆T, updating the minimum value Is = mini {Ii} of the sampled data and a transition counter C every time the absolute relative difference between the two last observed values is outside the predefined percentage threshold δI. These values, at the end of ∆T, will be used to check that the received signal intensities Ii are higher than or equal to a minimum acceptable threshold Imin and if they overcome the percentage threshold δI not too often.

42

In the example of Table 2, five intensity signal measurements in dBm are sampled over an observation period ∆T = 300 msec and checked against a minimum intensity threshold Imin = −85 dBm and a percentage threshold δI = 20%; a maximum transition frequency Fmax = 1/15 is also assumed, from which a maximum acceptable transition counter α = ∆T · Fmax = 300 · 1/15 = 20. The transition counter C is increased every time the percentage ratio of the last two sampled dBm values overcomes the threshold. Note that the minimum intensity threshold is never violated (i.e. IS = −80 > Imin = −85). So, the link is stable because also the transition counter C = 3 does not overcome the maximum α = 20.

Route Discovery Packet Fields The route discovery packet contains the following fields: destination node address; sender node address; sender node time-stamp; hop-count (number of links, or nodes, passed through); number of stable links;

Mobile Ad Hoc Networks

Table 2. Link stability table I i − I i −1 I i −1

Sample #

Intensity Ii (dBm)

Minimum value IS (dBm)

1

-50

-50

0%

0

2

-70

-70

40%

1

3

-80

-80

14%

1

4

-40

-80

50%

2

5

-70

-80

75%

3

pointer to a stack containing addresses of nodes traversed from the sender (bottom) to the recipient (top.)

Route Discovery Algorithm The route discovery process can be summarized as follows. Every node initiating a transmission activates a route discovery process. The transmitter node sends a packet including the destination address and the above mentioned fields. Every node receiving the packet checks if the destination address matches itself.

Matching The receiving node stores the return path, and the percentage of stable links over all links traversed. Later, after receiving the first packet, it waits for any other route discovery packet related to the pair sender-timestamp for a specified time ∆TB. If in this time another route discovery packet arrives, the node compares the percentage of stable links over all links passed through with the previous stored percentage. If it is higher, the new relative path and new percentage will be stored, otherwise the packet will be ignored. All other arriving route discovery packets will be treated the same manner until ∆TB expires. The recipient, once elected

% of oscillation =

Transition counter C

the best route among all considered in the above said interval: sends an ACK using the final reverse route. This acknowledgement will be uniquely associated with the route discovery packet transmitted by sender with its time-stamp included; inserts the reverse route (the winning route) in its routing-path table, binding it to a local time-stamp. Reverse-route will be used as long as the routing is valid, i.e. while the recipient is reachable.

No Matching The receiving node checks whether its address is in the stack of node addresses traversed: if yes, it drops the packet, since it is a broadcast route discovery packet previously handled by itself, so the broadcast storm effect will be excluded; if not, it sends a broadcast route discovery packet to the same destination address adding the node address from which the packet is coming from, plus a stable link counter increased by 1 (in case the receiving node has detected a stable link) and a counter storing all links traversed. Return to Step 3. It would be useful to conduct proper simulations to test the described algorithm and obtain significant values for the following parameters: statistical time interval ∆T;

43

Mobile Ad Hoc Networks

number n of samples considered in the time interval; minimum threshold signal intensity Imin in dBm; percentage threshold of transmission swings δI ; allowed frequency swings limit Fmax to define a stable link; % of stable links over total traversed links for routing-path table; sender node (waiting for acknowledgement) time-out to initiate a new route discovery; time-out to declare an “old” route in the routing-path table; recipient node wait-time ∆TB to receive the route discovery.

CONCLUSION AND FUTURE RESEARCH DIRECTIONS Future research trends reflect the general goals introduced at the end of paragraph Distributed computing, which characterize the new scenario for the understanding and the design of open networks. First of all, it is useful to understand − with respect to some potential application scenarios for ad hoc networks − under which conditions the known “classical” protocols guarantee the achievement of the requested social goals. This kind of analysis has been done with respect to some Internet protocols, such as BGP routing (Markakis & Saberi, 2003.) Future work must possibly answer the following questions, among others. Under which scenario (i.e., traffic model, node payoff, etc.) can we avoid the formation of coalitions, that is, the formation of subsets of nodes that can increase their payoff by seceding from the rest of the network? In the negative case, can some relatively simple payment mechanisms help driving the network activity towards socially better outcomes? It is likely that, under many potential scenarios, the known protocols for ad hoc networks are not strategy proof (i.e., the rationality and truthfulness properties are not both satisfied.) Hence, the more

44

ambitious research directed towards the design of new protocols resilient to strategic manipulations. As pointed out in paragraph Protocol design issues, there are already some proposals for the routing level, although no strategy proof protocol is known to date. However, it would be interesting to investigate strategic issues also at the topology control level as well as at the application level. The simple techniques exposed are suited to any type of mobile ad hoc network and any kind of speed, by the definition of the indicated parameters. Therefore, this methodology can probably be implemented in any type of network environment, even in networks with very high density of nodes, as wireless networks in delimited environments such as university campus, airports, shopping malls, etc. It would be useful to study how these techniques, when implemented, impact on the energy consumption of nodes. This study would not aim at finding an absolute value of absorbed energy, rather a percentage value relative to the generated network overhead.

REFERENCES Anderegg, L., & Eidenbenz, S. (2003). Ad hocVCG: a truthful and cost-efficient routing protocol for mobile ad hoc networks with selfish agents. In Proceedings ACM Mobicom (pp. 245–259). Awerbuch, B., Holmer, D., & Rubens, H. (2007). The medium time metric: High throughput route selection in multirate ad hoc wireless networks. Mobile Networks and Applications, 253–266. Blough, D. M., Leoncini, M., Resta, G., & Santi, P. (2002). On the symmetric range assignment problem in wireless ad hoc networks. Proceedings IfiP Conference on Theoretical Computer Science (pp. 71–82).

Mobile Ad Hoc Networks

Blough, D. M., Leoncini, M., Resta, G., & Santi, P. (2003). The k-neigh protocol for symmetric topology control in ad hoc networks. Proceedings ACM MobiHoc, 03, 141–152.

Dube, R., Rais, C., Wang, K., & Tripathi, S. (1997). Signal Stability-Based Adaptive Routing (SSA) for Ad Hoc Mobile Networks. IEEE Personal Communications, February (pp. 36–45).

Buchegger, S., & Le Boudec, J. (2002). Cooperative routing in mobile ad hoc networks: current efforts against malice and selfishness. Lecture Notes on Informatics, Mobile Internet Workshop, Informatik 2002, Dortmund, Germany.

Fan, Z. (2004). High throughput reactive routing in multi-rate ad hoc networks. Electronics Letters, 40(25), 1591–1592. doi:10.1049/el:20046622

Buchegger, S., & Le Boudec, J. (2002a). Nodes bearing grudges: towards routing security, fairness, and robustness in mobile ad hoc networks. Proceedings of the Tenth Euromicro Workshop on Parallel, Distributed and Network-based Processing (pp. 403–410). Buchegger, S., & Le Boudec, J. (2002b). Performance analysis of the CONfiDANT protocol: cooperation of nodes – fairness in dynamic ad hoc networks. Proceedings of ACM Symposium on Mobile Ad Hoc Networking and Computing (MobiHoc). Buttyan, L., & Hubaux, J. (2000). Enforcing service availability in mobile ad hoc WANs. Proceedings of IEEE/ACM Workshop on Mobile Ad Hoc Networking and Computing (MobiHoc). Buttyan, L., & Hubaux, J. (2001). Nuglets: a virtual currency to stimulate cooperation in selforganized ad hoc networks. Technical Report EPFL. DSC. Buttyan, L., & Hubaux, J. (2003). Stimulating cooperation in self-organizing mobile ad hoc networks. ACM/Kluwer. Mobile Networks and Applications, 8(5), doi:10.1023/A:1025146013151 Clementi, A. E. F., Penna, P., & Silvestri, R. (1999). Hardness results for the power range assignment problem in packet radio networks. Proceedings 2nd International Workshop on Approximation Algorithms for Combinatorial Optimization Problems (RANDOM/APPROX ’99). LNCS (1671) (pp. 197–208).

Frodigh, M., Johansson, P., & Larsson, P. (2000). Wireless ad hoc networking: the art of networking without a network. Ericsson Review, No. 4 (pp. 248–263). IEEE (1999). Wireless LAN Medium Access Control (MAC) and Physical Layer (PHY) specifications: Higher speed Physical Layer Extension in the 2.4 GHz Band. IEEE Supplement to Part 11, Rev. IEEE STD 802.11b. IEEE (2003). Wireless LAN Medium Access Control (MAC) and Physical Layer specifications: Further Higher-Speed Physical Layer Extension into the 2.4 GHz Band. IEEE Draft Supplement to Part 11, Rev. IEEE STD 802.11g/D8.2. Johansson, P., Larsson, T., Hedman, N., Mielczarek, B., & Degermark, M. (1999). Scenariobased performance analysis of routing protocols for mobile ad hoc networks. Proceedings of the ACM/IEEE International Conference on Mobile Computing and Networking (pp. 195–206). Jung, E. S., & Vaidya, N. H. (2002). A power control MAC protocol for ad hoc networks- Proceedings ACM MobiCom 02 (pp. 36–47). Liu, L., & Li, B. (2002). Capacity-aware topology control in mobile ad hoc networks. Proceedings IEEE International Conference on Computer Communications and Networks (pp. 570–574). Markakis, E., & Saberi, A. (2003). On the core of the multicommodity flow game. Proceedings ACM Electronic Commerce conference.

45

Mobile Ad Hoc Networks

Mas-Colell, A., Whinston, M., & Green, J. (1995). Microeconomic theory. New York: Oxford University Press. Mobile landscape - Graz in real time. Retrieved July 16, 2009, from http://senseable.mit.edu/graz/ Nisan, N., & Ronen, A. (1999). Algorithmic mechanism design. Proceedings of the 31st Annual ACM Symposium on the Theory of Computing (pp. 129–141). Paoliello-Guimaraes, R., & Cerda-Alabern, L. (2007). Adaptive QoS reservation scheme for ad hoc networks. Lecture Notes in Computer Science, 102–112. doi:10.1007/978-3-540-70969-5_8 Papadimitriou, C. H. (2001). Algorithms, games, and the Internet. Proceedings of the 33rd Annual ACM Symposium on the Theory of Computing (pp. 749–753). Ramanathan, R., & Rosales-Hain, R. (2000). Topology control of multihop wireless networks using transmit power adjustment. Proceedings - IEEE INFOCOM, 404–413. Roughgarden, T., & Tardos, E. (2000). How bad is selfish routing? Proceedings of the 41st Annual Symposium on the Foundations of Computer Science (pp. 93–102). Zhong, S., Yang, R., & Chen, J. (2003). Sprite. A simple, cheat-proof, credit-based system for mobile ad hoc networks. Proceedings of INFOCOM (pp. 1987–1997). Zou, S., Cheng, S., & Lin, Y. (2005). Multi-rate aware topology control in multi-hop ad hoc networks. Proceedings IEEE Wireless Communications and Networking Conference (WCNC’05) (pp. 2207–2212).

KEY TERMS AND DEFINITIONS Link: a hardware connection linking two or more electronic devices, normally using different

46

types of cables each designed for a certain standard of data transmission. In addition to the electric cable is also possible to make a connection through fiber optics, radio waves and infrared (IrDA). Link Stability: transmission error rates (due to signal weakness or external environment factors such as white noise and wireless interference) can result in an unstable link. A link is stable if it has a low signal variation frequency with a minimum acceptable signal intensity. MANET: (Mobile Ad hoc NETwork): an independent system of mobile devices connected by ad hoc wireless links. All nodes in the system cooperate in order to correctly route packets in multihop forwarding mode. Due to the unpredictable mobility of nodes, the network topology may change constantly. Ad hoc networks are built and used as appropriate in extremely dynamic environments, not necessarily with the help of an already existing infrastructure, such as after natural disasters, military conflicts or during emergencies. Multi-Rate: a transmission system in which speed (rate) can vary from point to point. The use of different sampling rates within the same system offers several advantages such as lower computational complexity, adequate transmission rate and less memory capacity requirements. Network: a system that allows the sharing of information and resources (both hardware and software) among several devices (hosts), providing an information transport service to a user population distributed over a more or less extensive area. Computer networks generate potentially high volume traffic, unlike the telephone, efficiently managed through the technology of packet switching. Protocol: a set of rules formally described, defined in order to facilitate communication between one or more entities (network protocol, in case of remotely connected entities). All these rules are defined by specific standards (see ISO/ OSI), of many different types, depending on the entities and communication means involved.

Mobile Ad Hoc Networks

Route Discovery: a request to find the best available route to the destination, when sending a message. Routing: in packet-switching networks, routing is the function of a device (router) that decides the best path along which to send a received packet. Each packet is forwarded from the source to a router, and from this to the next, until reaching the desired destination. The router often uses a table with destination network addresses to decide where to send each packet. The format

of this table and the way it is populated and possibly updated are specific to the different routing protocols adopted. Wireless: the term covers communication systems between electronic devices, which do not use cables (traditional cabled systems are called wired). Generally, the wireless uses radio waves at typically low power, but also infrared radiation or laser.

47

48

Chapter 4

Convergence of Wireless Technologies in Consolidating E-Government Applications in Sub-Saharan Africa Bwalya Kelvin Joseph University of Johannesburg, South Africa Chris Rensleigh University of Johannesburg, South Africa Mandla Ndlovu Botswana Accountancy College, Botswana

ABSTRACT The convergence of wireless applications presents a greater hope for consolidating e-Government (ICTenabled or online government) applications even in resource-constrained countries such as those in Africa. This chapter presents an exploratory study that aims at discussing the extent as to how convergence of wireless technologies from different vendors promises to contribute to the consolidation of e-Government applications in Sub-Saharan-Africa (SSA). This is done by reviewing the different adoption stages of ICT and e-Government in SSA. It looks at challenges facing adoption of wireless technologies (GSMs, Wireless Internet Access, satellite transmission, etc.) across all the socio-economic value chains in SSA. The chapter looks at Botswana and South Africa as case studies by bringing out the different interventions that have been done in the realm of facilitating a conducive environment for the convergence of different wireless technologies. Out of the analysis of legal, regulatory, market and spectrum policies affecting the adoption of wireless communications in SSAs, the chapter draws out recommendations on how to consolidate wireless communications to be adopted in different socio-economic setups (e.g. e-government, e-Health, e-Banking, etc.). DOI: 10.4018/978-1-60960-042-6.ch004 Copyright © 2011, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited.

Convergence of Wireless Technologies in Consolidating E-Government Applications in Sub-Saharan Africa

INTRODUCTION Wireless technologies (especially Internet-ready mobile phone technology) have seen higher adoption rates in the SSA region partly due to an escalation in the existence of Wireless Application Service Providers (WASPs) in SSA. This penetration has encouraged computing mobility which has encouraged spill-over applications to almost all the major sectors of the socio-economic hierarchy. One such e-application with great potential is the application of wireless technologies in the e-Government domain. Appropriate convergence of the different wireless technologies will enable increase access to information and public services in the framework of e-Government. As Africa’s technological landscape has kept growing exponentially, it is important to take advantage of this growth and promote e-Government which will hopefully bring about responsive, transparent and accountable management of national resources. Many literature sources that point to the fact that mobile technologies are growing at an alarming rate in Sub-Saharan Africa (SSA) (TRASA report, 2009; Bwalya, 2009; Touré, 2006; Graham et al., 2006; ITU Report, 2008). Telecommunication service providers in the SSA region have made Internet access possible through a multitude of Information and Communication Technology (ICT) devices such as personal digital assistants (PDAs), mobile phones etc. This means there is an increase in the number of people accessing the Internet making it possible for governments in SSA to consider implementing the e-Government governing model. In addition, e-Government can be a prerequisite to strategic initiatives avoiding rampant corruption and red-tape that characterizes most Africa’s government organs and thwarts effectiveness of public service delivery systems (Bwalya, 2009). This chapter defines E-Government as the use of ICTs to provide interactive public services linking then government, citizens and businesses. E-Government enables transparent government processes ushering in improved

service delivery/transactions which encourage participatory governance where citizens and businesses are accorded the chance to interact effectively with government departments, organs and line ministries. However, setting up a full-fledged e-Government interactive environment for SSA countries has been a huge challenge because of the different costs that come with its implementation. Before we look at the different challenges of e-Government implementation, let’s briefly look at the different benefits of using e-Government. The convergence of wireless technologies entails that different wireless ICT platforms cannot decode certain signals from a named source because of differences in IEEE standards. An example of this is that you can have a situation where WiFi (IEEE 802.11) can decode some signals from a source and WIMAX (IEEE 802.16) cannot decode the same signals that WiFi can. This may be because of protocol mismatch, different in synchronization and frequency bands. The figure below presents one context of convergence of wireless technologies where a WiFi device tries to join a WiMAX network and it can’t decode the signals because the base station which encodes the said signal uses WiMAX-OFDM. This convergence of wireless technologies can happen amongst many technologies such as GSM, CDMA, TDMA and iDEN or 1G, 2G, 2.5G, 3G to 4G in cellular phones. This convergence may also entail ability of mobile agents/devices to decode encoded signals on networks such as wireless LAN, MAN, sensor networks, RFID, and so forth. Since, generally, even ordinary individuals now have access to Internet-enabled mobile phones; implementation of e-Government on mobile platforms has much promise for Africa. This is compounded in the understanding that the universality in the signal decoding capabilities of different mobile devices is key to global access of e-Government information and public services in ubiquitous environments. This form

49

Convergence of Wireless Technologies in Consolidating E-Government Applications in Sub-Saharan Africa

Figure 1. Convergence of wireless technologies

of e-Government is called m-Government (discussed in the next section). In the SSA, a study by Maumbe and Owei (2006) investigated the adoption of m-Government in South Africa and outlined the different challenges that prevent the transition ofm e-Government to m-Government. Many authors (Bwalya, 2009; ) have outlined the different benefits (Mutula et al., 2010; Aladawi, Yousafzai, & Pallister, 2005; Warkentin, Gefen, Pavlou, & Rose, 2002) that come with the implementation of e-Government: a) the reduction in cost and the increase in the efficiency levels in public service delivery platforms brought about by appropriate coordination mechanisms amongst different government line ministries, organs and branches, b) growth in confidence levels in government processes as with the proper implementation of e-Government, there is resource accountability and responsiveness, c) reduction in corruption levels by public servants, d) facilitation of participatory democracy as citizens are involved in the different governance processes, e) ensures appropriate functioning for interaction mechanisms between the state, business and citizens, f) ensures openness and transparency of activities carried out by the public administration bodies for business and society, g) enables citizens to participate in the governance processes anywhere

50

and at anytime and h) encourages the interaction amongst different developmental partners (e.g. government, businesses and ordinary citizens) in exchanging governing ideas that could later transform the socio-economic status of a country. These different benefits are achieved after the major principles for e-Government applications have been appropriately applied. Reviewing the different countries that have successfully implemented e-Government reveals that the following principles should be employed for e-Government to succeed: a) platform for common national strategy formulation that would define goals, principles, objectives, indicators evaluating the process of e-Government implementation (implementation roadmap or e-Government strategy); b) synchronization of e-Government implementation process with administrative reforms and mainstreaming of e-Government systems at all levels of the government hierarchy e.g. at local government levels, c) creating tax and institutional mechanisms promoting demand for ICT and d) openness of e-Government introduction program development and accountability on its implementation. E-Government is looked at as a multi-dimensional entity which depends on availability of appropriate ICT infrastructure, convergence of

Convergence of Wireless Technologies in Consolidating E-Government Applications in Sub-Saharan Africa

ICT technologies, ICT skills of people, acceptable usability levels of web portals, cost of Internet access, cost of ICT devices, ICT culture, and so forth. For e-Government to be intermittent challenges to e-Government adoption can be attributed, principally, to lack of proper ICT infrastructures e.g. lack of fibre optical networks for broadband Internet communications. Very limited fixed networks are available in most parts of Africa (Bwalya, 2009). This further justifies the need to encourage provision of e-Government services through mobile platforms (m-Government). On another hand, in some places, telecommunications operators and other added-value service providers face a lot of difficulties in competing with stateowned competitor companies who enjoy wide monopolies and are fond of practices that do not encourage fair competition. This means it becomes difficult to have a mature telecommunications environment further negatively impacting on eGovernment efforts. All this is compounded by lack of proper monitoring, legal, regulatory and institutional frameworks in place. Other challenges have been lack of reliable power supply in many zones (mainly rural zones) where people can not charge the mobile terminals, or connect a computer to access to Internet, lack of the technological culture to understand the possibilities that these technologies could offer, and the difficulties to sure the investment return due the economical situation of great part of population, and so on. Another issue that confronts massive adoption of wireless technologies is the security issue. Enforcing of security inn wireless environment is more challenging and expensive that may be in convertible wired networks. The aforementioned challenges clearly justify the need to seek understanding of the premise of convergence of wireless technologies and its impact on e-Government. This further brings us to investigate the adoption trends of m-Government in the SSA and to explore on what interventions have been put in place to encourage the proliferation of e-Government in SSA. This chapter does

just that. To do this, the chapter surveys how the convergence of different wireless technologies, and other interventions, have aided in promoting adoption of e-Government in South Africa and Botswana. The next section discusses the background and gives a succinct motivation of this survey in a view to unearth the potential for the convergence of wireless technologies in promoting e-Government. It also discusses the pursuit of wireless e-Government applications in the pursuit of promoting the mobile governance model (m-Government), which is an emerging form of e-Government. Thereafter, the chapter presents findings on the exploratory studies of different convergence aspects of e-Government models in South Africa and Botswana. The next section after this gives the solutions and recommendations for advancing convergence of wireless technologies in facilitating m-Government applications in SSA and immediately thereafter, the future research directions are given. The chapter ends with a conclusion that gives a recap of what has been discussed in this chapter. Background In this plethora of technologies, convergence of telecommunications and wireless communications brings together voice telephony and more affordable wireless (mobile) services into more efficient interactive platforms. Broadband, a key requirement for productive Internet access, is still unavailable and/or unaffordable in most of the developing world (Lanwin, 2002; Bwalya, 2009; Pheko, 2009; Odinma, Oborkhale, Kah, 2007). These different convergence platforms aim for appropriate information dissemination and exchange of ideas, pulling people of different backgrounds out of different situations such as poverty and ameliorating their socio-economic standing. Convergence of wireless technologies offers great hope as a platform for e-Government applications in SSA as Africa has the highest growth rate of mobile technology adoption worldwide (Lanwin, 2002; Bwalya, 2009; Du Preez, 2009). As aforementioned, convergence of wireless technologies has brought about a re-

51

Convergence of Wireless Technologies in Consolidating E-Government Applications in Sub-Saharan Africa

ality of the proliferation of a converged form of e-Government where public services and sharing of information is done through mobile platforms (mobile government: m-Government). Before we proceed further, let’s have a look at definitions and implications of m-Government. The extension of e-Government services to m-Government coupled with massive growth in mobile and wireless technologies has been acknowledged by several authors (Amitava and Agnimitra, 2005; Amailef and Jie; Karantjias, Papastergiou and Polemi, 2007). In a nutshell, m-Government is looked at as the use of mobile and wireless communication technology in government for service and information delivery to citizens and organizations (Fasanghari and Shamimi, 2009). It is thus in order to refer to m-Government as a complex business strategy for the efficient utilization of all wireless devices providing for instant availability of services and information for better benefits to users Amitava and Agnimitra, 2005). The ultimate goal of m-Government is to push forward the e-Government agenda by utilizing mobile and ubiquitous platforms. Further, Fasanghari and Shamimi (2009) quoting from other sources, mention that m-Government further facilitates service presentation to the citizens with access to the information and services in any time and any place made possible possible through connected wireless tools to the Internet. The paper by Amitava and Agnimitra (2005) suggested that the network architecture that provides m-Government comprises a hierarchical one with various wireless/wired access layers. These layers include the following: the cellulular, short range and the personal network layer (for detailed descriptions of these layers please refer to Amitava and Agnimitra, 2005). Amailef and Jie (2008) has recognized the fact that within the context of e-Government, mGovernment offer more access to information and services for citizens, businesses, and non-profit organizations through wireless communication networks and mobile devices and platforms

52

such as pagers, PDAs, cellular phones, and their supporting systems. This has all been facilitated with the advent advances in Internet and wireless technologies which have brought forth mobility in e-applications. With intermittent growth of mobile technologies in Africa, m-Government presents a new way for enhancing interaction amongst citizens, businesses and government through mobile platforms towards improved content access, transparency, participations, and so forth. At the center of m-Government are issues of interoperability as different sets of mobile devises are involved in mobile e-transactions. This is where convergence of different wireless gadgets in heterogeneous computing environments is sought. In fact, interoperability requires easily identifiable and publishable mobile services as well as clear electronic and mobile interfaces for the establishment of secure and reliable connection points (Karantjias, Papastergiou and Polemi, 2007). Friske (2003) stated that convergence can be perceived through its division into two components: the technical and functional. The technical element refers to the ability of any infrastructure to transport any type of data whereas the functional aspect characterizes the means by which users are able to integrate the functions of computers, television, media and voice into a single device. However, it is worth mentioning that having lots of mobile devices does not enable m-Government. The true measure of convergence in this context is checking which applications are available, and how understandable, accessible and useful they are to the common citizen through the use of different mobile platforms. Different devices within different spectra can now do the same functions and are highly interoperable. Apart from the obvious uses of e-Government (access to government information from portals, downloading of public service forms, and so forth), some of the practical uses of m-Government have been in emergency response systems (Amailef and Jie, 2008). This suffices to mention that mGovernment can be used in many e-applications

Convergence of Wireless Technologies in Consolidating E-Government Applications in Sub-Saharan Africa

not limited to governance in the broader sense of the word. The convergence of different wireless ICT devices can encourage the growth of e-Government in the form of mobile government (m-Government). M-Government implementation involves the use of all wireless and mobile technology, services, applications and devices for improving benefits to the parties involved in e- Government including citizens, businesses and all government units (Kushchu, 2003). In most countries trying to implement e-Government, some of the early adopters of m-Government services include law enforcement, fire fighting, emergency medical services, education, health and transportation, immigration, border and coastal security and disaster response and management (Amine, 2005). M-Government implementations are emerging as one of the additional value-added features for the integrated and flexible data communication and exchange mechanism among government units, citizens and businesses. This can further be supported by the penetration of mobile technology and the relative low cost of entry into mobile connectivity, the convergence of wireless Internet and telecommunication networks, allowing information once only available on a computer to be received through mobile phones, and the shift towards higher data transfer rates and 3G services which promises to make more information available at faster speeds. M-Government offers a relatively lower cost of mobile phone technology versus Internet technology which has drastically lowered the entry barriers for ordinary individuals to participate in e-Government activities. This is because the use of WIMAX and WiFi technologies presents a real and low cost alternative to fixed networks to allow Internet access. This further presents an opportunity for WIMAX to combine with other current ICT infrastructures in these SSA countries where overreliance on high cost Internet connection through GSM/3G networks may not provide wide band for appropriate Internet connectivity.

Different wireless platforms have undergone transformations in order to be of much value to its customers and to be interoperable with other technologies, ICT devices and applications. Odinma et al. (2007) identifies broadband wireless networks as mainly comprising two types: fixed and mobile wireless. These can be Wireless Fidelity (Wi-Fi), which is an IEEE 802.11 standard and Worldwide Interoperability for Microwave Access (WiMax), which is an IEEE 802.16 standard. TRASA (2009) gives a historical overview of different transformations and convergence of different wireless technologies. The second generation (2G) wireless communication employed digital technologies such as the GSM, CDMAOne, iDEN, IS-136 and PDC. In between the 2G and Third Generation (3G) is the 2.5 Generation (2.5G) technologies, namely, GPRS which evolved from the GSM, EDGE which evolved from the GSM but it has higher speed than the GPRS and CDMA2000 1XRTT which evolved from CDMAOne. Al-Sherbaz (2008) acknowledges that WiFi, the widely deployed predecessor technology, is inexpensive and readily available with multiple vendors. But WiMAX offers better spectrum efficiency, data rate, and long-distance capability. Developing of the wireless environment can be facilitated by adopting the Fixed-Mobile Convergence (FMC) which is an analysis of the interworking between cellular networks and a variety of wireless technologies such as WLAN, WiMAX, RFID and UWB (Ghetie, 2008). Rossel (2006) states that the first understanding of mobile government is linked to the increase of space and time in accessing administration services to provide “mobile” with enhanced linkage and informational and transactional options or to develop new citizen opportunities linked with mobility issues and enhanced government services. This entails adapting government and administration services to be accessible in a ubiquitous manner, developing new options and services based on ubiquitous access and multi-channel possibilities, mobile services for mobile organizations, facilitat-

53

Convergence of Wireless Technologies in Consolidating E-Government Applications in Sub-Saharan Africa

ing the mobile state, and mobile administration of administration agents or agencies. Even after having achieved full convergence of wireless technologies or having one of the fastest growth rates in mobile and wireless and convergence networks, there are still some challenges that need to be addressed to facilitate appropriate facilitation of e-Government systems. Lanwin (2002) identifies the different challenges that come with m-Government as a) Privacy and security, b) securing wireless e-mail, c) protection against loss of data, d) securing handheld access to the corporate network, e) deploy and manage security policies for many devices, f) accessibility, g) mobile authentications and h) mobile payments. Specifically, South Africa has faced a number of challenges in implementing m-Government as outlined by Maumbe and Owei (2006). Some of the challenges outlined include the following: a) islands of automation and weak inter-agency information sharing, b) inadequate integration of poor citizens into e-Government, c) slow pace in multi-lingual content development due to lack of content development specialists, d) lack of empirical research to evaluate impact of e-Government on citizen livelihoods, e) privacy and security concerns and f) lack of an empirically validated model for e-Government development in Africa. In the midst of these challenges, a new wave of mgovernment has descended on South Africa. Patel (2005) acknowledges that m-Government in South Africa is in its infancy. The current m-government landscape is one of isolated pilot projects and a handful of full-scale implementations. However, there is increasing interest and acknowledgement that m-government has potential value. South Africa, in conjunction with the ITU and other cooperating partners has launched a project on Rural Telecoms, ICT Services and Entrepreneurship Development. The objective of this project is threefold: 1) to encourage telecommunication operators to provide services in rural/remote areas with appropriate private investments; 2) to foster development of the content of the ICT services

54

through private public sectors collaboration; and 3) to trigger provision of services to the general public by local entrepreneurs with support from microfinance institutions and other banks (ITU, 2008). Such kinds of interventions make sure that there is chance for wireless technologies to further converge. Since e-Government, and correspondingly, m-Government are new technologies that are just being introduced, it is important to understand technology adoption theories and principles. The following section presents a discussion of technology adoption especially e-Government and briefly mentions empirical findings of initiatives aimed at encouraging technology adoption. Fan and Zhang (2006) specifically proposed a conceptual model for government to government (G2G) information sharing in the context of the e-Government environment. This was done in the spirit of reducing bureaucracy that may be evident in many government agencies (Titah and Barki, 2006). With this model, information sharing amongst different government branches was made easier. After reviewing several technology acceptance models such as the Davis’ technology acceptance model (TAM), the Diffusion of Innovation (DOI) model, the unified theory of acceptance and use of technology (UTAUT), Fan and Zhang (2006) identified 8 common factors that have been utilized in different models such as: a) perceived benefits, b) perceived risks, c) top management support, d) IT support, e) costs, f) external pressures, g) critical mass, and h) championship. These 8 factors were accordingly incorporated into a conceptual e-Government adoption model that they proposed. Kamal and Themistocleous (2006) have also conducted a study to ascertain technology adoption in a complex environment such as a local government authority with hierarchical bureacratical structures with utter commitments to outmoded cultural values. Kamal and Themistocleous have identified a set of factors from literature that influence the uptake of e-Government. These are: knowledge of technology risks, IT capabilities,

Convergence of Wireless Technologies in Consolidating E-Government Applications in Sub-Saharan Africa

market knowledge on new technologies, managerial capabilities, project championships, external pressures, citizen’s data privacy and security, and Return on Investments (ROI). In their pursuit to better understand the factors that affect e-Government, Al-adawi et al (2005, P. 2), identified 4 questions they called ‘critical’ for the adoption and encapsulation of e-Government into the socio-economic setups. The following questions were identified and incorporated into a model for e-Government adoption: 1) How are intentions towards the use of e-government formed and to what extent are they related to the actual use of e-government? 2) To what extent the intentions to get information and to conduct transactions differ from each other? 3) What are the beliefs that influence citizens’ propensity to use e-government? How do these beliefs affect their intentions towards the use of e-government? 4) Are there any perception and adoption differences between segments of citizens on the basis of their technology readiness and demographic characteristics? Bélanger and Carter (2008) undertook a study which analyzed the impact of trust and risk perceptions on one’s willingness to use e-Government services and developed a model that include constructs supporting trust of the internet (TOI) and trust of government (TOG). Their study utilized data from a citizen survey that indicated that the disposition to trust positively affects TOI and TOG. Several authors (e.g. Warkentin, Gefen, Pavlou, & Rose, 2002; Bwalya, 2009; Mutula et al., 2010, and so forth) have pointed out that data security, accessibility and perceived confidentiality influence individual adoption of e-Government adoption to a greater extent. The factors affecting e-Government can be divided into individual and organizational. Titah and Barki (2006) have suggested that apart from organizational factors, individual beliefs of citizens have a significant influence on the adoption of e-Government services. With strong reference to Davis’ technology acceptance model of 1989, it is known that individual beliefs such as

perceived usefulness (PU) and perceived ease of use (PEOU) have been considered as the dominant beliefs that affect the intention to adopt or use the technology in a business to consumer (B2C) model (Warkentin, Gefen, Pavlou, & Rose, 2002). Heeks and Santos (2009; p. 3) unleashed a study where they gave some perspectives on the adoption of public sector innovations – “one that understands adoption as based on the behavior of individual actors set within a contextual framework”. Their motivation came from the fact that e-Government faces low usage rates and stated that there exists huge gaps between designers and adopters. In their study, they employed case research to study the practice-based problem of e-Government where it was thought that the experiences of the actors are important and the context of action is critical (also cited in Benbasat et al., 1987). Their study identified different factors that affect the role of the designers (as principals) and adopters (as agents). Colesca (2009) identifies five different steps to making e-Government happen and these are: Develop a vision; conduct an e-readiness assessment; identify realistic goals; get the bureaucracy to buy-in and develop a change management strategy; and build public-private partnerships. These e-Government constructs are accordingly applicable to m-Government. There is utmost need for the potential of wireless technologies to be explored as Africa’s mobile and wireless access to Internet resources keeps on growing. Boyinbode & Akinyede (2008) notes that Africa’s mobile usage has a growth rate of about 904% from 2000 to 2008 and that has made the continent the fastest growing region in terms of mobile usage. With the promising ubiquitous world, convergence of wireless technologies is very vital in that it will encourage a varying degree of e-applications to be performed on any wireless device such as mobile phone or PDA. Thus, for complete convergence to take place, as put by Javaid (2006), there are two fundamental issues: the first being the need for seamless

55

Convergence of Wireless Technologies in Consolidating E-Government Applications in Sub-Saharan Africa

interoperability and pervasive integration of heterogeneous networks in order to offer a variety of diverse services in an always best connected environment, and the second being the need for optimized network resource management, which has been obstructed by the conventional wireless and cellular networks. For the case of Africa, the convergence of wireless technologies depends on the local context. The next section looks at the two cases of the role on wireless convergence as applied in the e-Government platform for Botswana and South Africa.

CASE STUDIES A) Botswana Botswana is an emerging ICT-usage-powerhouse in SSA looking at the number of interventions being implemented towards the same. Recent endeavors have seen it implement massive projects such as the construction of the Kgalagadi optical fibre network (for broadband Internet), full liberalizations of the telecommunications sector, establishment of sound ICT sector regulatory and institutional frameworks, dedicated ICT policy, setting up of Botswana IT hub, etc. In an environment like this, convergence of wireless technologies may be a reality looking at the already established appropriate ICT environment. Gillwald & Stork (2008) did a study where Botswana emerged as one of the countries with a higher fixed-line penetration with 11 to 18% of households having a working fixed-line phone. By the end of March 2002, there were 278 000 mobile subscribers as compared to 142 000 fixed line subscribers (Monnane, 2003). Compared to other countries, Botswana’s mobile penetration (teledensity) has been remarkable, from nothing in 1997 to 14.24 in 2002, only second to South Africa in the region. In terms of mobile lines as percentage of total lines, Botswana is behind Lesotho, South Africa and Tanzania only.

56

With the introduction of cheap Taiwan or Chinese phones on the local telecommunications market, acquiring a handset is no longer a far-fetched dream for Botswana citizens (Batswana). Telecom operators e.g. Mascom and Orange Botswana have started offering Internet accessibility on mobile phones. People in the rural communities can also access Internet as mobile signal coverage is being improved courtesy of appropriate ICT infrastructure being erected even in rural areas as guided by the rural telecommunications initiative. This initiative has brought modern telecommunications, including Internet access, for the first time to 147 villages. When fully implemented, the project will ensure that more than 50% of Batswana living in the remote areas of the country will be provided with basic telecommunications services. Further the Botswana Telecommunications Corporation (BTC) has launched VSAT technology that, it hopes, will play a role in bringing services to remote areas through the use of satellite and overcome limitations placed on traditional services by vast distances and difficult terrain. It is anticipated that Botswana’s telecommunication industry is likely to record growth in the year 2009 placing it ahead of the telecoms market of South Africa and Nigeria in terms of teledensity. This forecast follows the launch of BeMobile by the BTC in 2008. A lot is being done in the telecoms sector to encourage the growth of this sector. Lewin et al. (2004) has noted that a Convergence Bill is under preparation to update the 1996 Telecommunications Act. Debate has focused on how far this Bill can include provisions that would extend liberalization. Another intiative has been the putting in place of the Cybercrime Bill, originally published in the Government Gazette in October 2007 (Mesa, 2007). The Convergence Bill encourages further liberalizations and convergence of both wireless and fixed technologies. Pheko (2009) notes that the effect of this has been the following: a) licensing of beMobile (the BTC mobile arm), b) rollout of product packages and price offerings to attract

Convergence of Wireless Technologies in Consolidating E-Government Applications in Sub-Saharan Africa

more customers by Orange Botswana and Mascom Wireless, c) establishment of own International Voice Gateways by Mascom Wireless and Orange Botswana, d) introduction of VoIP services by PTOs and VANS; e) upgrading of Public Land Mobile networks to introduce 3G services, f) rollout of WiMax services to some urban areas and g) increase in the number of registered VANS. In order to promote adoption and acceptance of wireless technologies in the pursuit of accessing public services and exchange of information, the government and different other stakeholders such as non-governmental-organizations have put in place the following interventions: a) The Botswana Telecommunication Authority (BTA) has been formed with the mandate to coordinate the ICT and telecommunications landscape in Botswana. This includes allowing mobile Internet service providers to compete fairly with other competitors, put in place legal and institutional frameworks, and so forth. b) About 4% of the country’s GDP has been directly allocated to interventions to encourage use of ICT (Mutula et al., 2010). These interventions include, among others. Carrying out massive awareness campaigns to the general citizenry on the benefits of taking full advantage of ICT platforms to access government services and participate in decision-making. Out of these awareness campaigns, many people now understand the importance of computers in their lives. c) Subsidizing computer and mobile gadgets importation into Botswana and reduction in duty paid for computer accessories into Botswana. d) Government’s implementation of Thuto.net project which aims to encourage the penetration of computers and mobile technologies (especially mobile phones) into Botswana’s schools, universities and companies. Out of this initiative, all the universities and schools

have computers and computer awareness courses have been introduced in all the high countries throughout the country. e) The BTA is also mandated to protect the consumer from exposure to harmful digital content by making sure that acceptable security is implemented to digital resources by the producers. f) The government further sponsors radio talks on how public services can be accessed through ICTs (e-Government). g) Establishment of an e/m-Government task force that has been mandated to make sure that the implementation of the e-Government agenda is realized. This e-Government task force is house within the Ministry of Transport and Communication under the Department of Information Technology. The task force has so far come up with a draft e-Government strategy for Botswana, and is currently working hand-in-hand with BTA to make sure m-Government applications are made a reality. To this end, a government portal has been designed and is in its pilot phase awaiting commissioning after considerable feedback from the users has been obtained. Feedback from the users of the limited mGovernment initiatives that are in place suggest that although many people are not directly aware of e-Government interventions and benefits, there are positive signs that m-Government will thrive in Botswana. Given these scenarios, it is fair to state that there is a lot of potential of m-Government in Botswana (Mutula et al., 2010) as more and more people are having access to mobile phones which can further be used as platforms for Internet access.

B) South Africa South Africa has done some significant strides in taking advantage of what convergence of wireless

57

Convergence of Wireless Technologies in Consolidating E-Government Applications in Sub-Saharan Africa

technologies has to offer by putting in place a committed team for m-Government implementation (Maumbe & Owei, 2006, 2009). Although there are challenges that have been met, use of the m-Government model in South Africa presents a promising platform for citizen-governmentbusiness interactions. Between 2003 and 2008, there has been over 40% increase in the number of mobile subscribers per 100 population in South Africa (Du Preez, 2009). The phenomenal growth in mobile subscribers has been coupled by a rapid increase in commercial and other e-applications that take advantage of the ubiquitous nature of mobile phones. With its relatively well developed and diverse infrastructure, South Africa is taking a regional lead role in the convergence of ICTs with the media and entertainment sector, promising reductions in telecoms costs and better availability of information and services. Billions of dollars are being invested into IP-based Next Generation Networks that are capable of delivering converged services more efficiently (Research report, 2009). Telecom carriers and ISPs are moving into delivering audio and video content over their networks, while in turn the traditional electronic media carriers are discovering the potential of their infrastructure for telecommunications service delivery. Mobile TV licenses are expected to be issued in 2009. Graham et al. (2003) further stipulates that the South African telecommunications market is poised at a critical point in its development. On the wireless side, South Africa has witnessed a spectacular growth in users, far beyond original expectations. Despite high levels of income inequality, approximately 37 percent of South Africans now have a mobile phone (Du Preez, 2009). Whether through WAP-based applications on mobile phones or via cellular data cards connected to laptop computers, wireless technology will increase Internet usage, and will drive some convergence with the fixed Internet world (Graham et al., 2003). South Africa presents a case

58

where convergence is taking place not only in the wireless medium, but blended convergence between broadcast services and mobile communications. This type of convergence is opening way for the implementation of mobile and digital TVs which will impact the way people consume digital content. South Africa’s public sector institutions largely engage with mobile solutions in an isolated and case-specific manner. There is no comprehensive strategy guiding choices around the use of mobile technology. Trying to take advantage of what the convergence of wireless networks has to offer, South Africa has put in place a well-defined vision for e/m-Government. The m-Government strategy, being driven by the South African Center of Excellence on m-Government, will ensure that the public sector can use its spending power and leverage far more strategically, and hence extend access to services whilst the Centre of Excellence will provide a meeting place where government and the private sector can, on a on-going basis, identify and test m-Government solutions as well as to look at how these solutions can positively contribute to improved operations and public service delivery (Maumbe, 2006; Du Preez, 2009). Maumbe and Owei (2006) analyzed the status of m-Government implementation in South Africa. In their study, they carried out an information audit on what endeavors have been done in the line of m-Government implementation, identified the different factors that affect e-Government implementation and developed a model for m-Government implementation in South Africa (please refer to Maumbe and Owei, 2006). Another study by Du Preez (2009) assessed the m-Government readiness in the province of Western Cape. That study found out that countrywide, there are more than 30 million mobile services subscribers and that Western Cape presents a case where mobile technology is going to be used as a major e-Government platform in future. Many of the participants in that survey gave positive experiences with regards to mobile technolo-

Convergence of Wireless Technologies in Consolidating E-Government Applications in Sub-Saharan Africa

gies and m-Government. The major limitation identified in both studies by Maumbe /Owei and Du Preez point to limitation of public services that can be decoded on mobile platforms. Most public services require a considerable bandwidth to access major applications.

SOLUTIONS AND RECOMMENDATIONS Botswana and South Africa show a case where different stakeholders are committed to advancing the wireless technologies for e-Government applications. This is evidenced by different institutional and regulatory frameworks that have been set up starting from the liberalization of the telecommunications sector. Both countries have put in place an enabling environment for wireless technologies proliferation. However this being the case, there are several challenges that are faced in both cases concerning benefits for the transformation of convergence of wireless technologies into tangible socio-economic values and e-applications such as m-Government. These challenges may be different from context-to-context. The following list shows the general recommendations that should be considered when implementing m-Government applications. Looking at the cases presented above, it is evident that there are still some issues that need to be addressed before m-Government were to be a success or mean anything tangible to the ordinary citizens of both South Africa and Botswana. Whilst major interventions are being implemented in both countries, it is evident that not much is being done in regards to protecting the user in these m-Government environments. a) Encourage and nurture m-Government applications by further encouraging awareness campaigns on mass scales to educate people on the benefits of engaging in m-Government applications. These campaigns should be

directed at highlighting both the pros and cons of engaging in m-government. The strategy taken by Botswana in strengthening radio e-Government campaigns should be encouraged. In fact, these campaigns can even be replicated to being promoted using simple text messages pointing citizens to where e-Government services can be found on the cyber space. b) Engage more of Wireless Application Service Providers (WASPs) in pushing the convergence of wireless technologies agenda. This is because some government entities may not have the technical staff and capacity to support the required programming and deployment to link applications to multiple types of wireless devices. For both South Africa and Botswana, this can be done through the regulatory bodies. The regulatory body in respective countries can engage the telecommunications service providers to provide their services through common wireless platforms and technologies. Further, these bodies can also help in encouraging the understanding and adoption of m-Government by ordinary citizens. c) There should be efficient liberalization platforms that encourage competition in the telecoms sector. The more service providers, the more competition and the better the services provided become. This may have a direct impact on m-Government. These liberalization plans should also be spearheaded by the regulatory bodies (BTA for the case of Botswana). d) The future importance of m-Government in the service delivery agenda will be dependent on resolving a range of range of technical and non-technical issues and challenges which are context-specific as identified in this chapter. This should be done through closer cooperation between industry and government, and not forgetting ordinary citizens. It is to be mentioned that a consul-

59

Convergence of Wireless Technologies in Consolidating E-Government Applications in Sub-Saharan Africa

tative process in drawing any strategy for m-Government is desired. Both South Africa and Botswana have put in place teams driving the e/m-Government agenda and drawing up / refining existing e/m-Government roadmap. It is desired that an e-Government agenda done in a participatory manner is appropriate because it can install a sense of ownership into the people and thus they can easily buy in into the technology. e) It is not only with putting in place institutional and regulatory frameworks that can make m-Government to thrive. Developing a better understanding of the legal and institutional issues that facilitate m-Government deployment and addressing weaknesses that may exist is also equally important. Such kind of initiatives should be embedded into e-Government strategy. f) There is a need to introduce an appropriate partnership between network operators, citizens, public sector institutions, and application developers (locally bred) that would allow for the identification of possibilities of the technology for government where this is cost effective and efficient. Once these set of recommendations are considered, coupled with other technical and managerial issues that may affect a local context have been taken care of, it is anticipated that the likelihood that the convergence of wireless technologies will bring forth m-Government success is high.

FUTURE RESEARCH DIRECTIONS The convergence of wireless technologies to presenting a better platform for m-Government summons a number of issues and challenges if full value of these technologies were to be retrieved. Wireless technologies are still under an active developmental phase so technical issues such as security and data rates are still in their infancy

60

of investigations. Most commonly used security platforms such as the wireless intrusion prevention system (WIPS) which monitors the radio spectrum for the presence of unauthorized access points (intrusion detection) and can automatically take countermeasures (intrusion prevention) may have several limitation concerns. Also, devices such as WIPS may not be cheaply accessible to ensure secure application in wireless technologies which would further install confidence on the part of m-Government participants. Further research directions should concentrate on finding cheap, yet appropriate security platforms that can be used in wireless technologies especially as these technologies converge. Also, it is imperative that future research directions be devoted towards finding optimal solutions to the following problems: a) The lack of technical professionals to develop applications, services and contents to increase the ROI for operators with great value solutions adapted to the need of population in the different countries. b) The computer penetration in home and companies and the need of develop an Internet culture to prepare the future citizens which demands new technologies like basic tools for the Society. This entails finding appropriate technology adoption models commensurate to the African technological setup. Other research directions include finding a synergy on how different evolving technologies can be made to perform common e-applications in the name of convergence. These may include further investigating both technical and managerial issues from an African perspective of convergence between broadcasting and telecommunications, fixed and wireless networks/services and voice and data networks. Further work is also needed to accommodate the different stakeholders that utilize e/m-Government applications in the same harmonized e-applications environment.

Convergence of Wireless Technologies in Consolidating E-Government Applications in Sub-Saharan Africa

CONCLUSION Evolving technologies and the convergence of different wireless technologies present a chance that ICT can be used as a public good in facilitating e-applications such as e-Government. Wireless technologies have provided a platform which offers mobility to different ICT devices. As these devices’ capacity to access the Internet is enhanced, and as this presents one of the fastest growing technologies adopted in Africa, the mGovernment model has been made possible. The opportunity provided by m-Government comes with many issues and challenges according to the context where this e-Government model is being implemented. This chapter has looked at what convergence of wireless technologies is and what it entails. Botswana and South Africa have been presented as case studies to ascertain the status of encapsulation of wireless technologies into the m-Government model. The chapter has shown that convergence of wireless technologies is occurring at different levels within the vertical value stack, from networking technology, services and applications to end-user mobile terminals. The chapter has also outlined the different interventions and actions that are necessary for effective convergence to take place and which should be focused on if different stakeholders such as the industry players and regulators are to benefit from the full range of opportunities that are and will be presented by the wireless industry now and in the future. The general consensus of this chapter is that since mobile and wireless technologies have shown great signs of growth in Africa, it is vital that the benefits (such as enhancing e-applications on heterogeneous mobile ICT devices) that this has to offer are taken advantage of. For this to be realized, the different challenges that come with convergence of different wireless technologies, especially for the case of Africa, have to be taken care of. M-Government (which is a replica of eGovernment but with the case where the ICTs used

are not static and where e-Government is mostly accessed using mobile technologies allowing a great deal of ubiquity and pervasiveness) has shown a lot of potential in Africa given the higher costs in acquiring standard computers and higher costs for Internet access which are prerequisites for e-Government facilitation. M-Government, however, offers a flexible and affordable platform as people can access the Internet for interaction with different m-Government stakeholders anytime and anywhere. For the case of typical e-Government, Internet access is only possible on static computers and the Internet provided by ISPs may be expensive than one provided by telecom companies on mobile devices. Now that m-government is inevitable, extending activities to wireless devices and networks will enable SSA countries to be more proactive in their operations and services by providing realtime and up-to-date information to the officials on the move and by offering citizens a broader selection of choices of interactions.

REFERENCES Al-adawi, Z., Yousafzai, S., & Pallister, J. (2005). Conceptual Model of Citizen Adoption of EGovernment. Paper presented at the Second International Conference on Innovations in Information Technology. Al-Sherbaz, A., Jassim, S., & Adams, C. (2008). Convergence in wireless transmission technology promises best of both worlds. SPIE Optoelectronics & Optical Communications. Amailef, K., & Jie, Lu. (2008). m-Government: A framework of mobile-based emergency response systems. Paper presented at the ISKE 2008 3rd International Conference on Intelligent System and Knowledge Engineering. 1, 1398 - 1403. Retrieved March 16, 2009, from http://ieeexplore. ieee.org/stamp/stamp.jsp?arnumber=4731151&is number=4730884

61

Convergence of Wireless Technologies in Consolidating E-Government Applications in Sub-Saharan Africa

Amine, B., & Yosra, K. (2005). M-Government: The Convergence between E-Government and Wireless Technology. Paper presented at the 2005 International E-Business Conference, Hammamet, Tunisia, June. 23-25.

Fricke, M. (2003). Convergence within a Developing Economic Environment: a Strategic Perspective of Interoperability in the SADC. Retrieved May 15, 2009, from home.intekom.com/satnac/ proceedings/.../580%20-%20Fricke.pdf

Amitava, M., & Agnimitra, B. (2005). Simple implementation framework for m-government services, Mobile Business. Paper presented at the ICMB ’05 International Conference on Mobile Business. 1, 288-293. Retrieved September 17, 2005, from http://ieeexplore.ieee.org/stamp/ stamp.jsp?arnumber=1493622&isnumber=32116

Ghetie, J. (2008). Fixed-Mobile Wireless Networks Convergence: Technologies, Solutions, and Services. Cambridge, UK: Cambridge University Press. doi:10.1017/CBO9780511536748

Boyinbode, O.K., & Akinyede R.O. (2008). Mobile learning: An application of mobile and wireless technologies in Nigerian learning system. IJCSNS International Journal of Computer Science and Network Security, 8(11). Bwalya Kelvin Joseph. (2009). “Factors Affecting Adoption of e-Government in Zambia”. EJISDC (2009) 38, 4, 1-13. Retrieved December 15, 2009, from http://www.ejisdc.org/ojs2/index.php/ejisdc/ article/view/573/286 Colesca, S. (2009). Increasing E-trust: A solution to minimize risk in e-Government adoption. [JAQM]. Journal of Applied Quantitative Methods, 4(1), 31–44. Fan, J., & Zhang, P. (2007) A Conceptual Model for G2G Information Sharing in E-Government Environment. Paper presented at the 2007 Sixth Wuhan International Conference on E-Business - e-Business Track. 199-204. Conference proceedings. Fasanghari, M., & Samimi, H. (2009). A Novel Framework for M-Government Implementation. Paper presented at the 2009 International Conference on Future Computer and Communication. 627-631. Retrieved April 30, 2009 from http:// ieeexplore.ieee.org/stamp/stamp.jsp?arnumber= 5189859&isnumber=5189730.

62

Gillwald, A., & Stork, C. (2008). Towards Evidence-based ICT Policy and Regulation: ICT access and usage in Africa. 1(2). Retrieved June 9, 2009, from www.researchictafrica.net/.../riapolicy-paper_ict-access-and-usage-2008.pdf Graham, F., Lewis, C., Lonergan, D., Mendler, C., & Northfield, D. (2008). South African Communications, 2002-2008: Market Review and Analysis. South African Communications Market Study. South Africa: Prepared for the Department of Communications. Heeks, R., & Santos, R. (2009). Understanding Adoption of e-Government: Principals, Agents and Institutional Dualism. iGovernment Working Paper Series. Retrieved October 20, 2009 from http://www.sed.manchester.ac.uk/idpm/research/ publications/wp/igovernment/index.htm Javaid, U., Meddour, D. E., Rasheed, T. M., & Ahmed, T. (2006). Cooperative wireless access networks convergence using ad-hoc connectivity: opportunities and issues. In wireless world research forum No. 16, Shanghai China. Du Preez, J. (2009). Assessing the m-Government Readiness within the provincial government western cape. University of Stellenbosch, school of public management and planning. Unpublished Masters’ Degree thesis. Kamal, M. M., & Themistocleous, M. (2006). A Conceptual Model for EAI Adoption in an eGovernment Environment. Paper presented at the 2006 European and Mediterranean Conference on Information Systems (EMCIS), Costa Blanca, Alicante, Spain.

Convergence of Wireless Technologies in Consolidating E-Government Applications in Sub-Saharan Africa

Karantjias, A., Papastergiou, S., & Polemi, D. (2007). Innovative, Secure and Interoperable E/MGovernmental Invoicing. Paper presented at the 18th International Symposium on Personal, Indoor and Mobile Radio Communications (PIMRC 2007). IEEE. 1-5, 3-7. Retrieved September 13, 2007 from http://ieeexplore.ieee.org/stamp/stamp. jsp?arnumber=4394010&isnumber=4393983 Kushchu, I., & Kuscu, M. (2003). M-Government: facing the inevitable. Paper presented at the 3rd European Conference on e-Government, Frank Bannister and Dan Remenyi (eds.), Trinity College, Dublin Ireland. Lanwin, B. (2002). A Project of Info Dev and The Center for Democracy & Technology. The E-government handbook for developing countries. Retrieved November 15, 2009, from http://www. cdt.org/egov/handbook/2002-11-14egovhandbook.pdf Lewin, D., Hall, R., & Milne, C. (2004). Further liberalization of Botswana’s telecommunications industry. Consultation document for interested parties issued by the Botswana Telecommunications Authority. Retrieved August 8, 2009, from www. bta.org.bw/pubs/Liberalisation%20Consult2.pdf Maumbe, B. M., & Owei, V. (2006). Bringing M-government to South African Citizens: Policy Framework. Cape Town: Delivery Challenges and Opportunities. MESA IT&T Newsleter, fourth quarter. (2007). Retrieved August 17, 2009, from http://www. linexlegal.com/content.php? Monnane, M. (2003). Botswana Telecommunications Services. Southern African Trade Research Network (SATRN). Working Paper No. 3 Mutula, S., Grand, B., Zulu, S., & Sebina, P. (2010). Towards an information society in Botswana: ICT4D Country Report. Research paper presented at Thetha Sangonet ICT4D Forum.4

Odinma, A. C., Oborkhale, L. I., & Kah, M. O. (2007). The Trends in Broadband Wireless Networks Technologies. The Pacific Journal of Science and Technology, 8(1). Retrieved June 25, 2009, from http://www.akamaiuniversity.us/PJST.htm. Patel, I., & White, G. (2005). M-government: South African Approaches and Experiences. Paper presented at EURO mGOV 2005, Brighton, UK, 313-323. Pheko, T. G. (2009). Botswana Telecommunications Authority. Botswana Country Report to CRASA. Retrieved August 15, 2009, from www. crasa.org/.../reports/.../Botswana%20Report%20 2009%20Presentation.pdf ITU Report. (2008). Report on the World Summit on the Information society stocktaking. Research Report. (2009). South Africa – Convergence – VoIP, NGN & Digital Media. Retrieved June 22, 2009, from http://www.researchandmarkets. com/reports/1031255/ Rossel, P., Finger, M., & Misuraca, G. (2006). Mobile” e-Government Options: Between Technologydriven and Usercentric. The Electronic [from www. ejeg.com]. Journal of E-Government, 4(2), 79–86. Retrieved September 20, 2009. Titah, R., & Barki, H. (2006). E-Government adoption and acceptance: A literature review. International Journal of Electronic Government Research, 2(3), 23–57. Touré, H. I. (2006). Competitiveness and Information and Communication Technologies (ICTs) in Africa International Telecommunication Union. Retrieved July 12, 2009, from www.weforum.org/ pdf/gcr/africa/1.5.pdf TRASA report. (2009). Report on the Next Generation Wireless Technologies Conference for Southern Africa. Retrieved August 4, 2009, from www.crasa.org/.../Next%20Generation%20Conference%20Report.pdf

63

Convergence of Wireless Technologies in Consolidating E-Government Applications in Sub-Saharan Africa

Wangpipatwong, S., Chutimaskul, W., & Papasratorn, B. (2008). Understanding Citizen’s Continuance Intention to Use e-Government Website: a Composite View of Technology Acceptance Model and Computer Self-Efficacy” The Electronic [from www.ejeg.com]. Journal of E-Government, 6(1), 55–64. Retrieved September 7, 2009. Warkentin, M., Gefen, D., Pavlou, P., & Rose, G. (2002). Encouraging citizen adoption of E-Government by building trust. Electronic Markets, 12(3), 157–162. doi:10.1080/101967802320245929

KEY TERMS AND DEFINITIONS Wireless Technologies: Technologies that enable communication to be facilitated through wave propagation with no use of a wired medium. The wave propagation may happen between a stationery base station or access node and a mobile platform such as a PDA. Convergence: Is the approach toward a definite value, a definite point, a common view or opinion, or toward a fixed or equilibrium state. Convergence of wireless technologies means different wireless technologies such as WiFi and WIMAX can decode encoded signals from the same base station either WiFi or WIMAX technology.

64

Electronic Government (e-Government): Is a platform through which the government interacts with its citizens and business entities for the sake of exchange of information, public services and participatory democracy through the use of ICT platforms. Mobile Government (m-Government): This is replica version of e-Government, only that in this mode, mobile ICT devices are used for interactions and access to the Internet may be facilitated anywhere and anytime (ubiquitous nature of e-Government). ICT: This acronym stands for information and communications infrastructure. Sub-Saharan Africa (SSA): Is a geographical term used to describe the area of the African continent which lies south of the Sahara or those African countries which are fully or partially located south of the Sahara. This excludes North African countries because they are considered as belonging to the Arab world. ICT Infrastructure: It encompasses all the devices, networks, protocols and procedures that are employed in the telecoms or information technology fields to foster interaction amongst different stakeholders. Regulatory Framework: A set of guidelines that coordinates the following of set principles, e.g. market competition and liberalization, ethic, etc.

65

Chapter 5

Process Innovation with Ambient Intelligence (AmI) Technologies in Manufacturing SMEs:

Absorptive Capacity Limitations Kathryn J. Hayes University of Western Sydney, Australia Ross Chapman Deakin University Melbourne, Australia

ABSTRACT This chapter considers the potential for absorptive capacity limitations to prevent SME manufacturers benefiting from the implementation of Ambient Intelligence (AmI) technologies. The chapter also examines the role of intermediary organisations in alleviating these absorptive capacity constraints. In order to understand the context of the research, a review of the role of SMEs in the Australian manufacturing industry, plus the impacts of government innovation policy and absorptive capacity constraints in SMEs in Australia is provided. Advances in the development of ICT industry standards, and the proliferation of software and support for the Windows/Intel platform have brought technology to SMEs without the need for bespoke development. The results from the joint European and Australian AmI-4-SME projects suggest that SMEs can successfully use “external research sub-units” in the form of industry networks, research organisations and technology providers to offset internal absorptive capacity limitations.

INTRODUCTION Through case study research, this chapter discusses some of the challenges Small and Medium Enterprises (SMEs) in the manufacturing sector face DOI: 10.4018/978-1-60960-042-6.ch005

in identifying and adopting Ambient Intelligence (AmI) technologies to improve their operations. Ambient Intelligence technologies are also known as Pervasive computing or Ubiquitous computing, and we include the descriptions of these terms when we refer to AmI technologies. Our study includes case studies of three Australian

Copyright © 2011, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited.

Process Innovation with Ambient Intelligence (AmI) Technologies in Manufacturing SMEs

SMEs and a comparison with similar application requirements in a German SME manufacturer. The outcomes of the study are likely to be applicable to small firms in many nations. The 1980s and 90s saw the operations of many large manufacturers revolutionized by the introduction of process and technological innovations (Gunasekaran & Yusuf, 2002). While there have been uneven adoption rates in smaller businesses and across different nations (Chong & Pervan, 2007; Oyelaran-Oyeyinka & Lal, 2006) it is clear that technological innovations such as Electronic Data Interchange, Business Process Re-engineering, Enterprise Resource Planning and robotic automation, amongst others, have played key roles in increasing manufacturing productivity. At the beginning of the twenty first century this transformation continues. Ambient Intelligence (AmI) technologies are being positioned as the next performance and productivity enhancing purchase for manufacturers, and a potential means for manufacturers in developed nations to counter perceived threats from lower labour cost countries (Kuehnle, 2007). Thus, the key objectives of this chapter are to consider potential applications of AmI technologies in Australian SME manufacturers, and discuss the opportunities and shared challenges faced by such firms in adopting these technologies. In doing this we will compare different levels of absorptive capacity and technological readiness in Australian firms, seeking possible reasons for similarities and differences in their comparative technology adoption processes. The chapter also examines the role of intermediary organisations in alleviating these absorptive capacity constraints. Our overarching research question is: “Can external intermediaries overcome absorptive capacity limitations in SMEs seeking process innovation through the application of AmI technologies?” In order to understand the issues surrounding this problem, a brief overview of ICT (Information and Communication Technologies) adoption in

66

manufacturing and an explanation of Ambient Intelligence (AmI) technologies are provided in the following section. Following that we examine the role of SMEs in the Australian manufacturing industry plus the impacts of government innovation policy and absorptive capacity constraints in SMEs in Australia.

BACKGROUND ICT Adoption for Business Performance Improvement Brown and Bessant (2003) described the global manufacturing environment developing in this new century as an increasingly competitive landscape, characterised by on-going demands for improved flexibility, delivery speed and innovation. A frequently occurring element in manufacturers’ responses to these pressures is the implementation of increasingly sophisticated ICTs. The benefits of incorporating ICTs on business responsiveness have been identified as: more effective and more efficient information flows; assisting in value-adding improvements for current processes; greater access to efficiency enhancing innovations throughout the value chain (Australian Productivity Commission, 2004); and the ability to access world markets through e-commerce (Kinder, 2002). ICT adoption has been considered worth the risk, given the competitive pressures placed on business to keep pace with technology. For example, in Australia, the uptake of ICTs has increased dramatically towards the later part of the 90’s and into the 21st Century. Reports show that in 199394, 50 per cent of firms used computers with 30 per cent having internet access; by 2000-01 these figures had increased to 85 per cent and 70 per cent respectively (Australian Productivity Commission, 2004). Recent figures (Australian Bureau of Statistics, 2009) reveal that almost all Austra-

Process Innovation with Ambient Intelligence (AmI) Technologies in Manufacturing SMEs

lian SMEs use ICTs, and 96% of them access the internet through a broadband connection. One of the latest developments in the application of ICTs to business improvement is that of Ambient Intelligence (AmI) technologies. The objective of AmI is to broaden and improve the interaction between human beings and digital technology through the use of ubiquitous computing devices. By using a wider range of simplified interaction devices, ensuring more effective communication between devices (particularly via wireless networks) and embedding technology into the work environment, AmI provides increased efficiency, more intuitive interaction with technology (Campos, Pina, & Neves-Silva, 2006) and improved value and productivity (Maurtua, Perez, Susperregi, Tubio, & Ibarguren, 2006).

Ambient Intelligence (AmI) Technologies Existing literature (Kopacsi, Kovacs, Anufriev, & Michelini, 2007; Li, Feng, Zhou, & Shi, 2009; Maurtua et al., 2006; Vasilakos, 2008; Weber, 2003) points to the co-existence of three features in any AmI technology: ubiquitous computing power, ubiquitous communication and adaptive, human-centric interfaces. Regardless of arguments about terminology and definitions (the terms “pervasive computing” and “ubiquitous computing” are in common use in the US, while “ambient intelligence” is favoured in the EU), these technologies are already commonplace. The beep signalling the automatic deduction of a road toll from your account as your car passes under a toll gate is one aspect of an AmI technology known as Radio-Frequency Identification (RFID). RFID technology is having an impact in many industries, some of which are not normally associated with high levels of ICT adoption. For example, during 2006, in NSW alone (one of the 7 states and territories within mainland Australia), more than 1.2 million head of cattle were automatically

tracked from farm to saleyard to abattoir as their RFID ear tags passed through RFID sensor gates (NSW Farmers Association, 2007). In addition to increasing process speed and efficiency, AmI technologies have the potential to provide tracking of employee and customer activity. While concerns about the impact of technology upon power relations in the workplace are not new (Zuboff, 1988), the characteristics of AmI technologies present new challenges to worker privacy, informed consent and dignity. AmI technologies may, intentionally or not, dramatically increase employee surveillance and monitor consumer activity over the entire product life cycle. This potential and the very nature of ‘ubiquitous computing’ raises important ethical issues. Proposals to use RFID tags to track sufferers of Alzheimer’s disease (Caprio, 2005) and children provide examples of the ethical dilemmas AmI technologies can present. While these issues are beyond the scope of this chapter, we suggest Cochran et al (2007) for a review of ethical challenges associated with RFID. Social factors associated with the introduction and implementation of AmI technologies may be exacerbated in small and medium businesses. In addition to concerns shared with corporate workers, such as disquiet about their personal data potentially being sold to marketing groups and anxiety about the security of the information gathered, members of small businesses are particularly prone to AmI’s ability to ‘break the boundaries of work and home through their pervasiveness and ‘always on’ nature’ (Ellis, 2004, p.8). While some profit-maximising small business owners welcomed this blurring of work and home boundaries, others did not, preferring to keep work and family spheres separate. Ellis cautions that in order to overcome existing negative preconceptions of AmI technologies, users must feel they control the devices and the data they produce, and be able to override them and cope with systems failure. In particular, AmI needs to be presented

67

Process Innovation with Ambient Intelligence (AmI) Technologies in Manufacturing SMEs

as “smarter” than existing ICTs and able to correct some of the problems associated with traditional forms of ICT support. In short, Ellis (2004, p. 9) asserts, “AmI needs to be associated with undoing some of the more problematic aspects of existing ICTs, to be accepted and not resisted as a more invasive, insidious and controlling form of what already exists.” Much of the promise of Ambient Intelligence (AmI) technologies rests upon connecting increasingly sophisticated and powerful sensors with existing computing facilities. McCullough (2001) identified the need to expand our thinking beyond the notion of filling environments with physical objects when considering Ambient Intelligence technologies. There is no longer such a thing as “empty space” when sensors and processing power combine to produce an environment that is “aware” of the locations, actions and information needs of humans. Clearly, the extent of existing Information and Communications Technology (ICT) infrastructure in an organisation will impact AmI technology implementations, providing either a “clean slate” from which to start or the opportunity to integrate new AmI capabilities with existing ICT systems and processes. In much the same way as the advent of mobile phones in China and India provided an opportunity for people unable to afford a landline, to access telephonic services, AmI technologies may prove a way for SME organisations to “leap frog” a stage of ICT implementation, and move directly to wireless and similar AmI technologies. Many other applications of AmI technologies are appearing as technologists extend the concept into areas such as “wearable technology” (clothing that incorporates sensors and interface devices), more intuitive home space designs, shopping assistance and the creation of seamless interfaces between work, home and leisure activities. While many of these applications currently seem unrelated to improving business productivity, it is clear that the applications for business can only grow as the technologies become more sophisticated and

68

less expensive. As Rao and Zimmerman (2005, p.3) state “there is a gap in the scholarly discussion addressing the business issues related to it, and the role of pervasive computing in driving business innovation”. It is in this context that the following case studies of four small-to-medium (SME) manufacturers - three Australian and one German firm – have been undertaken. In each firm, critical process analysis was carried out to examine possible process weaknesses and existing ICT systems, and recommendations were made concerning a selection of AmI technologies with the potential to boost business performance.

AMBIENT INTELLIGENCE TECHNOLOGY IN MANUFACTURING This section considers the applicability of several emerging AmI technologies to three SME manufacturers in New South Wales, Australia and compares the situation within these SMEs with one German SME manufacturer undertaking a similar technological adoption. In doing this the section also addresses questions about the preparedness of SMEs, particularly concerning their absorptive capacity limitations and how these may be overcome. Later sections also consider the potential impact of Ambient Technologies on the employees of the organisations studied. AmI technology is much more than RFID inventory control systems. Wireless, multi-modal services and speech recognition systems have the potential to increase manufacturing flexibility by supporting dynamic reconfiguration of process and assembly lines, and improving human-machine interfaces to reduce process times (Maurtua et al., 2006). Also, maintenance and distribution processes may be improved by linking common mobile wireless devices, such as mobile phones, Personal Digital Assistants (PDAs) or even pagers to production alert systems (Stokic, Kirchhoff, & Sundmaeker, 2006).

Process Innovation with Ambient Intelligence (AmI) Technologies in Manufacturing SMEs

Small and Medium Manufacturers in Australia Organisations with between 20 and 199 workers employ 56% of Australia’s workforce (Wiesner, McDonald, & Banham, 2007). The Australian Bureau of Statistics (ABS) defines a small business as employing less than 20 people, and a medium enterprise as employing between 20 and 200 employees (ABS, 2001). The most recent ABS figures available (2007) for Australia indicate that there are around 47,000 manufacturing firms employing between 1 and 20 people, around 10,000 employing between 20 and 200 people, and only 873 employing over 200 people. In turnover terms, around 29,000 manufacturing firms reported annual turnover between $500,000 and $10 million, while only 3,300 firms reported turnover of $10 million or above. It is clear that the bulk of manufacturing in Australia occurs in small-to medium firms. While SME firms employ the majority of manufacturing workers their expenditure on R&D notably lags behind that of large manufacturers. Within the manufacturing industry companies with more than 200 employees were responsible for 73% of total industry R&D expenditure, with only 27% being contributed by the SME sector (ABS, 2007). However, in their exploration of the cost and impact of logistics technologies in US manufacturing firms Germain, Droge & Daugherty (1994) found that for manufacturing managers wanting to innovate with logistics technology, organisational size provides an advantage that transcends both the cost and nature of the technology. These authors confirmed the established view in 1994; that organisational size was positively correlated with technology adoption, as found in many previous studies. This link between manufacturing organisation size and increased ability to extract benefit from technological innovations may provide some explanation for the fact that while Australia’s manufacturing output has quadrupled since the mid 1950s, the Australian Government

Productivity Commission (2003) states that overall, it has not grown at the same rate as the service sector. The Productivity Commission also describes Australia’s manufacturing sector as having “missed out on the productivity surge” of the mid 1990s while noting signs of improved manufacturing productivity in 2002 and 2003. The widespread availability of off-the-shelf ICT systems has probably meant that a great many more SMEs are adopting ICTs today than in 1994, however the limited resources of many such firms (both financial and human) almost certainly mean reduced awareness and limited capacity to exploit newer technologies commonly appearing in larger manufacturers. Given the significance of SMEs in Australian employment and the perceived need to increase manufacturing productivity, examination of the potential improvements available through the systemic application of AmI technology to SME manufacturers forms an important topic for research and government policy.

Absorptive Capacity This chapter applies the concept of absorptive capacity to manufacturing SMEs. We argue that SMEs can benefit from AmI technologies, using specialised intermediary organisations to overcome the “absorptive capacity” limitations evident in many SME organisations. Cohen and Levinthal (1990) proposed that internal Research & Development activities serve two purposes: to generate innovations, and to provide the ability to absorb relevant knowledge appearing in the external environment. The absorptive capacity of a firm is comprised of these two categories of activity. Their foundational paper conceptualised absorptive capacity in the context of large U.S. manufacturers, as evidenced by their survey of identifiable “R&D lab managers” (Cohen & Levinthal, 1990, p. 142) and their discussion of “communication inefficiencies” between business units. But what of small- and medium-sized manufacturers? Does the notion of absorptive

69

Process Innovation with Ambient Intelligence (AmI) Technologies in Manufacturing SMEs

capacity have relevance to SMEs outside narrow, industry-segment specific technologies? If so, can external intermediaries assist SMEs to overcome absorptive capacity limitations regarding ambient technological innovations? The preparedness of SMEs to adopt and benefit from the mobile capabilities of AmI technologies is likely to be linked to their ability to overcome the limited absorptive capabilities associated with their small size. However, it is possible for SMEs to successfully deploy AmI technologies. Clear evidence of the benefits produced by a range of AmI technologies have been observed by the researchers at Costa Logistics’ distribution centre in Western Sydney. Costa Logistics specializes in the distribution of fresh fruit and vegetables, and achieves astounding levels of accuracy and throughput: staff average zero to three errors per million cartons picked in the warehouse, the daily inventory turn ratio is 100% and over A$2 billion of product is handled annually on behalf of their clients (Game-Lopata, 2008). The company was a SME when it started to use wrist-mounted bar-code scanners coupled with wireless communications. These successfully implemented AmI technologies are key factor that have enabled Costa Logistics rapid expansion to well over 200 employees in late 2009.

Innovation, Manufacturers, SMEs and Government Policy Several previous studies (Cutler, 2008; Philips, 1997) have concluded that innovative Australian firms of all sizes (both manufacturing and servicebased) tend to be more successful in terms of sales growth and market share than non-innovating firms. In addition, the impact of innovation is considered to be cumulative (Chapman, Toner, Sloan, Caddy, & Turpin, 2008) with some level of innovative behaviour or research and development being required to equip a firm to identify, assess and adopt technologies. The innovativeness and absorptive capacity of SMEs is a matter of concern for other nations besides Australia. For example,

70

in its 2008 budget, the UK government signalled its intention to set a goal for innovative SMEs to win 30% of its ₤150 billion public procurement spending (Kable’s Government Computing, 2008), equating to $98 billion (AUD) of incentives to encourage UK SMEs to innovate. While similar incentives are yet to appear in Australia, there are clear signs of government interest in the ability of SMEs to innovate (Department of Innovation Industry Science and Research, 2008). There is a growing body of work in the innovation literature on the limited absorptive capacity of SMEs to identify relevant innovations, understand and appreciate possible applications, and finally adapt and implement innovation in their organisations (Beckett, 2008; Liao, Welsch, & Stoica, 2003; Muscio, 2008). Many points concerning “constraining factors” and “implementation challenges” support the notion that SMEs can experience organisational absorptive capacity limitations. Beckett (2008) identifies knowledge and resource constraints that impede the ability of SMEs to develop absorptive capacity, but also provides an example of how absorptive capacity is built when the outlays of time and money required match the SME’s available resources. While the benefits of AmI technologies are already accruing in large organisations (Angeles, 2005) if manufacturing SMEs are to benefit from AmI technologies, one challenge requiring attention will be that of their limited absorptive capacity for technological innovations. Our research considered the possibility of external intermediaries being used to facilitate SME manufacturers’ assessment of the application of AmI technologies for process innovation, thus overcoming, at least partially, the problems of limited absorptive capacity within the partner SMEs.

Process Innovation with Ambient Intelligence (AmI) Technologies in Manufacturing SMEs

Figure 1. Representation of the Three Phase AmI-4-SME Methodology (Source (Kirchhoff et al., 2006))

METHODOLOGY AND DATA COLLECTION Case research has been used to review and compare the operations of the three NSW manufacturing SMEs, identify process weak points (in partnership with the SME executive managers) and suggest potential Ambient Intelligence Technologies to assist each organisation. The data has been gathered as part of an Australian government funded International Science Linkages Project, which in turn was part of a larger EuropeanAustralian project on Ambient Intelligence in manufacturing - Ambient Intelligence Technology for Systemic Innovation in Manufacturing SMEs (AmI-4-SMEs). The European component of the AmI-4-SMEs project involved six SMEs, three research partners and three Information and Communications Technology (ICT) providers located in Germany, Ireland, Spain and Poland. The Australian AmI-4-SMEs project consists of six SMEs, two research partners (University of Melbourne and University of Western Sydney) and two ICT providers. Six SMEs were selected from those responding to a request for expressions of interest in participating in the study. Using the Australian Bureau of Statistics metrics (2001), all are classified as small-to-medium sized manufacturers and all are privately owned.

The EU AmI-4-SMEs project aimed to design and develop a coordinated approach for process innovation using, ICT “building blocks” and a software platform to support the improvement of manufacturing processes in SMEs. These improvements were achieved by re-engineering processes and introducing appropriate ICT tools. The method used to analyse and re-engineer business processes is an extension of the COST-WORTH methodology (ATB Institute for Applied Systems Technology Bremen GmbH, 2004) and has three main phases: Analysis and Conception, Selection and Specification, and finally Implementation. Due to lower levels of government funding support, the Australian Ami-4-SMEs project performed only the Analysis and Conception, and the initial aspects of the Specification and Design phase but not the Implementation phase. The links between these phases are shown in Figure 1. The Analysis and Conception phase produces an implementation plan for the proposed AmI solution. Analyses of each SME’s business processes and bottlenecks form the majority of this phase which concludes with presentation of a business re-engineering recommendation and a Return on Investment Analysis (Kirchhoff, Stokic, & Sundmaeker, 2006). One challenge of working with SMEs is to gather sufficient information without intruding

71

Process Innovation with Ambient Intelligence (AmI) Technologies in Manufacturing SMEs

to the extent that the organisation is adversely affected. On-site interviews and observation, questionnaires (these were developed as a part of the precursor COSTWORTH project, see Nousala, Ifandoudas, Terziovski and Chapman, 2008), video recordings, a visit to a company (Costa Logistics) already using wireless, wearable and voice (Chang & Heng, 2006) technologies in its warehouse, and joint creation of process maps where they did not previously exist were used to collect data and minimise disruption to the SMEs. Interview and questionnaire data were used to select important, problematic processes for each SME. On-site observations and video recordings were analysed to create “as-is” maps of the process selected for improvement, and identify key limitations of each process. AmI technologies with potential to improve the selected business process were selected and the likely costs and benefits reviewed with the SME executive managers. A strength (and simultaneous limitation) of this approach, is that the time spent at each SME site is not extended or intensive. However, given the objectives of the analysis and conception phase of the AmI-4-SMEs project, and the need to minimise disruption to the operation of the manufacturing businesses, the methods are appropriate.

RESULTS

have been used for the SME companies. Ranked in order from low to higher technological capability they were: SwivelMould – a plastics manufacturer specialising in rotational moulding work; BottleTop – manufacturer of plastic packaging closures; and TechMakers – a contract electronics component manufacturer. As described in the methodology section, due to lower funding levels, the Australian AmI-4SMEs project did not fully incorporate the latter two phases of the project (see Figure 1), instead leaving it to the individual companies to complete these steps. The European project funding also included direct funding support for the SME partners, which was not provided under the Australian government support funds. However, the Analysis and Conception phase, and the initial aspects of the Specification & Selection phase, included a detailed initial assessment of each SME, identified possible AmI solutions and produced a rough implementation plan for the proposed AmI solution. Analyses of each SME’s business processes and bottlenecks formed the majority of this phase, and these were conducted on-site at the three NSW SMEs. Tables 1 and 2 present summary information and key issues related to technology adoption for each of the three SMEs from New South Wales, Australia.

Overview of the Three Australian SMEs

Similarities and Differences between the Three Australian SME Manufacturers

The three Sydney-based SMEs represent a wide range of existing ICT complexity and skill, from “craft-work” factories to highly sophisticated manufacturers of technology. Although the organisations were not intentionally recruited to represent a typology of low, moderate and highly integrated ICT installations and skill sets, they do display these characteristics and it is useful to consider how AmI technology capabilities may be added to each of these settings. Pseudonyms

Scalability issues related to specialised equipment appeared as a limitation shared by the three Australian SMEs. Production could not rapidly increase without the purchase of more machinery, and in the case of TechMakers, difficulty in arranging for the rapid importation of components also limited their growth. Key differences observed in the three Australian SMEs relate to the proportion of standardised production and organisational culture. Eighty per

72

Process Innovation with Ambient Intelligence (AmI) Technologies in Manufacturing SMEs

Table 1. Company summary BottleTop Turnover: $30M Employees 97 Profit: moving from break-even to profitable trading in 2008/09. Industry: Manufacturing of packaging closures Outlook: Growing in a shrinking market by taking market share from competitors. BottleTop has set an aggressive revenue growth target of $50M by 2011/12.

TechMakers Turnover: $7 M Employees 70 Profit: Trading profitably Industry: Contract Electronic Manufacturers. Outlook: Industry is shrinking, work is increasingly being sent off shore, TechMakers are diversifying by creating and selling the devices they created to assist in their contract electronic manufacturing activities.

SwivelMould Turnover: $20M Employees: 80 Profit: trading profitably, but recently burdened by a major bad debt. Industry: Rotational Moulding Outlook: Rapid growth during drought through manufacture of water tanks,

Table 2. Key issues related to technology adoption BottleTop

TechMakers

SwivelMould

The company managers view their employees as loyal, and based on a long association, almost like family. Some staff members resent the profits they believe the company owners are making and draw unfavourable comparisons to their hourly wage rate. A profit sharing scheme based upon reducing the scrap rate has been enthusiastically embraced by staff. BottleTop’s labour efficiency has increased from 76 to 92% in the last year primarily attributable to their process focus as they implement Six Sigma Manufacturing techniques

No staff-related comments or concerns appeared in interviews or observations.

SwivelMould have a high employee turnover as the work is repetitive, takes place in a hot and noisy environment and does not pay high hourly rates. SwivelMould had difficulty recruiting and retaining employees while unemployment rates were low. New employees are only given formal training after they have completed a probationary period and if they appear likely to work at SwivelMould for some time.

The Quality Manager describes product tracking as “the black hole” because once a job commences no job status information is available until the end of the manufacturing run. This issue can be addressed by the organisation as TechMakers employ a full time in-house programmer. When time permits the programmer will interface product tracking with the MRP system.

Orders are received by fax and transcribed onto job sheets by hand. As jobs are completed on the shop floor the quantity, colour etc details are confirmed by the foreman writing on the same sheet. Any variations in process are also hand written on the sheet.

TechMakers processes are well developed and highly automated and with one exception, their work in progress system, discussed in the “Tracking of Products” column. A more difficult process to address concerns the performance of the distributor that acts as the exclusive Australian agent for the US manufacturer of components used by TechMakers.

SwivelMould has not mapped its processes. The process map built for one process as part of the AmI-4-SME process was the first time the company had used process mapping. Tradition and the knowledge of gang foremen are used to guide manufacturing processes.

No comments made in relation to training.

SwivelMould has developed a competency based training program in conjunction with an external training consultant. The objectives of the training are to increase productivity, quality and produce a change in the organisational culture.

Staff

Tracking of Products Product tracking is highly automated, integrated links between moulding machines and software packages running on PCs provide a real time view of activity.

Process Improvements BottleTop is using six sigma methods to improve quality, reduce product variability and reduce waste. The company is focussing on improving each process prior to automating it. As the Operations Manager states “We want to have strong processes, we don’t want rubbish processes just being done more quickly.” When quality processes are in place BottleTop’s focus will shift to automation and then to monitoring. Training Courses A mix of in-house and external training is used at BottleTop. External training is used to provide Six Sigma training.

continued on following page 73

Process Innovation with Ambient Intelligence (AmI) Technologies in Manufacturing SMEs

Table 2. continued BottleTop

TechMakers

SwivelMould

BottleTop managers are optimistic about their ability to grow revenue in a shrinking market. However, growing acceptance of closure systems make in countries with low labour rates may limits their growth.

TechMakers is and is the oldest and fourth largest contract electronic manufacturer (CEM) in Australia. The CEO is content to remain the fourth largest firm and claims there are advantages in not being the biggest player in the Australian market. The market for CEMs in Australia is shrinking as work moves off-shore. While the Quality Assurance Manager believes the company is committed to increasing revenue, the CEO (who is the owner of the company) stated privately that his objective is to improve the profitability of operations.

SwivelMould’s ambitious expansion plans are on track to deliver the anticipated growth in revenue. However as rain water tanks provide 5560% of annual revenue SwivelMould’s plans are dependent to some extent upon the maintenance of government rebate policies, and continuation of Australia’s ten year drought. During the course of the study the drought broke on the eastern seaboard, reducing demand for rainwater tanks. While newer entrants to the market have been unable to cope with the downturn in the tank business, SwivelMould’s Managing Director is confident of its ability to continue its expansion due to investments made in new products and its pursuit of new markets.

TechMakers are pursuing export opportunities for the technologies they have invented in house. They do not intend to attempt to compete with CEMs located overseas.

A reduction in demand for water tanks has occurred with the breaking of the drought in two cities (Sydney and Brisbane). To some extent this has been offset by increases in exports; the quantity of goods exported to the USA and China has trebled in the last year.

Highly integrated with and automated by ICT systems.

Rudimentary documentation, handwritten and physically carried between office and factory floor. Little integration with ICT systems.

Growth

Overseas Operations Sixty per cent of the company’s revenue is derived from importation and distribution activities. Logistics is important and problematic for the company as products imported as finished goods create significant supply chain challenges. BottleTop is considering employing a logistics outsourcing firm to manage these challenges Documentation & Processes Highly integrated with and automated by ICT systems.

cent of BottleTop’s product volume results from standing contracts with its top twenty customers. Beyond the top twenty customers is what the Operations Manager refers to as a “long tail” of hundreds of small customers. In contrast, TechMakers survives in a shrinking market due to its flexibility, rapid prototyping turn-around and high quality output. The CEO and Quality Assurance manager concur that “every day is different.” SwivelMould’s operations sit between the other two SMEs. SwivelMould offers a full service from product concept through design to manufacture. As each type of product requires a specialised mould, batch runs are used in production. While safety equipment is in use at TechMakers and BottleTop, SwivelMould employees resist using

74

standard safety equipment such as hard hats and hearing protection. Only foremen wear high visibility vests. The Managing Director describes these actions as symptomatic of the “70s culture” that he is trying to change.

AmI Technology Recommendations The Australian SMEs The wide range of technical skill levels in the three Australian SME manufacturers results in very few similarities in their technological capacities. The staff in two of the SMEs are happy with the opportunity to enhance their skills and comfortable with process change. In all the SMEs the arrival

Process Innovation with Ambient Intelligence (AmI) Technologies in Manufacturing SMEs

of technology that is industry-specific is seen as a form of “natural progression” from previous machines. In one SME (SwivelMould) the CEO is experienced difficulty in moving staff away from established skills and procedures. The recommendations for the SMEs were tailored to the existing environment of each organisation in terms of industry segment and their legacy ICT systems. In the case of SwivelMould, recommendations were made for AmI technologies that either link together existing “islands” of ICT equipment or provide quickly implemented, stand-alone solutions to environmental issues. For example, a recommendation to implement an integrated manufacturing system, taking advantage of the flexibility of wireless communications was made. As Kirchoff, Stockic and Sundmaeker (2006) assert, if insufficient or poorly integrated ICT infrastructure exists the first step towards obtaining the benefits of AmI technologies is to introduce ICT systems to support general manufacturing processes. Given the size of SwivelMould’s operations, it is likely that a standardised manufacturing package working on a PC/Windows, Unix or Linux operating environment will address this need. To provide a small example of the benefits available, using a secure on-line order entry system will avoid the need to manually transfer order details faxed to SwivelMould by their distributors onto worksheets. The second organisation, TechMakers presents a challenge as their business is manufacturing electronic devices and sub-assemblies, effectively acting as an outsourced technology design and manufacture facility for their clients. TechMakers are well aware of AmI technologies and would have already incorporated them into their operations if they had identified potential applications. Furthermore, their key business issues concern their contracting market and their dependency upon an intermediary distributor to order components from the USA. For TechMakers, the benefits available from AmI technology may

exist in opportunities to design, manufacture or modify AmI systems for other companies. At BottleTop an alert system based on AmI technologies in the form of wireless communications has the potential to improve productivity by decreasing the number and duration of production stoppages caused by machine parameters moving out of set tolerances, and release personnel from repetitive inspection tasks for higher value, and more rewarding and interesting work. The following section compares the detailed findings from one Australian SME (BottleTop) with those from a German SME participating in the European AmI4-SMEs project. Despite differences in industry and location, both these SMEs have very similar opportunities to address production issues using AmI Technologies, pointing to the potential for standardised AmI based solutions to improve SME manufacturing operations.

Comparison between European and Australian SME Manufacturers German SME (Truckbody GmbH) Truckbody GmbH claims market leadership for EU manufacture of truck swap bodies (steel framed transport containers, and the legs on which they stand while waiting transfer from truck to truck, or truck to rail) primarily intended for the EU domestic market. A key competence is the manufacture and powder coating of large structures, up to 15m long, such as bus frames. The company employs 330 people, which places it in the EU classification of SME organisations, in contrast to Australia where a SME is classified as an organisation with between 20 and 200 employees. Truckbody’s production system is characterized by strong interdependencies between different task groups; a delay in one step impacts many other groups further down the production line. To reduce production delays the EU Ami-4SMEs project research and technology partners identified a need for automatic production alerts

75

Process Innovation with Ambient Intelligence (AmI) Technologies in Manufacturing SMEs

that interfaced to the company’s planning system. The ATB Institute for Applied Systems Technology, based in Bremen, Germany is the project leader of the EU AmI-4-SMEs project, and is currently finalising the implementation of a rule engine and user interfaces on mobile devices. When problems occur in Truckbody’s production, employees who need to know about the disruption, such as the shift supervisor and the person with the skill to solve the problem, receive an automatically generated alert message. The alerts are based on user profiles (e.g. manager, foreman), the current location of the user (e.g. meeting, office, home) and the severity of the situation (e.g. deviation threshold, breakdown, loss). Use of a multi-modal user interface, (specifically a wireless message sent to a mobile phone or PDA) leverages the capability of AmI technologies to provide timely alerts that are “pushed” to relevant employees regardless of their location. In this manner, the AmI technology provides immediate and mobile access to production information, warns of delays to the production line, and so supports reallocation of work and staff. Prototypes have been developed as part of the Selection and Specification phase of the study.

Australian SME (BottleTop Pty Ltd) BottleTop produces a very different product from that of Truckbody GmbH. BottleTop manufactures specialty packaging, with particular strengths in the personal care, pharmaceutical, health foods, chemical, cleaning, food, beverage and cosmetics markets. Operating for sixty years from its single Sydney manufacturing site, it has built a strong sense of loyalty among its 97 employees, and has extensive links to international fastening manufacturers. Although plastic manufacturing accounts for around 7% of all Australian manufacturing activity, the industry is quite mature (McCaffrey, 2006), and is shrinking at around 4% per year, mainly due to increased purchases from foreign injection

76

moulding companies. BottleTop is growing in this shrinking market, winning market share from its competitors by focussing on quality, service, technology and relationships within and outside the organisation. The company plans to more than double its revenue by 2011/12. While the revenue goal is ambitious, BottleTop’s revenue grew by 12% in 2006 even after allowing for a 5% reduction in revenue from its existing customer base due to some customers moving their business to off-shore suppliers. Discussions with BottleTop’s Production Management team identified the following AmI technology scenario as an attractive business concept: BottleTop’s moulding and assembly machines have in-built Programmable Logic Controllers (PLCs) which can monitor the six key variables that control the formation of the plastic closure. If a software program collects and monitors the PLC data, when any of these six parameters move outside pre-set limits an SMS alert to a mobile phone, or pager message could be automatically generated and sent to on-site maintenance personnel. This provides several potential business benefits, including minimization of machine downtime, reduction of defective, scrapped product and reduced need for visual inspection. Currently all the plastic fasteners are inspected by a human operator as they leave the machine. Previous attempts to use computers coupled to cameras to replace human visual inspection of parts leaving the injection moulding machines were not successful due to the camera’s inability to cope with reflective foil routinely used in BottleTop’s products. It is important to note that the company’s HR practices are likely to support the introduction of the proposed AmI solution. A bonus scheme rewarding operators for reducing the amount of defective caps produced from each machine has been enthusiastically embraced; operators have been heard to comment, “That’s my money on the floor” when the speed of the machine is set too fast and fasteners overshoot the hopper.

Process Innovation with Ambient Intelligence (AmI) Technologies in Manufacturing SMEs

The preceding comparison demonstrates that despite operating in unrelated industries in different countries, some SME manufacturing processes have sufficient commonality to permit the development of generic AmI solutions. Furthermore, the appearance of the same requirement in different manufacturing contexts shows that AmI technologies have the potential to be “general purpose” production enablers in diverse SME manufacturing settings. This in turn suggests the possibility that affordable “turn-key” AmI solutions may become available from technology providers. The next section considers the possibility of third party technology providers tailoring generic AmI solutions to the specific requirements of each SME, thus overcoming the absorptive capacity limitations inherent in SMEs.

DISCUSSION AND DIRECTIONS FOR FUTURE RESEARCH In Australian manufacturing SMEs, there is a very low likelihood of in-house R&D being used to build absorptive capacity to investigate AmI technologies. SMEs prefer to buy new technology when it is already embedded in an industry specific product rather than master the details of the underlying innovation (Oyelaran-Oyeyinka & Lal, 2006). Instead, we propose that SMEs are more likely to use industry or informal networks to become aware of potentially useful innovations, and then “buy” the innovation embedded in capital equipment or consulting services as a means to ‘recognise the value of new information, assimilate it, and apply it to commercial ends’ (Cohen & Levinthal, 1990, p.128). However, Cohen and Levinthal (1990) question the effectiveness of “buying” absorptive capacity in the form of consulting services or through acquisitions when the knowledge is to be integrated with existing business systems. They state “To integrate certain classes of complex and sophisticated technological knowledge successfully into

the firm’s activities, the firm requires an existing internal staff of technologists and scientists who are both competent in their fields and are familiar with the firm’s idiosyncratic needs, organizational procedures, routines, complementary capabilities, and extramural relationships” (Cohen & Levinthal, 1990, p. 135). Out-sourcing of deep absorptive capacity to equipment and software vendors able to provide “turn-key” solutions that match industry requirements seems to be a way for manufacturing SMEs to gain the commercial benefits of AmI technologies despite the resource and time constraints that prevent them building absorptive capacity in any area other than their core business competence. Similar requirements appear in two very different SMEs on two continents. The potential for the same AmI technology solution components to address these requirements, albeit tailored to the specifics of equipment in use at each site, suggests that SMEs can benefit from AmI technologies by using specialised intermediary organisations to provide the “absorptive capacity” on their behalf. This finding points to potential links between absorptive capacity and “make vs. buy” decision-making, and to “broad” or “deep” versions of absorptive capacity (Henard & McFadyen, 2006) as avenues for future research. In addition, an opportunity exists to track the spread of AmI technologies in SME Australian manufacturers and in doing so contribute to the diffusion of innovation literature. Additionally, AmI implementation challenges for Australian SME manufacturers extend beyond the boundaries of their own organisations. Large ICT manufacturers use a channel marketing approach to sell their products to the SME market segment. The channels may include retail and direct sales forces, but frequently hardware is “bundled” with service and software offerings from business partners, specialising in a particular industry segment, such as manufacturing. While intermediary business partners may supply specialised knowledge and generic AmI solutions to compensate for limited SME absorp-

77

Process Innovation with Ambient Intelligence (AmI) Technologies in Manufacturing SMEs

tive capacity, the organisations that partner with large ICT providers are often SMEs themselves. The ability and willingness of these business partners to gain AmI skills may in turn be a limiting factor in the adoption of AmI technologies by Australian Manufacturing SMEs. Absorptive capacity limitations of SME organisations can potentially affect uptake of AmI technologies at two points: within the manufacturing SME and within the SME technology partner. Low levels of in-house AmI skills and heavy level reliance on SME Australian technology providers suggest there may be an argument for the provision of government subsidies to encourage the adoption of AmI technologies in Australian manufacturing. A precedent exists in that subsidies have been provided for the purchase of RFID scanners for NSW meat producers (NSW Farmers Association, 2007). Without some form of government encouragement, the task of integrating AmI systems with existing ICT investments and the concomitant diversion from core manufacturing activities, may be enough to prevent the adoption of AmI technologies and, therefore, achievement of the elusive “productivity surge” in Australian manufacturing SMEs.

GmbH, 2008) suggest that SMEs can use “external research sub-units” in the form of experiences reported by members of their industry network and trade associations, and solutions proposed by research and technology providers, to offset internal absorptive capacity limitations.

CONCLUSION

REFERENCES

Advances in the development of ICT industry standards, and the proliferation of software and support for the Windows/Intel platform since Cohen and Levinthal’s 1990 paper have brought technology to SMEs without the need for bespoke development. Furthermore, Cohen and Levinthal appear to assume that investments in absorptive capacity only exist in the form of R&D spending, rather than networking with other organisations to use “connect and develop” models typical of Open Innovation (Chesbrough, Vanhaverbeke, & West, 2006). In contrast, the results from the EU and Australian AmI-4-SME projects (ATB Institute for Applied Systems Technology Bremen

Angeles, R. (2005). RFID Technologies: SupplyChain Applications and Implementation Issues. Information Systems Management, 22(1), 51–65. doi:10.1201/1078/44912.22.1.20051201/85739.7

78

ACKNOWLEDGMENT The authors would like to thank the SME participants in the Australian AmI-4-SME project for providing access to, and information about their organisations. We would also like to thank our colleagues in Australia, Assoc. Prof. Mile Terziovski and Richard Ferrers at the University of Melbourne, Dr David Low at the University of Western Sydney and our colleagues at the ATB Institute for Applied Systems Technology; Bremen, Germany for their helpful advice and the opportunity to compare Australian and EU based SMEs. The authors would also like to acknowledge the support of an International Science Linkages grant from the Australian Department of DIISR (Project Number CG110181), the University of Western Sydney and the University of Melbourne.

ATB Institute for Applied Systems Technology Bremen GmbH. (2004). COST-WORTH Methodology Independent Reference Scheme [Electronic Version]. Retrieved April 22, 2008 4:25 pm from ATB Bremen GmbH. ATB Institute for Applied Systems Technology Bremen GmbH. (2008). AMI-4-SME Platform. from http://ami4sme.org/ results/platform.php

Process Innovation with Ambient Intelligence (AmI) Technologies in Manufacturing SMEs

Australian Bureau of Statistics. (2001). 1321.0 Small Business in Australia. Retrieved April 21, 11am, 2008, from http://www.abs.gov.au/ Ausstats/[email protected]/ 0/97452F3932F44031CA 256C5B00027F19?Open Australian Bureau of Statistics. (2007). 8104.0 - Research and Experimental Development, Businesses, Australia, 2005-06. Retrieved September 3, 2:50pm, 2008, from http://www.abs.gov.au/ AUSSTATS/[email protected]/ Latestproducts/ 8104.0Main% 20Features32005-06? opendocument&tabname =Summary&prodno =8104.0&issue =200506&num =&view= Australian Bureau of Statistics. (2009). 8129.0 - Business Use of Information Technology, 2007-08 (Publication. Retrieved September 17, 2009: http://www.abs.gov.au/ AUSSTATS/abs@. nsf/ Latestproducts/ 29F02AF6A 0F6C9A3CA 2576170018FB56? opendocument Australian Productivity Commission. (2003). Trends in Australian Manufacturing. Retrieved April 21, 11:46 am, 2008, from http://www.pc.gov. au/ research/commission research/tiam/ keypoints Australian Productivity Commission. (2004). ICT Use and Productivity: A Synthesis from Studies of Australian Firms. Retrieved. From http://www. pc.gov.au/ research/commission research/ictuse Beckett, R. C. (2008). Utilizing and Adaptation of the Absorptive Capacity Concept in a Virtual Enterprise Context. International Journal of Production Research, 46(5), 1243–1252. doi:10.1080/00207540701224327 Brown, S., & Bessant, J. (2003). The Manufacturing Strategy-Capabilities Links in Mass Customisation and Agile Manufacturing - an Exploratory Study. International Journal of Operations & Production Management, 23(7), 707–730. doi:10.1108/01443570310481522

Campos, A., Pina, P., & Neves-Silva, R. (2006). Supporting Distributed Collaborative Work in Manufacturing Industry. IEEE Xplore. Retrieved Feb 22, 2010 Caprio, D. W. J. (2005). Radio-Frequency Identification (RFID): Panorama of RFID Current Applications and Potential Economic Benefits. Paper presented at the Committee for Information, Computer and Communication Policy (ICCP) of the Organization for Economic Co-operation and Development (OECD). Retrieved Viewed April 14, 2008, from http://www.oecd.org/ dataoecd/60/ 8/3546 5566.pdf Chang, S. E., & Heng, M. S. H. (2006). An Empirical Study on Voice-Enabled Web Applications. Pervasive Computing (July-September 2006), 76 - 81. Chapman, R., Toner, P., Sloan, T., Caddy, I., & Turpin, T. (2008). Bridging the Barriers: A Study of Innovation in the NSW Manufacturing Sector. Sydney: University of Western Sydney. Chesbrough, H. W., Vanhaverbeke, W., & West, J. (2006). Open Innovation: Researching a New Paradigm. Oxford: Oxford University Press. Chong, S., & Pervan, G. (2007). Factors Influencing the Extent of Deployment of Electronic Commerce for Small-and Medium-Sized Enterprises. Journal of Electronic Commerce in Organizations, 5(1), 1–29. Cochran, P. L., Tatikonda, M. V., & Magid, J. M. (2007). Radio Frequency Identification and the Ethics of Privacy. Organizational Dynamics, 36(2), 217–229. doi:10.1016/j.orgdyn.2007.03.008 Cohen, W. M., & Levinthal, D. A. (1990). Absorptive Capacity: A New Perspective on Learning and Innovation. Administrative Science Quarterly, 35(1), 128–152. doi:10.2307/2393553

79

Process Innovation with Ambient Intelligence (AmI) Technologies in Manufacturing SMEs

Cutler, T. (2008). Venturous Australia: Building Strength in Innovation; Report of the Review of the National Innovation System in Australia. Melbourne: Cutler and Company. Department of Innovation Industry Science and Research. (2008, Feb 19, 2008). Small Business Surveys - Finance and Banking, Innovation, and International Activity Retrieved June 3, 2008, from http://www.innovation.gov.au/ Section/ SmallBusiness/ Pages/Small Business Surveys Finance and Banking Innovation and International Activity.aspx Ellis, R. M. (2004). The Challenges of Work/ Home Boundaries and User Perceptions for Ambient Intelligence [Electronic Version]. Chimera Working Paper, 2004-14. Colchester: University of Essex. Retrieved July 1, 2008, from http:// www.essex.ac.uk/ chimera/content/ pubs/wps/ CWP-2004-14-RE -e-Challenges.pdf Game-Lopata, A. (2008). Exclusive: Secret Sauce [Electronic Version]. Logistics. Retrieved September 18, 2009, from http://www.logisticsmagazine. com.au/ Article/Exclusive -Secret-Sauce/ 172799. aspx Germain, R., Droge, C., & Daugherty, P. J. (1994). A Cost and Impact Typology of Logistics Technology and the Effect of its Adoption on Organizational Practice. Journal of Business Logistics, 15(2), 227–248. Gunasekaran, A., & Yusuf, Y. Y. (2002). Agile Manufacturing: a Taxonomy of Strategic and Technological Imperatives. International Journal of Production Research, 40(6), 1357–1385. doi:10.1080/00207540110118370 Henard, D. H., & McFadyen, M. A. (2006). R & D Knowledge is Power. Research Technology Management, 49(3), 41–47.

80

Kable’s Government Computing. (2008). Chancellor sets 30% Target for SME Procurement [Electronic Version]. KableNET. Retrieved April 21, 12:10pm from www.kablenet.com / kd.nsf/ FrontpageRSS/ 23B6ED 428208979 68025740A 004A2E25! Open Document Kinder, T. (2002). Emerging E-commerce Business Models: an Analysis of Case Studies from West Lothian, Scotland. European Journal of Innovation Management, 5(3), 130–151. doi:10.1108/14601060210436718 Kirchhoff, U., Stokic, D., & Sundmaeker, H. (2006). AmI Technologies Based Business Improvement in Manufacturing SMEs. Paper presented at the eChallenges e-2006. from http:// www.ami4sme.org/. Kopacsi, S., Kovacs, G., Anufriev, A., & Michelini, R. (2007). Ambient Intelligence as Enabling Technology for Modern Business Paradigms. Robotics and Computer-integrated Manufacturing, 23(2), 242–256. doi:10.1016/j.rcim.2006.01.002 Kuehnle, H. (2007). Post Mass Production Paradigm (PMPP) Trajectories. Journal of Manufacturing Technology Management, 18(8), 1022–1037. doi:10.1108/17410380710828316 Li, X., Feng, L., Zhou, L., & Shi, Y. (2009). Learning in an Ambient Intelligent World: Enabling Technologies and Practices. IEEE Transactions on Knowledge and Data Engineering, 21(6), 910–924. doi:10.1109/TKDE.2008.143 Liao, J., Welsch, H., & Stoica, M. (2003). Organizational Absorptive Capacity and Responsiveness: An Empirical Investigation of Growth-Oriented SMEs Entrepreneurship. Theory into Practice, 28(1), 63–85. Maurtua, I., Perez, M. A., Susperregi, L., Tubio, C., & Ibarguren, A. (2006). Ambient Intelligence in Manufacturing. Paper presented at the Intelligent Production Machines and Systems, 2nd I*PROMS Virtual Conference.

Process Innovation with Ambient Intelligence (AmI) Technologies in Manufacturing SMEs

McCaffrey, J. (2006). Version]. Retrieved Dec 17, 2007 from http://commercecan.ic.gc.ca/ scdt/bizmap/ interface2.nsf/ vDownload/ISA_ 5111/$file/X_ 4862570.PDF McCullough, M. (2001). On Typologies of Situated Interaction. Human-Computer Interaction, 16, 337–349. doi:10.1207/S15327051HCI16234_14 Muscio, A. (2008). The Impact of Absorptive Capacity on SMEs’ Collaboration. Economics of Innovation and New Technology, 16(8), 653–668. doi:10.1080/10438590600983994

Stokic, D., Kirchhoff, U., & Sundmaeker, H. (2006). Ambient Intelligence in Manufacturing Industry: Control System Point of View. Paper presented at the IACTED Conference on Control and Applications 2006 conference Vasilakos, A. V. (2008). Ambient Intelligence. Information Sciences, 178(3), 585–587. doi:10.1016/j.ins.2007.08.016 Weber, W. (2003). Ambient Intelligence - Industrial Research on a Visionary Concept. IEEE Xplore Retrieved Feb 22, 2010

Nousala, S., Ifandoudas, P., Terziovski, M., & Chapman, R. L. (2008). Process Improvement and ICTs In Australian SMEs: A Selection and Implementation Framework. Production Planning and Control, 19(8), 735–753. doi:10.1080/09537280802476169

Wiesner, R., McDonald, J., & Banham, H. C. (2007). Australian Small and Medium Sized Enterprises (SMEs): A study of high performance management practices. Journal of Management & Organization, 13(3), 227–248. doi:10.5172/ jmo.2007.13.3.227

NSW Farmers Association. (2007). National Livestock Identification System (Cattle) 013.07i NLIS Cattle [Electronic Version]. Retrieved April 10, 2008 from http://www.nswfarmers.org. au/ data/assets/ pdf_file/ 0012/3072 /FS_NLIS_ Cattle_0207.pdf

Zuboff, S. (1988). In the Age of the Smart Machine: the Future of Work and Power. Oxford Heinemann Professional.

Oyelaran-Oyeyinka, O., & Lal, K. (2006). SMEs and New Technologies; Learning E-Business and Development. Basingstoke, UK: Palgrave McMillan. doi:10.1057/9780230625457 Philips, R. (1997). Innovation and Firm Performance in Australian Manufacturing. Industry Commission Staff Research Paper. Canberra, AGPS Rao, B., & Zimmermann, H.-D. (2005). Pervasive Computing and Ambient Intelligence: Preface [Electronic Version]. Electronic Markets, 15, p. 3. Retrieved June 3, 2008 from http://www. informaworld.com/ smpp/ftinterface ~content= a713735017 ~fulltext= 713240928.

KEY TERMS AND DEFINITIONS Ambient Intelligence Technologies (AmI): Technologies that combine to create an environment that is are sensitive and responsive to the presence of people Pervasive and Ubiquitous Computing: Terms in use in the USA to refer to the same technologies as those named Ambient Intelligence Technologies in Europe Absorptive Capacity: The absorptive capacity of a firm is comprised of its ability to generate innovations, and absorb relevant knowledge appearing in the external environment. Small Medium Enterprise SME: Small to Medium Enterprise. This measure of a company’s size is generally based upon employee numbers, and varies across countries.

81

Process Innovation with Ambient Intelligence (AmI) Technologies in Manufacturing SMEs

Radio Frequency Identification Device (RFID): Data collection devices consisting of electronic tags for storing unique identifying data. Technology Adoption: A process that begins with awareness of a specific type of technology or device, and progresses through stages ending in use or rejection of the technology.

82

Open Innovation: Organisations using ideas and capabilities originating outside their boundaries in order to increase the rate with which innovation occurs and decrease innovation costs. Open innovation also includes an organisation selling innovative ideas it has generated but cannot use in its business.

83

Chapter 6

Providing Ubiquitous Access to Synthetic Sign Language Contents over Multiple Platforms Fernando López-Colino Universidad Autónoma de Madrid, Spain Jose Colás Universidad Autónoma de Madrid, Spain

ABSTRACT This work presents the design of a distributed sign language synthesis architecture. The main objective of this design is to adapt the synthesis process to the diversity of user devices. The synthesis process has been divided into several independent modules that can be run either in a dedicated server or in the client device. Depending on the modules assigned to the server or to the client, four different scenarios have been defined. These scenarios may vary from a heavy client design which executes the whole synthesis process, to a light client design similar to a video player. These four scenarios will provide equivalent signed message quality independently of the device’s hardware and software resources.

INTRODUCTION Sign language (SL) synthesizers have been designed as PC-based applications because they DOI: 10.4018/978-1-60960-042-6.ch006

require complex calculations, specific libraries, and 3D capable devices. Current mobile device hardware and software resources do not fulfill all these requirements. Although several mobile devices can manage simple 3D contents and real time animations, they cannot manage the previ-

Copyright © 2011, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited.

Providing Ubiquitous Access to Synthetic Sign Language Contents over Multiple Platforms

ous steps that generate the final animation of the virtual avatar. Unfortunately, there are many real situations where a mobile device is the only alternative to obtain access to signed contents, such as: real-time translation, virtual museum guides, transport information services, etc. Existing approaches to providing signed contents focus only on a single platform. The adaptation between different technologies is an expensive process that requires repeating the comprehensibility evaluations in order to check that the signed messages obtained with the new implementation are comprehensible. The purpose of this chapter is to present a global and unique solution to providing synthetic signed contents to most kinds of devices, from PCs to mobile phones and gaming consoles. Instead of reducing the synthesis features using a low quality avatar or adapting the synthesis method to the device resources, our approach assigns to the user’s device only the modules that it can manage. The signed messages must present the same quality and intelligibility device independently. The architecture uses a unique module responsible for the definition of the avatar’s animation, which defines the message comprehensibility; therefore we avoid repeating this evaluation.

BACKGROUND Literature provides several examples of SL synthesizers. In order to represent synthetic signed messages, two main techniques have been developed: 1. The first approach to SL synthesis consists of creating a composition of small segments of video (Solina, Krapež, Jaklič, & Komac, 2001). This approach to SL synthesis requires image processing and a great number of pre-recorded sequences in order to act as a synthesizer, and thus significant storage capacity.

84

2. The second main approach to SL synthesis uses virtual avatars. H-Anim (ISO/IEC 19774, 2005) is the most widely used avatar structure; it is a standard definition for human representation on VRML (ISO/IEC 14772-1, 1997) or X3D (ISO/IEC 19775, 2004). Within avatar animation category, there are two different approaches related to the definition of the animation. The first one uses continuous motion data obtained from (a) an expert signer using different motion capture techniques or (b) manual animations created by an expert animator. Although the results obtained with this technique are natural, Kennaway (2002) described several disadvantages of this approach based on the difficult adaptation of the recorded data to avatars with different anatomies. The second approach to define the animation for avatar-based SL synthesis uses a parametric definition of the signs in order to generate the animation (Bangham, Cox, Elliot, Glauert, & Marshall, 2000; Irving & Foulds, 2005; Kennaway, Glauert, & Zwitserlood, 2007; Zwiterslood, Verlinden, Ros, & van der Shoot, 2004). The resulting avatar animation is not as natural as the one obtained using the continuous motion data approach. However, the animation quality is the same over the whole sentence and the storage requirements are highly reduced. The parametric synthesis is the only approach that provides enough flexibility to define all the SL linguistic variations. Our synthesis approach presents the signed message using a 3D avatar and is based on parametric sign definitions. Hence, it is also interesting to review the existing research focused on handling 3D contents over mobile devices. All authors agree that mobile device resources, specifically those required resources for running 3D applications, are limited. Boukerche and Pazzi (2006) proposed the use of a remote rendering technique to deal

Providing Ubiquitous Access to Synthetic Sign Language Contents over Multiple Platforms

with the lack of resources of these devices. On the other hand, Nadalutti, Chittaro and Buttussi (2006) presented a render application for mobile devices using the X3D standard for VRML applications and Lukashev, Puresev, and Makhlushev (2006) reported the available 3D API for Java-based mobile phones. Both papers concluded that the scene complexity rendered on these devices cannot be the same as on PCs. Agu, Banerjee, Nilekar, Rekutin and Kramer (2005)proposed several techniques to simplify several aspects of the 3D scene, such as geometric complexity adaptation or the use of progressive meshes. However, these techniques cannot be automatically applied to the avatar definition, as they may reduce its intelligibility. Finally, we review several projects focused on providing signed contents to mobile devices. Ahmed and Seong (2006) proposed an application that displays signed contents using the SignWriting notation (Sutton, 1974). The SignWriting notation is more suitable for deaf people that text, however its relation to SL is the same relation between text and speech and it is not the preferred communication approach among deaf people. Representing signed contents on mobile devices follow the same research approaches described for desktop synthesizers: 1. Several projects use video recordings to provide signed contents: Kimura et al. (2008) retrieved tagged videos to be played in the mobile device, the project Blue Sign Translator (Bennati, Capasso, Giallombardo, Maggio & Giorgi, 2006) concatenated several pre-recorded videos in order to synthesize the signed message. These videos were stored and concatenated in a server previously to sending them to the mobile device.↜Although it cannot be considered SL synthesis, we would like to emphasize the work presented by Cherniavsky, Cavender, Ladner, and Riskin (2007). Their study of the frame rate required for SL video conferenc-

ing can be also applied to the streaming of synthetic SL contents. 2. The avatar-based approach for providing signed contents on mobile devices uses manually-defined animations (Chittaro, Buttusi, & Nadalutti, 2006; Xu, Wang, Yao, Zhang, & Zhao, 2008) and VRML viewers installed on the mobile device. A SL expert defines each sign by means of its most relevant static gestures and next the system creates the final animation interpolating the orientation of each skeleton bone between these static gestures. This approach only handles the animation of the avatar, but the generation of the avatar’s gestures is a manual process and does not exploit the advances applied on desktop synthesizers.

DISTRIBUTED ARCHITECTURE The SL synthesizer has been designed to deal with the great diversity of the final user devices. In order to cover most hardware and software platforms, a distributed architecture has been established, dividing the whole process into five different and independent modules. Figure 1 depicts these five modules: the HLSML Parser module, the Retrieval module (composed by the Avatar Description Retrieval and the Sign Description Retrieval sub-modules), the Gesture Synthesis module, the Rendering module, and the Visualization module. The following subsections describe each of these modules.

HLSML Parser Module The first step of a synthesis system is the parsing of the input message. Written text is the graphical representation of speech and it is used for speech synthesizers. However, there is no standard written representation of SLs. Although there are several self contained notations that allow parametric definitions of established signs, such as SiGML

85

Providing Ubiquitous Access to Synthetic Sign Language Contents over Multiple Platforms

Figure 1. This image depicts the main modules of the SL synthesis architecture

(Elliot, Glauert, Jennings & Kennaway, 2004; Kennaway, Elliot, Glauert & Parsons, 2002), SWML (Rocha & Pereira, 2001) or SEA (Herrero Blanco, 2004a), none of them has been completely accepted by the deaf community as written representation of SLs. All of these notations require deep knowledge of SL phonology to define a message. We have chosen the HLSML notation, developed by López-Colino and Colás (2009), because it focuses on message definition and SL syntax, simplifying the manual message description task. The phonologic definitions of the signs are stored in a Relational Database which allows more complex and precise definitions compared to the previously mentioned notations (the structure of the Relational Database is briefly explained in the Retrieval module section). Two relevant elements from the HLSML should be noted: the Classifier Constructions (CCs) definition (Herrero Blanco, 2004b; Liddell, 2003) and the prosodic modifiers:

requires describing avatar’s hands animation, so the complexity in the message and the parsing process is increased; the last part of this subsection reports the increase of parsing time when a CC is present in a message. 2. The prosodic modifiers that can be defined using HLSML are relevant for the Retrieval module and the Gesture Synthesis module. The Relational Database stores different representations for every sign representing different mood states. Hence, the modifiers stated in the HLSML message are used during the retrieval stage in order to obtain the correct mood variation. These modifiers also describe other prosodic aspect such as global signing speed, the transition time between signs, and the movement gentleness. The Gesture Synthesis module applies these modifications to the sign definition retrieved from the Relational Database.

1. CCs are semantically complex units in signed messages, which may imply complex signer movements similar to mimicry. The CCs can be used to describe a situation or the spatial disposition of several objects mentioned in the message. Therefore, the CCs have to be defined in the input message and processed in the Gesture Synthesis module in order to create the relevant avatar’s animation that will represent these units. The CCs definition

The HLSML Parser module has been implemented using the XML SAX parser (Megginson, 2004). In order to estimate the increase of the parsing time when processing a CC, we have measured the parsing time of several sentences including only established signs and several sentences including only CCs. We have considered only the time required by the parsing method call. The obtained results have shown that the average parsing time required for a single sign is 14 ms

86

Providing Ubiquitous Access to Synthetic Sign Language Contents over Multiple Platforms

and 101 ms for a CC. Obviously, these parsing times depend on the implementation but they can be used to estimate the relation between these two message units. The CC parsing process takes approximately seven times more than parsing a established signs. This module generates (a) the sequence of signs whose phonologic descriptions have to be recovered, (b) the description of the message’s CCs, and (c) the avatar’s skin identifier (selecting avatar’s appearance) that the Retrieval module will recover from a server or a local directory, as we will describe next.

Retrieval Module This module comprises the retrieval of the avatar description (Avatar Description Retrieval submodule) and the sign descriptions (Sign Description Retrieval sub-module). The synthesizer main implementation is based on the JSR-184 standard (JSR-184 Expert Group, 2005), which is used as 3D API engine and for the scene description. One of the main contributions of this standard was the definition of the m3g-file format which allows the description of all the elements in the scene such as geometry, lights, cameras, animation tracks, mesh materials, textures, and hierarchical bone structures. The signing avatar is defined using a skeleton mesh with a multiple-material assignment. The mesh is a connected network of thousands of polygons, each of them assigned to a smoothing group and a material definition. This assignment to a smoothing group defines continuous surfaces whose illumination should be processed together in order to avoid edge effects in the visualization. The mesh is a static structure whose deformation, required to represent different gestures, implies the definition of a control structure: a skeleton. The skeleton is a hierarchical set of nodes, usually named as bones, whose reference system is defined as a set of 3D transformations applied to the reference system of its parent node. This structure defines

the inheritance of every transformation applied to a node, which emulates the behavior of a real skeleton. The skeleton mesh defines the relation between every mesh’s vertex and a group of bones. Hence, making a change to any skeleton’s bone will modify the position of the vertexes related to that bone, causing a deformation in the mesh. Every avatar’s gesture is defined by means of the orientations of several skeleton’s bones, so the avatar gesture can be specified with a very small amount of data (compared to defining the mesh deformation describing each vertex’s position). The structure of the avatar follows the design reported by López, Tejedor, Garrido, and Colás (2006), which presents two main advantages which simplified the synthesis: 1. The animation approach is exclusively based on the skeleton approach, including the face expression animation. Most approaches use a morphing approach to define several face expressions. Each of these expressions is a copy of the mesh that must be transmitted and stored in the user’s device. For mobile devices with limited storage and network capacity, the skeleton approach reduces the amount of data that has to be transmitted to the orientations of facial bones. 2. The arm’s structure has been redefined so the changes in the upper arm and forearm do not modify the orientation of the wrist. This new structure simplifies the inverse kinematics process reducing the seven degrees of freedom that must be calculated using a standard arm’s structure (like H-ANIM’s) to four degrees of freedom. The Avatar Description Retrieval sub-module recovers the avatar description and any required auxiliary file using the HTTP protocol, which allows the system to recover the m3g-formated file from a web server or from a local directory. Obtaining the avatar’s description from a web server ensures the latest version of the avatar.

87

Providing Ubiquitous Access to Synthetic Sign Language Contents over Multiple Platforms

However, if the network communication should be minimized (e.g. for cost reduction), using a local copy of the avatar is also possible. The size of the latest version of the avatar’s file, dated August 2009, is 170 kB uncompressed and 76 kB compressed using “gzip” compression; the size of the related auxiliary file is 8 kB (used for skeleton management). In the previous section we presented the HLSML notation as a high level notation approach focused on sign message description as it relies on another source for storing sign phonologic definitions. Our SL synthesis approach uses a Relational Database to store these phonologic sign definitions. The Relational Database stores the Phonologic Parameters1 (PPs) and their definitions; it also stores the sequences of these PPs that describe each sign and the alternative representations of the different mood states that the avatar can represent. Using a parametric approach, the amount of data required to describe a sign is lower than the required using the motion capture approach which stores the orientation of each avatar’s joint recorded between thirty and sixty times per second. The Sign Description Retrieval sub-module defines a series of SQL queries to the Relational Database in order to recover the phonologic descriptions of the message’s signs. For each PP of every sign, the synthesizer recovers the defined sequence for that sign. Each hand may define up to six sequences; there is another sequence for the non-hand parameter, if defined. The kind of sign defines the number of sequences: a static one-handed sign only requires five sequences whereas a dynamic two-handed sign describing non-hand parameters requires thirteen sequences. After recovering the sequences, the synthesizer obtains the description of each unit included in them, which are described using 3D orientations, 3D vectors, and anatomic references (using text strings). The average amount of data (measuring 120 random signs) transmitted from the synthesizer to the database server is 15 kB per sign, cor-

88

responding to SQL requests and network signaling, and the average amount of data corresponding to the SQL responses and the network signaling is 33.6 kB per sign. The CCs differ from the established signs in the amount of data recovered from the Relational Database. The CCs have only to recover the description of the PPs units included in the CC, as the sequences are described in the HLSML message. The average amount of data (measuring 50 different CCs) transmitted from the synthesizer to the database is 11 kB per CC, and the average amount of data recovered from the Relational Database is 21 kB per CC.

Gesture Synthesis Module The main task of the Gesture Synthesis module is to process the information obtained in the HLSML Parsing and the Retrieval modules and to define the avatar’s animation that represents the input message. Therefore, this module is the ultimate responsible of the synthetic message intelligibility. This module’s input are: (a) The message definition obtained from the HLSML Parsing module, which includes the signs sequence, the description of the CCs, and several prosodic modifiers that affect the final animation; (b) the morphologic descriptions of the signs included in the message; and (c) the definition of the avatar’s anatomy, both obtained in the Retrieval module. The avatar’s animation is described by means of Bone Animation Tracks (BATs). A BAT is a sequence of bone orientations with a timestamp that indicates the instant when the bone must reach that orientation. There are many approaches for 3D orientation definition; we have used quaternions (Hamilton, 1853) because they are the best approach for animating orientations, as stated Shoemake (1985). Therefore, the Gesture Synthesis module processes the PP definitions previously retrieved from the Relational Database to generate those bone orientations. The timestamps for each bone orientation is defined using the PP sequences (retrieved from the Relational

Providing Ubiquitous Access to Synthetic Sign Language Contents over Multiple Platforms

Database), the CCs definition from the HLSML message, and the prosodic modifiers. We have presented before that the PPs can be defined using 3D orientations, 3D vectors, and anatomic references: •

•

The 3D-orientation-defined PPs do not require further processing as the retrieved orientations can be directly used in the definition of the BATs. This PP definition is applied to hand shapes, hand orientation, body gestures, and face expressions. Both the 3D vectors and the anatomic references are used for hand positioning. The coordinates where the hands must be placed can be defined directly (by means of a 3D vector) or using a body reference, in this case, the Gesture Synthesis module obtains the coordinates of the referred anatomic point. The final sequence of coordinates require further processing in order to obtain the suitable shoulder, elbow, and wrist orientations to place the hand in the required coordinate and keeping the required hand orientation (defined by the PP hand orientation). This process is named inverse kinematics and will be briefly described next:↜The inverse kinematics process aims to obtain the required orientations of a sequence of bones in order to the final element of that sequence of bones achieves a defined position. The process may involve several orientation restrictions to some bones, increasing the complexity of the process. The standard definition of an avatar’s arm requires resolving a seven degree of freedom bone structure: the shoulder defines three degrees of freedom, the elbow defines one, and the wrist defines the last three degrees of freedom. As we have stated before, the avatar’s arm proposed by López et al. (2006) simplifies the inverse kinematics process as the wrist’s orientation is not modified by the

orientations of the shoulder and the elbow, removing the wrist’s degrees of freedom from the inverse kinematics process. Even with this simplification, the inverse kinematics process is the most time-consuming part of the synthesis process, as it is reported in the last part of this section.↜The transition between two consecutive coordinates defined for a hand can cause the hand to collide with the avatar’s body or head. This situation must be avoided using a collision detection and avoidance algorithm. The Gesture Synthesis module includes an algorithm that detects a collision between the hand and the body or the head and creates a collision-free trajectory. The collision avoidance algorithm requires a great amount of verifications in order to check if there is a collision in a defined instant; the algorithm conducts these verifications in different instants during each transition, converting the collision avoidance algorithm into the second most time-consuming process of the whole SL synthesis. The Gesture Synthesis module contains the two most time-consuming processes of the SL synthesis, the inverse kinematics process and the collision avoidance algorithm. As the execution of this module supposes a great challenge for the device, this module should be considered for server-side execution.

Rendering Module The rendering process consists of the generation of a 2D image (frame) from a 3D scene definition. The 3D animated scene definition is created by merging the scene definition retrieved from the Web Server and the BATs created in the Gesture Synthesis module. Each time the Rendering module has to generate one image, it has to carry out the update of the animated scene, which means that the orientation of each skeleton’s bone has to

89

Providing Ubiquitous Access to Synthetic Sign Language Contents over Multiple Platforms

be defined (this process will be explained next) and updating the mesh deformation as defined by the new skeleton bones’ orientation to obtain the positions of all the mesh’s polygons. When all the polygons are in their final position, the 3D engine defines the colors of each calculating the received illumination, their original color, and the applied texture. Finally, the system projects the 3D coordinates into the projection plane that is represented in the screen. The duration of this complex process depends on the number of polygons defined in the scene, the number of lights, and the complexity of the textures. The Rendering module generates a video signal, which can be stored in a video file if necessary, corresponding to the animation defined in the Gesture Synthesis module. This video signal is the input of the Visualization module. The previous subsection has defined the BATs as discrete sequences of quaternions that define the bone’s orientation in a precise instant of time; each of these timed-quaternions is named a keyframe. If the rendering instant does not coincide with one of these keyframes, the bone’s orientation must be calculated using interpolation algorithms. The Rendering module allows using two different interpolation algorithms: a linear interpolator which only considers the previous and the following keyframes and a spline-based interpolator which considers the whole animation definition in order to create a smoother transition between the values of all the keyframes. It must be noted that the complexity of the interpolation algorithm changes the duration of the rendering process. We can define two different rendering approaches: 1. The first approach is the real time rendering, as the application presents the resulting 2D image to the user after it has been processed. In this approach, the period of time between two consecutive images depends on the time required for processing and rendering each

90

image; this time also defines the animation instant to be presented. Therefore, the final animation fluidness, which directly affects the user’s experience, depends on the capacity of the final user’s device to process each image fast enough. The maximal time interval between two consecutive frames for a fluid animation is 60 ms approximately, which corresponds to a 16 frames per second animation rate. The time interval depends both on final user’s device hardware and software resources and system’s load. The first factor will define the highest rendering frame rate that the device can handle, but this frame rate is limited by system’s load. The system’s load depends on background programs activity; so changes in this activity will affect the rendering process and randomize the time interval between frames. 2. The second rendering approach is the nonreal time rendering approach. Previous rendering approach describes how the frame rendering time has influence on real time visualization, but if the visualization can be delayed, the frame rendering time has no effect. The time interval between two consecutive frames can be defined to fulfill any video format and ensure the fluidness of the animation, independently of the required time to process each frame. In addition, this approach also allows using complex rendering algorithms, advanced lightning, and more detailed textures in order to obtain a photorealistic result. The Rendering module allows two different approaches to generate the animated sequence: the real time rendering approach presents one advantage because it does not require storing the results of the rendering, as they are instantly displayed and every image is discarded when the following one has been processed. This real time rendering approach also has the disadvantage of the animation variable fluidness and the limitation

Providing Ubiquitous Access to Synthetic Sign Language Contents over Multiple Platforms

applied to image complexity and realism in order to reduce processing time. The non-real time presents the opposite advantages and disadvantages, as they complement each other.

Visualization Module The visualization of the resulting animated sequence is the last stage of the process. This module processes the video signal obtained from the Rendering module and presents it in the device’s screen. Therefore, the user’s device must run this module. Previous section described the relation between the Rendering module and the Visualization module. There are three different approaches for the Visualization module depending on whether it is run in the same device as the Rendering module or not and the rendering approach used: 1. The first approach is applied when both modules are run in the same device and the real time rendering approach is in use. The Rendering module video signal output is directly presented to the user. In this approach, the Visualization module acts as a simple interface between the Rendering module and the graphic device. This approach is equivalent to a 3D video game: real time rendering and local visualization. 2. The second approach is applied when the Rendering module is run in a server and the Visualization module in the user’s device and the real time rendering approach is in use. This approach is also known as remote rendering; each time a frame has been generated, the server sends it to the user’s device using any available network connection. This approach is similar to a video streaming, but this time the server generates the video “on the fly”. 3. The third approach is applied when the Rendering module uses the non-real time approach. The output of the Rendering

module has been previously stored in a video file and the Visualization module acts as a standard video player. In this approach, the resolution of the video and the compression codec can be adapted to the specifications of the user’s device.

Complexity of the Architecture’s Modules Previous subsections have described the five modules of the architecture. The aim of this subsection is to present the complexity of each module in order to define which modules should be run in a dedicated server depending on the user device’s resources. We have stated before that the final user device must run the Visualization module, so we will omit measuring its complexity. We also omit presenting the complexity of the Rendering module, as its performance is hardware related. Only devices which integrate dedicated 3D hardware will run the Rendering module, so assigning this module to the server depends on this feature. On the other hand, the complexity of the first three modules of the architecture has to be measured to decide whether they can be run in the user device or not. Using a desktop implementation of the architecture, we have measured the required time to run the HLSML Parsing, the Sign Description Retrieval, and the Gesture Synthesis modules. The Table 1 presents the results obtained using the following hardware configuration: a Pentium Core 2 Duo system with 3.12 GB of RAM memory and Java 1.6.0. The 3D API is the Hybrid Rasteroid J2RE implementation (Hybrid, 2006) of the JSR-184 standard. The table presents the average time measured after synthesizing forty times each kind of sign and CC. The static units (both signs and CCs) define a static gesture of the avatar whereas the dynamic units include the movement of the hands (so the collision avoidance algorithm is required).

91

Providing Ubiquitous Access to Synthetic Sign Language Contents over Multiple Platforms

Table 1. This table presents the time measures (in ms) for the first three modules of the synthesizer using different kinds of signs and CCs Sign One-handed sign Module

CC Two-handed sign

Static

Dynamic

14 (0.3%)

15.5 (5.2%)

101.5 (2.2%)

97.8 (10.7%)

101.3 (2%)

32 (10.7%)

56 (1.2%)

673.2

2674

238.5

2312

Static

Dynamic

Static

Dynamic

HLSML Parsing

14 (5.5%)

14 (1%)

14 (1.5%)

Sign Description Retrieval

65.1 (26%)

74.2 (5.3%)

161.3

718

Inverse kinematics Gesture Synthesis

Collision avoidance BAT generation Total

0

589.4

0

2130

0

2058

11.7

12.4

126.9

114.7

11.8

11.9

172 (68.5%)

1319.8 (93.7%)

800.1 (87.8%)

4918.7 (97.7%)

250.3 (84.1%)

4381.9 (96.6%)

The Gesture Synthesis module requires up to a 97.7% of the processing time consumed by the three modules. This time measure is consistent with the statement presented when describing the Gesture Synthesis module as the inverse kinematics and the collision avoidance processes require complex calculations. Therefore, mobile devices with low processing resources should avoid running the Gesture Synthesis module and rely on a server for generating the BATs.

FINAL USER DEVICE ADAPTATION The previous section presented the architecture of the synthesizer. This architecture can be divided in three main blocks: the first block groups the HLSML Parsing, the Retrieval, and the Gesture Synthesis modules; the second block is the Rendering module; and the third one is the Visualization module, which must be run in the user device. Depending on the user device’s hardware and software resources the first two blocks can be run on the user device or on a dedicated synthesis server. It must be noted that once a module has been assigned to the server, all the previous modules have to be run in the server. We propose four different scenarios depending on the blocks

92

distribution and the communication interface (see Figure 2): 1. The Alpha scenario assigns all the modules to the user device defining a heavy client approach. 2. The Beta scenario only assigns the Visualization module to the user device, leaving the other four modules run to the server. 3. The Gamma scenario assigns the Rendering and the Visualization modules to the user device and the server sends an m3g-formatted file containing the avatar description and the BATs to the client. 4. The Delta scenario also assigns the Rendering and the Visualization modules to the user device. In contrast to the Gamma scenario, the Delta scenario only sends the Animation Definition to the client, establishing a 3D API independent scenario.

Alpha Scenario: The Desktop Application The first scenario is equivalent to other desktop synthesis approaches, as it is designed for PC devices. The user device runs all the five modules

Providing Ubiquitous Access to Synthetic Sign Language Contents over Multiple Platforms

Figure 2. The modular architecture allows four different scenarios: Alpha) A heavy client running the whole SL synthesis system. Beta) The light client scenario focused on the visualization. Gamma) This scenario is oriented to 3D capable devices using the JSR-184 API. Delta) A multi 3D API scenario allowing devices based on different 3D architectures.

of the architecture, as PCs have enough resources to run them. The HLSML message can be manually defined (as shown in Figure 3), automatically obtained form a translation server, or downloaded from a server, depending on the architecture that includes this synthesizer. This is a heavy client approach and there is no need of a synthesis server. The network load is high as the application has to make several queries to the Relational Database and download the avatar description from the Web server, if required.

Figure 3. This figure represents a screenshot of the Desktop application developed using the Alpha scenario approach

Beta Scenario: Devices without 3D Capabilities This is the light client approach, mainly focused on weak and mobile client devices; it can be also applied to web browsers, or TDT subtitling. This scenario is useful for non-real time synthetic signed messages distribution, which implies storing the videos, and is the best solution for multimedia and offline applications. Figure 4 shows an example of this scenario, where the user device is

a Nintendo DS portable console. Developing an application for game consoles is more complex than the development of PC-based applications. For this scenario we have used the “Moonshell” video player for the Nintendo DS and the server uses the specific codec for this video player. Using

93

Providing Ubiquitous Access to Synthetic Sign Language Contents over Multiple Platforms

Figure 4. This figure shows the synthetic signed message in a Nintendo DS using the Beta scenario

the wifi connection available on the Nintendo DS, the video file is transmitted to the gaming device. The communication between the server and the client is based on video transmission; Figure 2 depicts two alternatives for this transmission. The first approach is based on file transmission; the Rendering module output is stored in a video file and transmitted to the client. This approach requires the Rendering module to finish generating the whole message, so the total duration of the synthesis process is incremented. The second approach is based on video streaming: the server transmits each frame to the client after the Rendering module has finished its process. This approach neither requires storing the Rendering output in a video file, nor has to wait till the message has been completely generated. This scenario presents the lowest client load as the client device only has to run the Visualization module. On the other hand, this scenario involves the highest network load, due to video transmission, and the highest server load because it has to run the first four modules of the architecture, and handle the video streaming process.

94

Figure 5. This figure shows an m3g file player developed for a mobile device

Gamma Scenario: Java Mobile Based Devices This scenario focuses on devices with 3D capabilities, which allow handling 3D contents and animations, but their processing resources do not allow running the first three modules, just because it is impossible2 or quite complex. This scenario takes advantage of the JSR-184’s m3g file definition because it includes both geometry and scene description and animation definition. This simplifies the definition of the client application as it is a “3D file player”, a real time renderer (see Figure 5). The design of this scenario aims at Javabased mobile devices that include the JSR-184 API. These devices can handle 3D games (these 3D games use the device’s rendering capabilities), but most of them will not be able to run the complex Gesture Synthesis module. Therefore, this scenario is the best approach for providing synthetic signed contents to these devices. This scenario presents a high server load as it has to run the Gesture Synthesis module and has to generate the m3g file containing the scene definition and the BATs. The client presents a medium processing load as it has to render the

Providing Ubiquitous Access to Synthetic Sign Language Contents over Multiple Platforms

m3g file. Finally, this scenario presents a low network load as the server only requires sending a single file to the client. The average size of an m3g file is 225KB per sentence.

Figure 6. This figure shows the XNA-based implementation created to play sign animations

Delta Scenario: 3D API Independent Approach This scenario is similar to the Gamma approach; it focuses on devices with 3D capabilities. The main difference between this approach and the previous one is the contents sent to the client. This scenario only transmits the BATs created in the Gesture Synthesis module to the client. The format of these BATs has been defined as an array of orientations with a timestamp. All the arrays use the same time scale in order to synchronize all the movements. The client application adapts the BATs to the format defined by its 3D API. These 3D API can be the JSR-184, as in the Gamma scenario, or other 3D APIs as VRML, OpenGL, or XNA (Direct3D). Therefore, this scenario is useful for avoiding reimplementation of the first three modules of the architecture. Figure 6 shows the implementation of this scenario using XNA. We only had to implement the Rendering module using this 3D API (the programming language is C#), but we could reuse the HLSML Parsing, the Sign Definition Retrieval, and the Gesture Synthesis modules previously implemented, for the Alpha scenario using Java. The server load of this scenario is lower than the Gamma scenario’s because the server does not create the m3g file and it only sends the BATs, as they have been created in the Gesture Synthesis module, to the client. On the other hand the client load is higher than the Gamma scenario’s because the client has to adapt the BATs to its 3D API animation format. It also has to run the Rendering module. The network load is initially higher than the Gamma scenario approach as the client has to download the avatar’s definition and the BATs from different sources. But as the avatar’s definition is only downloaded once, the

second and subsequent sentences will imply lower network load, as the BATs required for a single sentence imply the average data transmission of 13kB.

Multi-Scenario Approach and Dynamic Scenario Switching Each of the previous scenarios has presented the kind of device that motivated its design. However, there are devices that can handle several of the presented scenarios, like a PC. For this reason, the last final user adaptation approach defines a multi-scenario approach applied for those devices that can handle several ones. The choice of the scenario is based on server’s load and network’s load and cost. User’s preferences are only considered to minimize economic costs, as the quality of the signed message is scenario independent and the user cannot distinguish between them3. The dynamic scenario switch is only possible between different signed blocks (sentences or paragraphs). The Beta, Gamma, and Delta scenarios can be described as different approaches to a synthesis service, with different degrees of server processing load. The Alpha scenario does not assume any

95

Providing Ubiquitous Access to Synthetic Sign Language Contents over Multiple Platforms

server load as the whole synthesis process is run in the client device. Depending on the number of clients connected to the synthesis server and the scenarios that they can run, the server will handle their requests using the most suitable scenario. For example, a PC-based client may start using the Beta scenario but if the server’s load rises, the system can dynamically change to the Alpha scenario, so the server can manage the request of other devices that cannot handle the most demanding scenarios. The network load and cost is also considered when choosing the scenario. For example, a pocket PC device can be either connected using a free wifi connection or a payment 3G connection. Depending on the used wireless network, the system will switch between the Beta and the Gamma scenarios.

FUTURE RESEARCH DIRECTIONS Future work is closely related with new mobile devices, new wireless communications, and new 3D API standards. New portable devices with higher processing resources will allow the development of a pseudo-Alpha scenario, using the server resources only for the inverse kinematics and the collision avoidance algorithms. The new wireless communications (meaning faster and cheaper communications) will allow the transmission of more realistic textures and avatars for the Alpha, Gamma, and Delta scenarios and high resolution videos for the Beta scenario. In order to adapt to new 3D APIs, it will be necessary the development of new client applications for the Delta scenario-based approach. The XNA-based development, currently used for the Delta scenario and run on PC environment, can be adapted to the XBOX; this would represent the first approach which would integrate synthetic signed contents into one of the main gaming consoles.

96

CONCLUSION This chapter presents a global solution to providing synthetic signed contents on most kinds of devices with the same quality and reducing the development efforts. The architecture of the system is divided into several different modules that can be assigned to a dedicated server when they cannot be run in the client device. This approach allows providing the same synthetic signed contents regardless the available hardware and software resources. The final user’s device adaptation is done by means of one of the four different distributions of the main modules of the architecture. The system can also switch between different scenarios, when available, allowing network and server load dynamic balancing. The switch between the scenarios is transparent to the user and it does not deteriorate user’s experience. The main architecture is implemented using Java, J2RE for the desktop design and J2ME for the mobile implementation. Hence, the adaptation between different systems supporting Java is simplified. We also consider non Java-based devices such as Pocket PC and gaming consoles. For these devices we have proposed a 3D API independent scenario that allows providing the synthetic signed contents by means of the avatar animation definition. The defined animation is rendered using the application developed for the available 3D API, but the module that generates the animation (responsible for the message comprehensibility) is the same for each the scenario. Therefore, this approach avoids repeating the intelligibility tests required for evaluating a synthesizer implementation. The only tests that have to be repeated for each device are those related to image resolution, contrast, and visualization fluidness that can be performed in a laboratory environment.

Providing Ubiquitous Access to Synthetic Sign Language Contents over Multiple Platforms

ACKNOWLEDGMENT The authors would like to thank the FPU-UAM program for its financial support and the FCNSE linguistic department’s staff for their evaluation of the synthetic messages.

REFERENCES Agu, E., Banerjee, K., Nilekar, S., Rekutin, O., & Kramer, D. (2005). A middleware architecture for mobile 3D graphics. In Distributed Computing Systems Workshops, (pp. 617-623). Ahmed, A. S., & Seong, D. S. K. (2006). SignWriting on mobile phones for the deaf. In Proceedings of the 3rd international conference on Mobile Technology, Applications & Systems, (pp. 1-7). New York. Bangham, A., Cox, S., Elliot, R., Glauert, J., & Marshall, I. (2000). Capture, Animation, Storage and Transmission - an Overview of the ViSiCAST Project. In IEE Seminary on Speech and Language Processing for Disabled and Elderly People. Virtual Signing. doi:10.1049/ic:20000136 Bennati, P., Capasso, T., Giallombardo, F., Maggio, E., & Giorgi, R. (2006). Blue Sign Translator. Retrieved September 30, 2009, from http:// bluesign.dii.unisi.it Boukerche, A., & Pazzi, R. W. N. (2006). Remote rendering and streaming of progressive panoramas for mobile devices. Proceedings of the 14th annual ACM international conference on Multimedia, (pp. 691-694). New York. Cherniavsky, N., Cavender, A. C., Ladner, R. E., & Riskin, E. A. (2007). Variable frame rate for low power mobile sign language communication. In Proceedings of the 9th international conference on Computers and Accessibility, (pp. 163-170). New York.

Chittaro, L., Buttussi, F., & Nadalutti, D. (2006). MAge-AniM: a system for visual modeling of embodied agent animations and their replay on mobile devices. In Proceedings of the working conference on Advanced Visual Interfaces, (pp. 344-351). New York, Elliot, R., Glauert, J., Jennings, V., & Kennaway, R. (2004). An Overview of the SiGML Notation and SiGML Signing Software System. In Proceedings of Language Resources and Evaluation Conference, (pp. 98-104).Lisbon. da Rocha, A. C., & Pereira, G. (2001). Supporting Deaf Sign Languages in Written Form on the Web. The SignWriting Journal (Vol. 0). Association pour le Traitement Automatique des Langues. Hamilton, S. W. R. (1853). Lectures on Quaternions. Dublin, Ireland: Royal Irish Academy. Herrero Blanco, A. (2004a). A practical writing of sign languages. In Proceedings of Language Resources and Evaluation Conference, (pp. 3742).Lisbon. Herrero Blanco, A. (2004b). Una aproximación morfológica a las construcciones clasificatorias en la lengua de signos española. In ELUA (Vol. 18, pp. 151–167). Estudios de Lingüística. Hybrid (2006). Hybrid Rasteroid. Retrieved June 2, 2006 from http://www.hybrid.fi Irving, A., & Foulds, R. (2005). A parametric approach to sign language synthesis. Proceedings of the 7th international conference on Computers and Accessibility, (pp. 212-213). New York. ISO/IEC 14772-1:1997, I. (1997). Information technology -- Computer graphics and image processing -- The Virtual Reality Modeling Language (VRML) -- Part 1: Functional specification and UTF-8 encoding. In Geneva, Switzerland. ISO/IEC 19774:2005, (2005). Information technology -- Computer graphics and image processing -- Humanoid animation (H-Anim). Geneva, Switzerland.

97

Providing Ubiquitous Access to Synthetic Sign Language Contents over Multiple Platforms

ISO/IEC 19775:2004, I. (2004). Information technology -- Computer graphics and image processing -- Extensible 3D (X3D). Geneva, Switzerland. JSR-184 Expert Group. (2005). Mobile 3D Graphics API for Java 2 Micro Edition. Retrieved June 1, 2006 from http://www.jcp.org/ en/jsr/detail?id=184 Kennaway, R. (2002). Synthetic Animation of Deaf Signing Gestures. [International Gesture Workshop, London.]. Lecture Notes in Computer Science, 2298, 146–157. doi:10.1007/3540-47873-6_15 Kennaway, R., Elliot, R., Glauert, J., & Parsons, K. J. (2002). SiGML Document Type Definition (DTD). In ViSiCAST Project website. Retreived March 12, 2007, from http://www.visicast.cmp. uea.ac.uk/sigml/ Kennaway, R., Glauert, J., & Zwitserlood, I. (2007). Providing signed content on the Internet by synthesized animation. ACM Transactions on Computer-Human Interaction, 14, 1–29. doi:10.1145/1279700.1279705 Kimura, T., Katoh, M., Hayashi, A., Kanda, K., Hara, D., & Morimoto, K. (2008). Application System of the Universal Sign Code - Development of the Portable Sign Presenter. In Heidelberg (Ed.), Lecture Notes in Computer Science Vol. 5105/2008. (pp. 678-681). Liddell, S. K. (2003). Sources of Meaning in ASL Classifier Predicates. In Emmorey, K. (Ed.), Perspectives on Classifier Constructions in Sign Languages (pp. 199–220). Psichology Press. López, F., Tejedor, J., Garrido, J., & Colás, J. (2006). Use of a Hierarchical Skeleton for Spanish Sign Language 3D Representation over Mobile Devices. In Proceedings of INTERACCION, (pp. 565-568).

98

López-Colino, F., & Colás, J. (2009). HLSML, una notación de alto nivel para la descripción de mensajes sintéticos en LSE[HLSML, a high level notation for describing LSE synthetic signed messages]. In Proceedings of III National Conference on Spanish Sign Language, Madrid, Spain. Lukashev, D., Puresev, A., & Makhlushev, I. (2006). 3D Applications for 3G Mobile Phones: Design, Development, Resource Utilization. In 10th International Symposium on Consumer Electronics, (pp. 1-4). Megginson, D. (2004). Simple API for XML (SAX). Retrieved November 21, 2008 from www. saxproject.org Nadalutti, D., Chittaro, L., & Buttussi, F. (2006). Rendering of X3D content on mobile devices with OpenGL ES. Proceedings of the 11th international conference on 3D web technology, (pp. 19-26). New York. Shoemake, K. (1985). Animating rotation with quaternion curves. In Proceedings of the 12th annual conference on Computer graphics and interactive techniques, (pp. 245-254). New York Solina, F., Krapež, S., Jaklič, A., & Komac, V. (2001). Design and Management of Multimedia Information Systems. Chap 13: Multimedia Dictionary and Synthesis of SL (pp. 268-281). Idea Group Publishing. Sutton, V. (1974). SignWriting. Retrieved September 30, 2009, from http://www.signwriting.org/ Xu, X., Wang, X., Yao, L., Zhang, D., & Zhao, H. (2008). Avatars based Chinese Sign Language Synthesis System. In IEEE International Conference on Systems, Man and Cybernetics. (pp. 444-449). Zwiterslood, I., Verlinden, M., Ros, J., & van der Schoot, S. (2004). Synthetic Signing for the Deaf: eSIGN. In Proceedings of the Conference and Workshop on Assistive Technologies for Vision and Hearing Impairment, Granada, Spain.

Providing Ubiquitous Access to Synthetic Sign Language Contents over Multiple Platforms

KEY TERMS AND DEFINITIONS Remote Rendering: The 3D scene processing is carried out on a remote server and the results are sent as a video or an image to the client device which performs the visualization of the results. Remote Gesture Synthesis: The avatar animation definition process is carried out on a remote server, the obtained animation tracks are sent to the client using a pre-established format. Parametric Sign Language Synthesis: Generation of the synthetic signed message based on phonetic-like sign definitions, reducing the storage requirements and incrementing the linguistic flexibility. Modular Architecture: Distribution of a complex application into several and more simple units that can be run in different devices. The communication between these units is done using a predefined interface. Ubiquitous Sign Language Synthesis: Generation and access to artificial signed messages regardless the hardware and software resources of the client’s device and the quality of the network connection. Signing Avatar: Virtual anthropomorphic character used to represent signed messages. Its

animation can be manually defined, captured from a human signer, or parametrically described. HLSML: XML-based notation focused on signed message definition and its prosody; it does not require describing the phonology of the signs, as other notations do.

ENDNOTES 1



2



3



The Phonologic Parameters in SL are equivalent to the oral language phonology. Each Phonologic Parameter is composed by a set of Phonemes. The Phonologic Parameters comprise hand shape, hand orientation, location, plane, contact point, movement and non-hand parameter. The math libraries included in some mobile devices do not provide all the required features for the Gesture Synthesis module (e.g. floating point operations). The user only notices the different durations required for sign message generation when using different scenarios. In any case, no user could appreciate differences in the same signed message using different scenarios.

99

100

Chapter 7

The Impact of MIMO Communication on NonFrequency Selective Channels Performance Andreas Ahrens Hochschule Wismar, Germany César Benavente-Peces Universidad Politécnica de Madrid, Spain

ABSTRACT This chapter reviews the basic concepts of multiple-input multiple-output (MIMO) communication systems and analyses their performance within non-frequency selective channels. The MIMO system model is established and by applying the singular value decomposition (SVD) to the channel matrix, the whole MIMO system can be transformed into multiple single-input single-output (SISO) channels having unequal gains. In order to analyze the system performance, the quality criteria needed to calculate the error probability of M-ary QAM (Quadrature Amplitude Modulation) are briefly reviewed and used as reference to measure the improvements when applying different signal processing techniques. Bit and power allocation is a well-known technique that allows improvement in the bit-error rate (BER) by managing appropriately the different properties of the multiple SISO channels. It can be used to balance the BER’s in the multiple SISO channels when minimizing the overall BER. In order to compare the various results, the efficiency of fixed transmission modes is studied in this work regardless of the channel quality. It is demonstrated that only an appropriate number of MIMO layers should be activated when minimizing the overall BER under the constraints of a given fixed date rate. DOI: 10.4018/978-1-60960-042-6.ch007 Copyright © 2011, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited.

The Impact of MIMO Communication on Non-Frequency Selective Channels Performance

INTRODUCTION The need for high data rate communication systems has remarkable increased in the last years due to the rising demand of wideband services like video and TV especially in mobile applications. Nevertheless, several barriers must be broken to fulfill the requirements of future communication systems. A lot of efforts and research activity has been developed in order to give answers to the challenge of increasing the spectral efficiency and improving the bit-error rate (BER) with the constraint of finding a moderate complexity implementation. Various technologies, techniques and algorithms have been developed in the last decades and researchers are working at the definition of new standards and future communication systems. Adaptive modulation (AM) is a promising technique to increase the spectral efficiency of wireless transmission systems by adapting the signal parameters, such as modulation constellation or transmit power, dynamically to changing channel conditions (Zhou et al., 2005). However, in order to comply with the demand on increasing available data rates in particular in wireless technologies, systems with multiple transmit and receive antennas, also called MIMO systems (multiple-input multiple-output), have become indispensable and can be considered as an essential part of increasing both the achievable capacity and integrity of future generations of wireless systems (Kühn, 2006; Zheng and Tse, 2003). The well-known water-filling technique is virtually synonymous with adaptive modulation and it is used for maximizing the overall data rate. However, delay-critical applications, such as voice or streaming video transmissions, may require a certain fixed data rate. For these fixed-rate applications it is desirable to design algorithms, which minimize the overall BER at a given fixed data rate. Against this background, the novel contribution of this chapter is that we demonstrate the benefits of amalgamating a suitable choice of

activated MIMO layers and number of bits per symbol along with the appropriate allocation of the transmit power under the constraint of a given data throughput.

BACKGROUND There is no doubt about the key role that communication systems have in the information society. The various available technologies allow users to share, store and transmit information to others. The rising demands of new services especially broadband ones like video require the appropriate technologies to transmit and receive information with the expected quality. The demand for higher network capacity and for higher performance of wireless networks is enormous. There are two major challenges in the design of future wireless communication systems (i.e. in LTE, Long Term Evolution): increasing the spectral efficiency (channel capacity) and improving the link reliability (BER). MIMO Systems are able to improve the spectral efficiency significantly, and consequently MIMO plays a key role in many future wireless communication systems. MIMO technology has attracted a lot of attention in wireless communications, since it offers significant increases in data throughput and link range without additional bandwidth or transmit power. It achieves this by higher spectral efficiency (more bits per second per hertz of bandwidth) and link reliability or diversity (reduced fading). Because of these properties, MIMO is a hot topic in international wireless research. Multiple antennas techniques can be used for different objectives and the most common are beamforming and transmit/receive diversity. Diversity techniques provide some protection against channel fading and increase the system range. MIMO techniques are a different way in which multiple antennas are used in a communication system.

101

The Impact of MIMO Communication on Non-Frequency Selective Channels Performance

The signal propagating through the radio channel is disturbed by different effects. It suffers from fading, multipath effects, noise and interference from other users and systems. Diversity and coding are two well known techniques for combating fading. Diversity is a technique that provides the receiver with several replicas of the transmitted signal which can be processed appropriately in order to combat against fading and interference and are able to improve the link quality. There are different ways in which diversity techniques are applied (time diversity and frequency diversity). In the last years, the use of spatial diversity has become widely used due to the advantages this technique offers. One of the main properties is that it can be applied without losing spectral efficiency. The use of multiple antennas at the receive end is called receive diversity and has been studied in depth (Jakes, 1974). Space-time coding (Tarokh et al., 1998; Alamouti, 1998; Hochwald and Marzetta, 2000) is a technique that is being used in mobile communications that uses multiple antennas at the transmit side in combination with appropriate digital signal processing and coding algorithms and is attracting a lot of research activity. The use of multiple antennas at both the transmitter and receiver ends (the well-known multipleinput multiple-output (MIMO) technology) allows the use of techniques and algorithms to achieve high data rates and decrease the BER (Telatar, 1999). MIMO techniques take advantage of multipath to improve system performance. These techniques are known as spatial multiplexing (Bolcskei et al., 2002) or BLAST (Bell Laboratories Layered Space-Time, Foschini, 1996; Foschini and Gans, 1998) and provide a significant increase in spectral efficiency (channel capacity). The idea underlying the MIMO techniques is that the transmit and receive antenna signal’s are processed in such a way that the BER and the data rate of the communication is improved. Most of the past research activity on MIMO wireless communication techniques have been

102

focused on narrowband channels. Nevertheless, spatial diversity broadband MIMO channels offer higher capacity and frequency diversity due to delay spread. The use of OFDM (orthogonal frequency division multiplexing, LeFloch et al., 1995) is being widely used because it significantly reduces the complexity at both the transmitter and receiver ends in wideband communications. The combination of MIMO techniques and OFDM modulation (Bolcskei and Paulraj, 2000) is being adopted in different standards because their promising properties.

FROM SISO TO MIMO SYSTEMS This section describes the evolution from singleinput single-output (SISO) systems, through SIMO (single-input multiple-output) and MISO (multiple-input single-output), to multiple-input multiple-output ones. The SISO channel is composed of one antenna at the transmitter side and one antenna at the receiver side. It is the conventional communication channel and constitutes the basic radio channel access mode. Equipping the transmitter and/or the receiver with more than one transmit antenna leads to some extensions which are able to outperform the BER performance of a wireless link significantly. The single-input multiple-output (SIMO) channel is composed of one antenna at the transmitter side and several antennas at the receiver side. This solution has long be considered as the classical solution for outperforming the overall quality by processing different replicas of the transmitted signal appropriately. Another solution to improve system performance is the use of a multiple-input single-output (MISO) channel which is composed of multiple antennas at the transmitter side and one antenna at the receiver side, motivated by Wittneben’s pioneering work (Wittneben, 1991). Unfortunately, no improvements in the spectral efficiency are possible with all these setups.

The Impact of MIMO Communication on Non-Frequency Selective Channels Performance

The use of SIMO/MISO techniques improves the link robustness but they are not able to increase the data rate. However, this drawback can be compensated by equipping the transmitter as well as the receiver with different antennas. The resulting MIMO channel provides the ability to improve the data rate and BER.

MIMO AND LTE 3GPP Long Term Evolution (LTE) technology is a broadband system based upon TCP/IP. LTE provides significant performance improvement. A remarkable advantage is the easy integration with internet networks making the convergence between mobile and fixed simpler. MIMO systems have become a key element of LTE (Long Term Evolution) in order to achieve larger throughput and spectral efficiency. MIMO means using multiple antennas at transmitter and receiver sides and appropriate signal processing. The competitiveness of UMTS in the next years is based on the achievement of higher data-rates, low-latency and packet optimized radio access technology. This is because LTE technology is being investigated and it is still under definition and standardization. A variety of technologies, techniques and algorithms are involved in its development. LTE technology uses new multiple access schemes on the air interface: OFDMA (Orthogonal Frequency Division Multiple Access) in the downlink and SC-FDMA (Single Carrier Frequency Division Multiple Access) in the uplink. Spatial multiplexing allows transmitting simultaneously different data streams through the same downlink. The number of data streams that can be transmitted in parallel over the MIMO channel is given by the minimum number of antennas at both sides, the transmitter as well as the receiver side and is finally limited by the rank of the channels matrix as it is shown below. It is remarkable that especially under frequency-selective channel conditions, spatial multiplexing techniques

makes the receivers very complex, and in order to avoid this problem it is typically combined with Orthogonal Frequency-Division Multiplexing (OFDM) or with Orthogonal Frequency Division Multiple Access (OFDMA) modulation, where the problems created by the multi-path channel can be handled efficiently. The IEEE 802.16e standard incorporates MIMO-OFDMA. The IEEE 802.11n standard, which is expected to be finalized soon, recommends MIMO-OFDM. MIMO is also planned to be used in mobile radio telephone standards such as recent 3GPP and 3GPP2 standards. In 3GPP, High-Speed Packet Access plus (HSPA+) and LTE standards use MIMO techniques to improve system performance.

MIMO SYSTEM MODEL AND QUALITY CRITERIA When considering a non-frequency selective SDM (space division multiplexing) MIMO link composed of nT transmit and nR receive antennas, the system is modeled by u = H ⋅ c + w.

(1)

In (1), u is the (nR×1) received vector, c is the (nT×1) transmitted signal vector containing the complex input symbols and w is the (nR×1) vector of the additive, white Gaussian noise (AWGN) having a variance of U R2 for both the real and imaginary parts. Furthermore, we assume that the coefficients of the (nR × nT ) channel matrix H are independently Rayleigh distributed with equal variance and that the number of transmit antennas equals the number of receive antennas. The interference between the different antenna’s data streams, which is introduced by the off-diagonal elements of the channel matrix H, requires appropriate signal processing techniques. Common strategies for separating the data streams are linear equalization at the receiver side or

103

The Impact of MIMO Communication on Non-Frequency Selective Channels Performance

linear pre-equalization at the transmitter side, if channel state information is available. Unfortunately, linear equalization suffers from noise enhancement and linear pre-equalization of the transmit signal from an increase in the transmit power. Both schemes only offer poor power efficiency (Fischer, 2002). Therefore, other signal processing strategies have attracted a lot of interest. Another popular technique is based on the singular value decomposition (SVD) of the system matrix H (Haykin, 1991), which can be written as H = S ⋅ V ⋅ DH , where S and D H are unitary matrices and V is a real-valued diagonal matrix of the positive square roots of the eigenvalues of the matrix H H H that are commonly sorted in descending order. Throughout this chapter, the transpose and conjugate transpose (Hermitian) of D are denoted by DT and DH, respectively. The SDM MIMO data vector c is now multiplied by the matrix D before transmission. In turn, the receiver multiplies the received vector u by the matrix SH. Thereby neither the transmit power nor the noise power is enhanced. The overall transmission relationship is defined as  . y = S H (H ⋅ D ⋅ c + w ) = V ⋅ c + w

(2)

Here, the channel matrix H is transformed into independent, non-interfering layers (multiple SISO channels) having unequal gains. Figure 1 shows the layer-specific block diagram, where  refers to the layer and k to the transmitted data block. The transmitted data symbol c, k is weighted by corresponding positive square root of the eigenvalues of the matrix H H H , i.e., ξ ,k  ,k is added. and finally the noise term w In general, the quality of data transmission can be informally assessed by using the signal-to-noise ratio (SNR) at the detector’s input defined by the half vertical eye opening and the noise power per quadrature component according to

104

Figure 1. Resulting system model per MIMO layer  and per transmitted data block k

=

( Half vertical eye opening)2 Noise Power

=

 

2

  

2

U A  U R 

, (3)

which is often used as a quality parameter (Ahrens and Lange, 2007). The relationship between the signal-to-noise ratio  = U A2 / U R2 and the biterror probability evaluated for AWGN channels and M-ary Quadrature Amplitude Modulation (QAM) is given by (Kalet, 1987; Proakis, 2000) 

PBER



    2 1 − 1  erfc    . =     2  log 2( M )  M   

(4)

When applying the proposed system structure, the SVD-based equalization leads to different eye openings per activated MIMO layer  and per transmitted symbol block k according to U A( , k ) = ξ , k ⋅ U s  ,

(5)

where U s  denotes the half-level transmit amplitude assuming M  -ary QAM and ξ ,k represents the positive square roots of the eigenvalues of the matrix H H H (Figure 1). The probability density function of the layer-specific weighting factors are analyzed in Figure 2 ( nT = nR = 4 ).

The Impact of MIMO Communication on Non-Frequency Selective Channels Performance

Figure 2. PDF (probability density function) of the layer-specific amplitudes ξ   = 1, 2, , 4 )

(with

by min(nT , nR ) . Therein it is worth noting that with the aid of powerful non-linear near Maximum Likelihood (ML) sphere decoders it is possible to separate nR > nT number of layers (Hanzo and Keller, 2006). The bit-error probability per MIMO layer  and transmitted symbol block k after SVD is given by (Ahrens and Lange, 2007)

( , k ) BER

P

Together with the noise power per quadrature component, the SNR per MIMO layer becomes



( , k )

=

2

  

U A( , k )  U R2

= ξ , k

2

   

U s   

U R2

.

(6)

expressed as (Forney et al., 1984; Kalet, 1989) 2 2 U ( M  − 1). 3 s

(7)

Combining (6) and (7), the layer-specific SNR results in 

( , k )

= ξ , k

Ps  3 . 2 ( M  − 1) U R2

(8)

Using the parallel transmission over L ≤ min(nT , nR ) MIMO layers, the overall mean transmit power becomes Ps =

1 M

) erfc 

ξ , k U s   . ⋅  2 U  R  

log 2( M  )

(9)

The resulting average bit-error probability per transmitted symbol block k assuming different QAM constellation sizes per activated MIMO layer results in (k ) PBER =

1

∑

L ν =1

L

log 2( M ν )

∑ log (M  =1

2



( , k ) ) PBER .

(10)

Considering QAM constellations, the average transmit power Ps  per MIMO layer  may be

Ps  =

=

(

2 1−

∑

L

 =1

Ps  , where

When considering time-variant channel conditions, rather than an AWGN channel, the BER can be derived by considering the different transmission block SNRs. Assuming that the transmit power is uniformly distributed over the number of activated MIMO layers, i.e., Ps  = Ps / L , the half-level transmit amplitude U s  per activated MIMO layer results in Us  =

3 Ps 2 L ( M  − 1)

.

(11)

The signal-to-noise ratio per SDM MIMO data block k and MIMO-layer  , defined in (6), results together with (11) in

the number of readily separable layers is limited

105

The Impact of MIMO Communication on Non-Frequency Selective Channels Performance

 ( , k ) = ξ , k

Ps Es 3 3 = ξ , k , 2 2 L ( M  − 1) U R L ( M  − 1) N 0

(12)

with Ps U

2 R

Es . N0 / 2

=

(13)

Finally, the BER per activated MIMO layer  and transmitted symbol block k is given by:

( , k ) BER

P

=

(

2 1−

1 M

) erfc

log 2( M  )

      

3 ξ , k



Es   .  2 L ( M  − 1) N 0  

(14) The resulting average bit-error probability per transmitted symbol block k assuming different QAM constellation sizes is obtained as (k ) BER

P

      3 ξ , k   E 2 L  1   s  ,  erfc  = ∑ 1 −   2 L ( M − 1) N   R  =1  M    0    

(15) with L

R = ∑ log 2( M  ) ,

(16)

 =1

describing the number of transmitted bits per data block.

BIT- AND POWER ALLOCATION Existing bit loading and transmit power allocation techniques are often optimized for maintaining both a fixed transmit power and a fixed target bit-error rate while attempting to maximize the

106

Table 1. Investigated transmission modes transmitting 8 bit/s/Hz over non-frequency selective channels layer 1

layer 2

layer 3

layer 4

256

0

0

0

64

4

0

0

16

16

0

0

16

4

4

0

4

4

4

4

overall data-rate. The well-known water-filling technique is virtually synonymous with adaptive modulation and it is used for maximizing the overall data rate. However, delay-critical applications, such as voice or video streaming transmissions, may require a certain fixed data rate to provide an appropriate QoS (Quality of Service). For these fixed-rate applications it is desirable to design algorithms, which minimize the bit-error rate (BER) at a given fixed data rate. In order to transmit at a fixed data rate while maintaining the best possible integrity, i.e. bit-error rate, an appropriate number of MIMO layers has to be used, which depends on the specific transmission mode, as detailed in Table 1. In general, the BER per SDM MIMO data vector is dominated by the specific transmission mode and the characteristics of the singular values, resulting in different BERs for the different QAM configurations in Table 1. An optimized adaptive scheme would now use the particular transmission mode that results in the lowest BER for each SDM MIMO data vector. This would lead to different transmission modes for different SDM MIMO data vectors and a moderate signaling overhead would result. However, in order to reduce the signaling overhead further, fixed transmission modes are used in this contribution regardless of the channel quality.

The Impact of MIMO Communication on Non-Frequency Selective Channels Performance

Figure 3. Resulting system model per MIMO layer  and transmitted data block k including MIMOlayer PA

( , k ) PA



=

2

  

( , k )   U PA  2 R

U

= p , k ⋅

3 ξ , k

Es = p , k ⋅  (  , k ) . L ( M  − 1) N 0

(18)

Using (4) and (18), along with the MIMO detector’s input noise power, the resultant BER per MIMO layer and transmitted data block can be calculated according to

ADAPTIVE POWER ALLOCATION In systems, where channel state information is available at the transmitter side, the knowledge about how the symbols are attenuated by the channel can be used to adapt the transmit parameters. Power allocation can be used to balance the bit-error probabilities in the activated MIMO layers. Adaptive Power Allocation (PA) has been widely investigated in the literature (Ahrens and Lange, 2007; Krongold et al. 2000; Jang and Lee, 2003). The BER of the uncoded MIMO system is dominated by the specific layer having the smallest SNR. As a remedy, a MIMO transmit PA scheme is required for minimizing the overall BER under the constraint of a limited total MIMO transmit power. The proposed PA scheme scales the half-level transmit amplitude U s  of the  th

( , k ) PBER = PA

(

2 1−

1 M

) erfc

log 2( M  )

       

 

3 p , k ξ  , k

Es   .  2 L ( M  − 1) N 0  

(19) Finally, the BER per data block results in (k ) BER PA

P

     3 p, k ξ , k E  2 L  1   s   erfc  = ∑ 1 −  .  2 L ( M − 1) N  R  =1  M     0   

(20)

The aim of the forthcoming discussions is now the determination of the values

p, k for the

MIMO layer by the factor p, k . This results in

activated MIMO layers. A common strategy is to use the Lagrange multiplier method in order to find the optimal

a transmit amplitude of U s 

value of

p, k for the QAM

symbol of the MIMO transmit data vector transmitted at the time k over the MIMO layer  (Figure 3). Applying MIMO-layer PA, the half vertical eye opening per MIMO layer  and data block k becomes ( , k ) U PA =

p , k ⋅ ξ  , k ⋅ U s  .

(17)

Now the signal-to-noise ratio, defined in (12), is changed to

p, k for each MIMO layer  and each

data block k , which often leads to excessivecomplexity optimization problems (Ahrens and Lange, 2007). Therefore, suboptimal power allocation strategies having a lower complexity are of common interest (Ahrens and Lange, 2007; Park and Lee, 2004). A natural choice is to opt for a PA scheme, which results in an identical signal-to-noise ratio

( , k ) PA = equal

2

   

( , k )  U PA  equal  

U R2

= p , k ⋅  (  , k )

(21)

107

The Impact of MIMO Communication on Non-Frequency Selective Channels Performance

for all activated MIMO layers per data block k , i.e. in ( , k ) PA = constant equal

 = 1, 2, , L .

(22)

Figure 4. BER without PA when using the transmission modes introduced in Table 1 and transmitting 8 bit/s/Hz over non-frequency selective channels ( +  adaptive choice of the transmission mode and ∗  optimal bitloading)

The power to be allocated to each activated MIMO layer  and transmitted data block k can be shown to be calculated as follows (Ahrens and Lange, 2007): p , k =

( M  − 1) ⋅ ξ , k

L L

∑

( M ν −1)

ν =1

.

(23)

ξν, k

Taking (17), (23) and (11) into account, for each symbol of the transmitted MIMO symbol vector the same half vertical eye opening of ( , k ) U PA = equal

3 Ps

p , k ⋅ ξ  , k ⋅ U s  =

L

2∑ ν =1

( M ν −1)



ξν , k

(24)

can be guaranteed (  = 1, , L ), i. e., ( , k ) U PA = constant equal

 = 1, 2, , L . (25)

When assuming an identical detector input noise variance for each channel output symbol, the above-mentioned equal quality scenario (22) is encountered, i.e., ( , k ) PA equal



=

2  (  , k )  U PA equal   

U R2

=

Es

N0

L

∑ ν =1

3 ( M ν −1)

.

(26)

ξν , k

Analysing (26) for a given SDM MIMO data block, nearly the same BER can be achieved on all activated MIMO layer. However, taking the

108

time-variant nature of the transmission channel into account, different BERs arise for different SDM MIMO data blocks. Therefore, the BER of the MIMO system is mainly dominated by the data blocks having the lowest SNR’s. In order to overcome this problem, the number of transmit or receive antennas has to be increased or coding over the different data blocks should be used (Ahrens, Kühn and Weber, 2008).

RESULTS In this contribution fixed transmission modes are used regardless of the channel quality. Assuming predefined transmission modes, a fixed data rate can be guaranteed. The obtained BER curves are depicted in Figure 4 for the different QAM constellation sizes and MIMO configurations of Table 1, when transmitting at a bandwidth efficiency of 8 bit/s/Hz within a given bandwidth. Assuming a uniform distribution of the transmit power over the number of activated MIMO layers, it turns out that not all MIMO layers have to be activated in order to achieve the best BERs. More

The Impact of MIMO Communication on Non-Frequency Selective Channels Performance

Figure 5. Probability of choosing a specific transmission mode by using optimal bit loading

explicitly, our goal is to find that specific combination of the QAM mode and the number of MIMO layers, which gives the best possible BER performance at a given fixed bit/s/Hz bandwidth efficiency. The Es / N 0 value required by each scheme at BER 10-4 was extracted from Figure 4 and the best configuration is emphasized in bold in Table 1. Allowing a low signaling overhead, an adaptive choice of the transmission modes can be carried out. Since the BER per SDM MIMO data block is dominated by the chosen transmission mode and the distribution of the singular values, the different transmission modes, as depicted in Table 1, lead to different BERs per SDM MIMO data block. An adaptive modulation scheme would now use the specific transmission mode that results in the lowest BER per SDM MIMO data block. As depicted in Figure 4, the adaptive choice of the transmission mode outperforms our fixed modes at the cost of a small signaling overhead. However, the lowest BERs can only be achieved by using bit auction procedures leading to a high signaling overhead. Analyzing the probability of choosing a specific transmission mode by using optimal bit loading, as depicted in Figure 5, it turns out that at moderate SNR only an appropriate number of MIMO layers have to be activated,

Figure 6. BER with PA (dotted line) and without PA (solid line) when using the transmission modes introduced in Table 1 and transmitting 8 bit/s/Hz over non-frequency selective channels

e.g. the (16,4,4,0) QAM configurations. The expression lg(⋅) is considered to be the short form of log10(⋅) . Surprisingly, a neglectable loss in the BER results at low SNR by using the best fixed transmission mode even though different layerspecific BERs arise. Further improvements are possible by taking the adaptive allocation of the transmit power into account. Furthermore, from Figure 6 we see that unequal PA is only effective in conjunction with the optimum number of MIMO layers. Using all MIMO layers, our PA scheme would assign much of the total transmit power to the specific symbol positions per data block having the smallest singular values and hence the overall performance would deteriorate.

IMPLEMENTATION ASPECTS The implementation of a MIMO system involves considering several issues that affects the performance of the system. In this section some implementation aspects (as hardware and algorithms) are considered. The use of efficient hardware architectures and computationally efficient algorithms

109

The Impact of MIMO Communication on Non-Frequency Selective Channels Performance

are required to a real time implementation of the system. FPGA (field-programmable gate array) and DSP (digital signal processors) devices play a key role in the development of these systems where the computation exigencies are so high, as well as moderate power consumption is required. Multiple-input multiple-output (MIMO) wireless technology uses multiple antennas at the transmitter and receiver to produce significant capacity gains over single-input single-output (SISO) systems using the same bandwidth and transmit power. It has been shown that the capacity of a MIMO system increases linearly with the number of antennas at both the transmitter as well as the receiver side in the presence of a scattering-rich environment. This will ensure that the signals at the antennas in the array are sufficiently uncorrelated with each other. This is where antenna design comes in for MIMO systems. The SVD can be implemented in different ways and on various hardware solutions and architectures. DSP are powerful devices offering high processing speed as well as additional resources for computations and some flexibility (co-processors, input/output ports, interconnectivity). FPGA devices have increased in the last years their capabilities (including speed processing and calculation resources) and functionalities while maintaining the inherent flexibility that allow building appropriate processing architectures. FPGAs have evolved from being flexible logic design platforms to signal processing and powerful engines. They are now an essential component of digital signal processing systems due to their flexibility and real-time processing capabilities. Current advances in FPGA design technology have provided high-speed processing in a compact footprint, while maintaining the flexibility and programmability capabilities of signal processing systems. FPGA devices are popular for their highspeed, intensive computing, parallelism, common reconfigurable applications as fast Fourier transform (FFT), finite impulse response (FIR) filtering and other multiply-accumulate operations.

110

Although these devices have great processing capabilities, efficient algorithms are required to reduce the number of operations, computation complexity and the resources required for their implementation. The benefits are reducing the power consumption and the possibility of using cheaper devices or allowing the allocation of additional functionalities on the same device. Some research works (Benavente-Peces et al., 2008; Benavente-Peces et al., 2009) have demonstrated that the use of the CORDIC algorithm to implement the SVD decomposition is a computationally efficient solution with acceptable performance degradation. Once the channel is characterized and the SVD matrices are obtained, the information must be sent back to the transmitter in order to perform the pre-coding operation. This means that some overhead should be assumed in the communication. The channel conditions vary with time and that information is continuously updated. The updating frequency depends on the change rate in the channel parameters and must be study a priori. The use of vector quantization and other strategies may be used to diminish the overhead.

FUTURE RESEARCH Wireless MIMO systems currently are a very dynamic field of research. Therefore there are a lot of interesting topics open for further work, which can be derived from the results of this chapter: Bit and power loading for coded MIMO systems are of great practical interest as shown in (Ahrens and Kühn, 2007; Ahrens and Benavente-Peces, 2009, Ahrens and Benavente-Peces, 2009a) or (Ahrens, Kühn and Weber, 2008). Additionally, non-frequency selective MIMO links have reached a lot of research. By contrast, frequency selective MIMO links require substantial further research (Ahrens and Benavente-Peces, 2009). In addition to the currently very popular wireless MIMO systems, multiple-input multiple-

The Impact of MIMO Communication on Non-Frequency Selective Channels Performance

output channels are observed in a variety of transmission links and network parts. Currently, fixed access networks are mainly constituted of multi-pair copper cables which contain a number of wire pairs. These copper cables then by their nature compose a MIMO channel. Cables or cable binders can be treated as MIMO channels and crosstalk relations are taken into account for example by crosstalk equalization schemes (Lange and Ahrens, 2005) and they are exploited in the case of dynamic spectrum management (Kerpez, 2002; Cioffi and Mohseni, 2004), which is currently gaining a growing practical interest. Crosstalk in multi-pair copper cables has long been seen as a possible application for MIMO techniques (van Etten, 1975; Ahrens and Lange, 2009). Another important type of a fixed network medium is the optical fibre, where single- and multi-mode fibres are distinguished. In particular optical multi-mode fibres guide light of different modes: Therefore the multi-mode fibre can be interpreted as a MIMO channel. This is rarely discussed in the literature, since in almost all relevant transmission cases nowadays single-mode fibres are deployed due to their superior transmission performance. However, in case of fibre deployment over short distances (e.g. in-house networks for high data rates) the multi-mode fibre technology eventually may become more popular again (due to the cost advantages of the multi-mode fibre and the associated components compared to single-mode fibres and optical components). If in case of practical relevance - in particular in the optical fibre case - results from the currently very dynamic research on wireless and wireline MIMO techniques can be adopted and adapted to the fixed optical line problems, synergy effects and considerable improvements are expected.

CONCLUSION This chapter is focused in the analysis of MIMO systems in narrowband communications and using

fixed transmission modes. The modeling process allows describing the MIMO system as its decomposition into multiple SISO channels which defines the MIMO layers. The choice of the number of bits per symbol and the number of MIMO layers combined with the appropriate allocation of the transmit power substantially affects the performance of a MIMO system obtaining a noticeable improvement in the BER. After studying, analyzing and characterizing the uncoded system performance, it turns out that not all MIMO layers have to be activated in order to achieve the best BERs. On the other hand we must remark that MIMO systems are a key element of LTE in order to achieve larger throughput and spectral efficiency and additional analysis is required for broadband services.

REFERENCES Ahrens, A., & Benavente-Peces, C. (2009). Modulation-Mode and Power Assignment in SVD-assisted Broadband MIMO Systems. In International Conference on Wireless Information Networks and Systems (pp. 83-88), Milano, Italy, 6.-10. July. Ahrens, A., & Benavente-Peces, C. (2009a). Modulation-Mode and Power Assignment for Broadband MIMO-BICM Schemes. In IEEE 20th Personal, Indoor and Mobile Radio Communications Symposium (PIMRC), Tokio, Japan, 13.-16. September. Ahrens, A., & Kühn, V. (2007). Analysis of SVD-Aided, Iteratively Detected Spatial Division Multiplexing using EXIT Charts. In 12th International OFDM-Workshop (pp. 271-275), Hamburg, 29.-30. August. Ahrens, A., Kühn, V., & Weber, T. (2008). Iterative Detection for Spatial Multiplexing with Adaptive Power Allocation. In 7th International Conference on Source and Channel Coding, Ulm, 14.-16. January.

111

The Impact of MIMO Communication on Non-Frequency Selective Channels Performance

Ahrens, A., Lange, C. (2007). Transmit Power Allocation in SVD Equalized Multicarrier Systems. International Journal of Electronics and Communications (pp. 51-61), 61 (1). Ahrens, A., & Lange, C. (2009). Iteratively Detected MIMO-OFDM Twisted Pair Transmission Schemes. In Communications in Computer and Information Science (pp. 281–293). Heidelberg: Springer. Alamouti, S. M. (1998). A simple Transmit Diversity Technique for Wireless Communications, IEEE Journal on Selected Areas in Communications (pp. 1451-1458), 16 (8). Benavente-Peces, C., Ahrens, A., Arriero-Encinas, L., & Lange, C. (2008). Implementation Analysis of SVD for Modulation-Mode and Power Assignment in MIMO Systems. In International Conference on Communication Systems and Networks (CSN 2008). Palma de Mallorca, Spain, 01.-03. September. Benavente-Peces, C., Ahrens, A., Pardo-Martín, J. M., & Ortega-González, F. J. (2009). Fixed Point SVD Computation Error Characterization and Performance Losses in MIMO Systems. In International Conference on Wireless Information Networks and Systems (pp. 91-94), Milano, Italy, 6.-10. July. Bolcskei, H., Gesbert, D., & Paulraj, A. J. (2002). On the capacity of OFDM-based Spatial Multiplexing Systems. IEEE Transactions on Commuications (pp. 225-234), 50 (2). Bolcskei, H., & Paulraj, A. J. (2000). SpaceFrequency Coded Broadband OFDM Systems. In Proc. IEEE WCNC-2000, Chicago, IL. Cioffi, J. M., & Mohseni, M. (2004). Dynamic Spectrum Management - A Methodology for providing significantly higher Broadband Capacity to the Users. Telektronikk (pp. 126-137), 4.

112

Fischer, R. F. H. (2002). Precoding and Signal Shaping for Digital Transmission. New York: John Wiley. doi:10.1002/0471439002 Forney, G.D., Gallager, R.G., Lang, G.R., Longstaff, F.M., & Qureshi, S.U. (1984). Efficient Modulation for Band-Limited Channels. IEEE Journal on Selected Areas in Communications (pp. 632-647), 2 (5). Foschini, G. J. (1996). Layered space-time Architecture for Wireless Communication in a Fading Environment when using multi-element Antennas. Bell Labs Tech. Journal (pp. 41-59), autumn. Foschini, G. J., & Gans, M. J. (1998). On limits of Wireless Communications in a fading Environment when using multiple Antennas. Wireless Personal Communications (pp. 311-335), 6. Hanzo, L., & Keller, T. (2006). OFDM and MC-CDMA. New York: John Wiley. doi:10.1002/9780470031384 Haykin, S. S. (1991). Adaptive Filter Theory. Englewood Cliffs, N. J.: Prentice Hall. Hochwald, B. M., & Marzetta, T. L. (2000). Unitary space-time Modulation for multipleantenna Communications in Rayleigh at Fading. IEEE Transactions on Information Theory (pp. 543-564), 46 (2). Jakes, W. C. (1974). Microwave mobile communications. New York: Wiley. Jang, J., & Lee, K.B. (2003). Transmit Power Adaptation for Multiuser OFDM Systems. IEEE Journal on Selected Areas in Communications (pp. 171-178), 21(2). Kalet, I. (1987). Optimization of Linearly Equalized QAM. IEEE Transactions on Communications (pp. 1234-1236), 35(11). Kalet, I. (1989). The Multitone Channel. IEEE Transactions on Communications (pp. 119-124), 37 (2).

The Impact of MIMO Communication on Non-Frequency Selective Channels Performance

Kerpez, K. J. (2002). DSL Spectrum Management Standard. IEEE Communications Magazine (pp. 116-123). 40(11). Krongold, B.S., Ramchandran, K., & Jones, D. L. (2000). Computationally Efficient Optimal Power Allocation Algorithms for Multicarrier Communications Systems. IEEE Transactions on Communications (pp. 23-27), 48 (1).

Wittneben, A. (1991). Base Station Modulation Diversity for Digtal SIMULCAST. In: IEEE Vehicular Technology Conference (VTC-Spring), pp. 848-853 Zheng, L., Tse, D.N.T. (2003). Diversity and Multiplexing: A Fundamental Tradeoff in Multiple-Antenna Channels. IEEE Transactions on Information Theory (pp. 1073-1096), 49 (5).

Kühn, V. (2006). Wireless Communications over MIMO Channels - Applications to CDMA and Multiple Antenna Systems. Chichester: Wiley. doi:10.1002/0470034602

Zhou, Z., Vucetic, B., Dohler, M., & Li, Y. (2005). MIMO Systems with Adaptive Modulation. IEEE Transactions on Vehicular Technology (pp. 10731096), 54(3).

Lange, C., & Ahrens, A. (2005). Far-End Crosstalk Equalization in Multi-Pair Symmetric Copper Cable Transmission. In International Conference on Advances in the Internet, Processing, Systems, and Interdisciplinary Research, Pescara, Italy.

KEY TERMS AND DEFINITIONS

LeFloch, B., Alard, M., & Berrou, C. (1995). Coded Orthogonal Frequency Division Multiplex. Proceedings of IEEE (pp. 982-996), 83(6). Park, C.S., & Lee, K.B. (2004). Transmit Power Allocation for BER Performance Improvement in Multicarrier Systems. IEEE Transactions on Communications (pp. 1658-1663), 52 (10). Proakis, J. G. (2000). Digital Communications. Boston: McGraw-Hill. Tarokh, V., Seshadri, N., & Calderbank, A. R. (1998). Space-time Codes for High Data Rate Wireless Communication: Performance Criterion and Code Construction, IEEE Transactions on Information Theory (pp. 744-765), 44(2). Telatar, I. E. (1999). Capacity of multi-antenna Gaussian Channels. European Transactions on Telecommunications (pp. 585-595), 10. Van Etten, W. (1975). An Optimum Linear Receiver for Multiple Channel Digital Transmission Systems. IEEE Transactions on Communications (pp. 828-834), 23 (8).

AM (Adaptive Modulation): Adaptive Modulation refers to the capability to perform a modulation and power adjustment in the communication (wireless) system to prevent channel conditions and disturbances changing along time (known as fading). This technique increases the spectral efficiency of wireless transmission systems by adapting the signal parameters appropriately. MIMO (Multiple-Input Multiple-Output): It’s a technique used to increase channel capacity (spectral efficiency) by adapting the spatial dimension of the transmission channel. This technology incorporates at least two antennas at the transmitter side and at least two antennas at the receiver side. MIMO takes advantage from multipath to improve the system performance. PA (Power Allocation): It’s a technique used to distribute the total available power at the transmitter along the various antennas (or layers). SDM (Space Division Multiplexing): It’s a multiplexing technique in which physical separation of transmitting (antennas) is used to deliver simultaneously different data streams. SDM technique is an approach to MIMO systems and it improves capacity by increasing the number of

113

The Impact of MIMO Communication on Non-Frequency Selective Channels Performance

antennas in the fading channel. The most popular algorithm is the BLAST. BLAST (Bell Laboratories Layered SpaceTime): BLAST is a spectrally efficient algorithm applied to wireless communication which uses the spatial dimension (SDM) to transmit and receive different data streams using multiple antennas. There are various improvements of this algorithm

114

called V-BLAST (vertical BLAST) and D-BLAST (diagonal BLAST). Beamforming: It is a signal processing technique applied to multiple antennas system in order to obtain the required radiation pattern for each case in order to transmit most of the radiated power in a concrete direction (or directions).

115

Chapter 8

Node Localization in Ad-hoc Networks Zhonghai Wang Michigan Tech University, USA Seyed (Reza) Zekavat Michigan Tech University, USA

ABSTRACT This chapter introduces node localization techniques in ad-hoc networks including received signal strength (RSS), time-of-arrival (TOA) and direction-of-arrival (DOA). Wireless channels in ad-hoc networks can be categorized as LOS and NLOS. In LOS channels, the majority of localization techniques perform properly. However, in NLOS channels, the performance of these techniques reduces. Therefore, non-line-of-sight (NLOS) identification and mitigation techniques, and localization techniques for NLOS scenarios are briefly reviewed.

I. INTRODUCTION Node localization in ad-hoc networks have emerging applications in homeland security, law enforcement, defense command and control, emergency services, and traffic alert. These systems promise to considerably reduce the society’s vulnerabilities

to catastrophic events and improve the quality of life. Typical examples include emergency 911 (E911) (Mayorga, C.L.F. et al. 2007), tracking a fire fighter (Ingram, S.J. et al. 2004), battlefield command and control (Venkatesh, S. et al. 2008), vehicle safety, etc. These techniques also help to enhance wireless routing (Stojmenovic, I. 2002), (Karimi, H.A. et al. 2001) and resource alloca-

DOI: 10.4018/978-1-60960-042-6.ch008 Copyright © 2011, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited.

Node Localization in Ad-hoc Networks

tion (Haddad, E. C. et al. 2003) performance in ad-hoc networks. In E911, the cell phone starting an emergency call is localized with an error that is in the order of 100 meters (for 95 percent of calls). In a building on fire, it is hard for fire fighters to find their position and their way out. While if fire fighters’ 3-dimentional (3-D) position can be tracked by commander or fire fighters themselves, then they can make their way out by themselves or under the commander’s guidance. In battlefield, soldiers’ position information allows commanders to maintain central command procedure. In addition, it allows them to monitor the health of soldiers, and provide support in emergency situations. Accordingly, many localization methods have been proposed. Examples include GPS (global positioning system) plus communication (Karimi, H.A. et al. 2001), received signal strength indication (RSSI) (Bulusu, N. et al. 2000), (Konrad L. et al. 2005), time-of-arrival (TOA) fusion (Goud, P. et al. 1991), time difference-of-arrival (TDOA) fusion (Gillette, M. D. et al. 2008), directionof-arrival (DOA) fusion (Girod, L. et al. 2006), (Niculescu, D. et al. 2003), joint TOA-DOA estimation (Tong, H. et al. 2007), and multi-node TOA-DOA fusion (Wang, Z. et al. 2009), etc. In all of these systems, usually two types of nodes are available: (1) Target nodes which their position should be found, and (2) Base nodes which enable finding the position of target nodes. In some techniques, such as RSSI, target nodes’ position is calculated by themselves and can be reported to base nodes. In some techniques such as TOA fusion and joint DOA-TOA estimation, base nodes are in charge of finding the position of target nodes. These systems are usually categorized as active remote positioning system, because target nodes actively contribute in the process of remote positioning. RSSI incorporates the signal strength of multiple beacon-signals generated by base nodes at the target node. Here, the base nodes are also called anchor nodes as their positions are assumed

116

known. Two RSSI localization approaches have been proposed: (a) based on the communication of the anchor nodes and target nodes: If a target node communicates with an anchor node, then the target node would be in the coverage area of the anchor node. When the target node communicates with multiple anchor-nodes, the target node is localized at the centroid (Bulusu, N. et al. 2000) or weighted centroid (Shen, X. et al. 2005) of these anchor-nodes with respect to a reference point; and (b) based on mapping the measured signal strength set (signature) into a premade received signal strength map of the environment (Konrad L. et al. 2005). Here, the environment should be perfectly known, and a signal strength map should be available. Techniques that are based on range measurement consist of two categories: (a) those that measure the distances between target node and base nodes to determine multiple circles with their center at base nodes via TOA estimation (Goud, P. et al. 1991). In this case, the target node is localized at the crossing point of these circles; and (b) those that are based on measuring the range differences between pairs of base nodes and the target node (TDOA) to determine multiple hyperbolas with focal points at the two base nodes (Gillette, M. D. et al. 2008). The target node is localized at the crossing point of multiple hyperbolas. In the techniques based on angle measurement, either the angles of the target node with respect to base nodes is measured by base nodes via DOA estimation (Girod, L. et al. 2006) or the angle of base nodes (beacon nodes) with respect to the target node is measured at the target node (Niculescu, D. et al. 2003). The target node is localized at the crossing point of multiple lines determined by the base nodes position and the measured angles. In this technique, the nodes that are in charge of DOA estimation should be equipped with antenna arrays. Antenna arrays are an array of antennas, e.g., dipole antennas, which are usually located at a fixed distance from each other.

Node Localization in Ad-hoc Networks

The MDS (multidimensional scaling) localization methods (Yi, S. et al. 2004), (Jose, A. et al. 2006) computes the relative position of nodes in a system. Here, no base node is needed: the relative position of nodes in the system is attained via the distance information between nodes pairs. If multiple base nodes (or anchor nodes) are available, the localization method can be used to compute the absolute nodes position. A shortcoming of this method is its high computational complexity. Thus, it is not suitable for localizing fast moving nodes. Hybrid techniques integrate different kinds of measurements to localize targets. The measurement can be two or more of the following: range, angle and signal strength, e.g., multi-node TOADOA fusion in (Wang, Z. et al. 2009). To localize a target node, parameters such as received signal strength, distance (based on TOA), and/or DOA are measured. Next, these parameters are fused (processed) to estimate the target node’s position. The received signal is usually sensitive to the wireless channel between target nodes and base nodes. Hence, the wireless channel impacts the signal received at those nodes. In the localization process, the impact of the wireless channel should be considered. The most significant impact of wireless channel on localization is the error originated by the lack of availability of line-of-sight (LOS) between base and target nodes. If the base node does not realize that LOS scenario is not available, considerable localization error would occur in the measurement made by one or some nodes. These erroneous measurements can negatively impact the fusion results. Hence, base nodes should identify non-LOS (NLOS) situations to avoid those errors. In this case, localization techniques developed for NLOS scenarios should be implemented. Available localization techniques that are designed to deal with NLOS error can be grouped into two categories. One category finds NLOS base node and exclude them from the localization process, i.e., only use the information from LOS base nodes to localize target node (Xiong, L. 1998), (Chen, P.C. 1999). Another category uses

the statistics of measurements in NLOS scenarios to mitigate the NLOS impact on localization performance (Güvenç, I. et al. 2008), (Cong, L. et al. 2005). If LOS base nodes are not enough to perform localization or the statistics of the measurement in NLOS scenario is not available, these available techniques would not perform, and the method that directly use the measurements in NLOS is needed. An example of these techniques has been discussed in Subsection V.3. Section II introduces the wireless channel and its impact on localization performance. Section III introduces the parameter measurement and available localization techniques. Section IV compares these techniques and Section V introduces methods that discriminate LOS and NLOS, and localization methods in NLOS scenarios. Section VI concludes the chapter.

II. WIRELESS CHANNELS AND LOCALIZATION Node localization techniques developed for ad-hoc networks that are based on signal strength, range and angle are impacted by the wireless channel between base and target nodes. The wireless channel usually consists of a LOS path and multiple reflected (NLOS) paths between source and destination. Here, we assume that the transmitted signal from the source is X (t ) , and the channel between the source and the destination is linear time invariant (LTI). In addition, here, we assume the availability of an antenna array at the nodes to enable DOA estimation. It should be noted that in general wireless channels are dynamic; however, within a short time period, they can be well modeled as LTI. In other words, these channels are considered quasi-static. The impulse response of an LTI channel is modeled by K −1

h(τ ) = ∑ k =0 αk iΦ(θk )iδ(τ − τk ).

(1)

117

Node Localization in Ad-hoc Networks

In Equation (1), K is the number of paths between the source and the destination, including one direct path (k = 0) and multiple reflected paths (k = 1, 2, …, K-1); ak , θk and τk are the attenuation, angle-of-arrival and the time delay of the kth path, respectively; Φ(θk ) is the corresponding array response vector determined by the receive antenna array geometry and the angle-of-arrival when an antenna array is installed at the destination, and it is taken Φ(θn ) = 1 if Omni-directional antennas are installed at the destination. Note that in Equation (1), τ 0 , τ 0 ≠ 0 , is the time delay of the LOS path; in addition, we assume τ 0 < τ1 < τ2 <  < τK −1 . Ignoring the noise effect, the received signal at the destination corresponds to R(t ) = X (t ) ∗ h(t ) =

∑

K −1

a i(θk )iX (t − τk ). k =0 k



(2)

If a 0 is large enough to be detected, we categorize the channel as the LOS channel in the study of node localization. In the LOS channel, the time delay of the first detectable path corresponds to the true distance (R) between source and destination, i.e., R= c⋅τ0

(3)

In Equation (3), c is the speed of light, τ 0 is the time that the signal travels from the source to the destination. Hence, in LOS channels, the distance between source and destination can be computed by measuring the first arrival’s time delay. In LOS channel, the angle of the first detectable path ( θ0 ) is the true angle of the source with respect to the destination. Therefore, the angle of the source with respect to the destination can be calculated by measuring the angle of the first arrival.

118

If a 0 is too small to be detected, the channel is categorized as NLOS. In this case, the first detectable path is from a reflector. Thus, the first arrival’s time delay and angle would be form a reflector, and accordingly the localization would involve with major error. Most localization techniques (e.g., (Bulusu, N. et al. 2000), (Goud, P. et al. 1991), (Gillette, M. D. et al. 2008), (Niculescu, D. et al. 2003), (Tong, H. et al. 2007), (Wang, Z. et al. 2009)) are designed assuming a LOS between source and destination is available. In some applications, such as open areas (e.g., desert in battlefield applications), the LOS between the transmitter and receiver is almost available in all scenarios. Thus, in these scenarios, NLOS identification is not an important issue. However, there are many other scenarios, such as in urban or indoor area, in which the probability that the LOS between the transmitter and receiver is obstructed is high. Thus, it is important to identify NLOS conditions. Many techniques have been proposed to identify NLOS scenarios (Venkatesh, S. et al. 2007), (Venkatraman, S. et al. 2002), (Wang, Z. et al. 2009) and maintain localization in NLOS scenarios (Güvenç, I. et al. 2008), (Cong, L. et al. 05). These techniques will be investigated in Section V.

III. LOCALIZATION TECHNIQUES This chapter introduces four categories of measurements for the available localization techniques in ad-hoc networks that include: (1) received signal strength (RSS), (2) Range, (3) Angle, and (4) Hybrid measurements. This section reviews the theoretical foundations of these techniques.

Node Localization in Ad-hoc Networks

III.1 Parameter Measurement in Localization In this subsection, we review the measurement techniques applied in node localization in ad-hoc networks. These techniques include received signal strength measurement, range measurement and angle measurement.

a) Received Signal Strength Measurement The received signal power ( Pr ) in a receiver at the measurement point can be calculated by Pr =

PG G t at re (4π / λ)µ

.

(4)

Where, Pt is the transmitter outage power,

Gat is the transmit antenna gain, Gar is the re-

ceiver antenna gain, Gre is the gain from antenna output to the measurement point in the receiver, d is the distance between transmitter and receiver, λ is the carrier wavelength, and µ (usually larger than 2) is the fading parameter determined by the channel. Based on Equation (4), RSS is determined by the transmitting power, transmitting and receiving antennas’ gains, the receiver structure, the distance between transmitter and receiver, the carrier wavelength and the channel fading parameter, and it is independent of the signal bandwidth. A real antenna beam pattern is not ideally Omni directional. Thus, the power in one direction might be higher than the other direction. In addition, the channels between multiple base station transmitters and the target receivers are not the same; hence, when the RSS is mapped into the distance, error may occur. RSS can be measured in the intermediate frequency (IF) stage before the IF amplifier or in the baseband signal chain before the baseband ampli-

fier in zero-IF systems. In a complex sampling system, when I and Q samples (samI and samQ) of the received signal are evaluated, the power of the received signal can be calculated using

(5)

Where E(â‹Ž) is the expectation operator and Rin is the input load. Fading effects occur because of channel variations. Thus, the received signal amplitude is not constant. To achieve a better power measurement, in practice, (assuming the signal to be a mean ergodic process), a large number of samples are collected and sample mean is applied to calculate the expectation of Equation (5).

b) TOA Measurement TOA estimation techniques are divided into two categories: round-trip and single-trip methods. Radar and wireless local positioning system (WLPS) (Tong, H. et al. 2007) apply the roundtrip method. Radar stands for RAdio Detection And Ranging. It is a device that transmits a burst of radio energy and receives its reflections from objects and processes them to detect the desired targets and find their range. In addition to detection of the target and finding its range, advanced radars are capable of extracting other information such as speed, size and nature of the target. For example they can easily discriminate a fighter jet from a Boeing 747. Radar computes the TOA of the round trip via pulse detection, and incorporates the detected TOA to find the target range. In Radar, the target does not contribute in the process of localization. Thus, radar is called a passive remote positioning system. In WLPS, a base station broadcasts a directsequence spread spectrum (DSSS) inquiry signal. When a target node receives the inquiry signal, it generates and transmits a DSSS response with a fixed and known delay back to the base node. The round-trip TOA and the know delay are incorpo-

119

Node Localization in Ad-hoc Networks

rated at the base node to find the position of the target. DSSS transmission is mainly suggested for wireless environments to create path diversity at the receiver and improve its performance. Here, the target contributes in the process of positioning. Thus, WLPS is called an active target remote positioning system. GPS uses single-trip method from the satellite to the ground receivers for localization (Elliott D. Kaplan, 1996). The timings of all GPS satellites are synchronized with a clock in master control station (MCS) located at Schriever AFB in Colorado. Each satellite broadcasts ranging code and navigation data including its position and the time that the ranging code is transmitted. When a GPS receiver receives the signal from a satellite, a TOA estimation technique is applied to find the TOA of the ranging code and compare it to a local clock to find the time delay from the GPS satellite to receiver. A hybrid single trip TOA measurement technique is presented in (Fukuju, Y. et al. 2003). Here, radio frequency (RF) and ultrasonic signals are used to measure the TOA of ultrasonic signal between source and target nodes. The received RF signal enables the TOA counter and the received ultrasonic signal ends the counter. In air, the ultrasonic speed (about 340 m/s) is much lower than that of RF (about 3×108 m/s). The speed difference is applied to measure TOA. The accuracy of range measurement based on RF signal is limited by the signal bandwidth (Peebles, Peyton Z. 1998), measurement techniques, signal-to-noise ratio (SNR), and numberof-reflectors in the area. For a single measurement (no filtering or average method is applied) based on cross correlation similar to the one used in GPS, or based on pulse detection such as the one used in RADAR, ignoring the impact of SNR, the range resolution corresponds to P = E(

120

samI2 + samQ2 Rin

).

(6)

Where, Tres is the TOA resolution, Beff is the effective bandwidth, ρres is the range resolution and c is the light speed. Because c ≈ 3×108 m/s, if high ranging accuracy is requested (e.g., 5 meters), then higher bandwidth should be used (the required bandwidth corresponding to 5 meters is 60 MHz). A low bandwidth signal always generates low range accuracy. Note that in the hybrid RF – ultrasonic technique, the ranging accuracy is not sensitive to the RF signal bandwidth. In this hybrid system, the ranging error 1 c Tres = , Pres = . corresponds to: Beff Beff (∆ρ)

(7)

In Equation (7), V is the ultrasonic speed in air, and V = 340 m/s. ∆p = V i∆TOA. includes two parts, one is generated by RF signal and the other is generated by ultrasonic signal. If the RF signal bandwidth is 1MHz, the TOA measurement error generated by RF signal is about 1μs. The corresponding ranging error is 340m/s × 10-6 s = 0.34mm. This error is small in any ad-hoc network application and can be ignored.

c) DOA Measurement In ad-hoc networks, antenna array and electronically steerable passive array radiator (ESPAR) antennas are usually used to measure DOA, because their size is small, and they are cost effective. Reflector antennas are capable of DOA estimation as well, but their size is large and they need mechanical scanning. This limits their applications in ad-hoc networks. In a linear antenna array, delay-and-sum (DAS) (Liberti, J. C. et al. 1999), multiple signal classification (MUSIC) (Schmidt, R.O. 1986) and root-MUSIC (Barabell, A.J. 1983) are usually applied to measure DOA. In DAS, the single received by each antenna element is assigned a complex weight to change its phase. The weight is determined by the assumed DOA of the signal, antenna array parameters

Node Localization in Ad-hoc Networks

(element distance, number of elements) and the signal carrier wavelength. Then the delayed signals are summed, and its output power is calculated. When the assumed DOA matches the true one, the output power of the weighted sum reaches its maximum value. Hence, when the maximum output power of the weighted sum is observed, the corresponding assumed DOA is taken as the received signal DOA. In MUSIC, the received signal of an antenna array is modeled by ∆TOA

(8)

In (8), X is the received signal vector, A is the array vector of the antenna array determined by the DOA (θ) of the signal, antenna array parameters and the signal carrier wavelength. Antenna array parameters and the carrier wavelength are fixed; hence, A is only a function of DOA (θ). In addition, W is the received noise vector. The eigenvectors of the covariance matrix of X is calculated and the eigenvectors corresponding to the smallest eigenvalues are selected and used to construct a matrix E. Essentially, E represents the noise components. MUSIC exploits the orthogonality of noise and signal components: noise components are represented by E (matrix is denoted with bold capital letter), and the signal components received from the angle θ are represented by A(θ). Thus, MUSIC estimates the DOA of received signal by finding the peaks of the MUSIC spectrum X = A(θ)iS + W .

(9)

While the root MUSIC (Barabell, A.J. 1983) directly finds the root of the polynomial Pmu (θ) =

1 . A (θ)iEiE∗ iA(θ) ∗

(10)

Where, M is the total number of antenna eleM +1

ments, D(z ) = ∑ l =−m +1 al z −1. , al =

∑

m −n =l

Bmn

is the element on the mth row and nth column of matrix B, Bmn , E is defined in (9), and B = E × E* When the roots of the polynomial are calculated, the corresponding DOA can be calculated, for z = exp(−j 2πd sin θ / λ) . For details, see (Barabell, A.J. 1983). ESPAR antenna (Ohira, T. et al. 2001) is made by a single active element surrounded by multiple parasitic elements loaded with variable reactance. By controlling the reactance of these parasitic elements, the ESPAR antenna beamforming is implemented, and DOA is measured via electronic beam scanning. DOA estimation accuracy is a function of the technique used, the SNR, number of elements in the array, the channel structure (i.e., the availability of multiple paths), the mutual coupling between antenna elements which itself is a function of the distance between neighboring elements, and the calibration of antenna array. It should be noted that in general the RF components connected to the antenna receivers does not operate fully equivalent. Thus, the phase and amplitude of signals received through each antenna element may vary from one RF element to another. This effect can highly reduce the DOA estimation performance. DOA complexity varies with the selected method. The complexity and performance of DAS (MUSIC) is very low (high). Recently, a new fused DASMUSIC high performance and low complexity technique (suitable for ad-hoc networks) has been introduced (Zekavat, S.A. 2007).

III. 2 Localization Techniques Based on Signal Strength Measurement In the localization techniques based on signal strength measurement, anchor (reference) nodes broadcast beacon signals periodically. The position of anchor nodes is usually kept fixed or known

121

Node Localization in Ad-hoc Networks

(for example by installing a GPS on them). The beacon signal includes the anchor-node ID and the power level if the nodes’ transmitting power can be controlled. Hence, the mobile target node can identify that the received signal is coming from which anchor-node and with what power. This technique can be implemented using two methods: connectivity fusion, and signature mapping. In connectivity fusion (Bulusu, N. et al. 2000), mobile target node listens to all anchor-nodes for a while. In this time period, each anchor-node broadcasts Nsent times. If the mobile target node receives Nrece times from an anchor-node, and Nrece is larger than a predetermined threshold Nthresh, then it is considered that the mobile target node is in the coverage area of the anchor-node (the mobile target node is connected with the anchornode). The coverage area of the anchor-node is in a circle centered at the anchor-node position and its radius is the maximum distance that the anchor-node and the mobile target node can communicate. When it is determined that the mobile target node is connected to multiple anchor-nodes, the mobile target node is localized at the centroid of these anchor-nodes position. For example, the mobile target node is connected to anchor-nodes 1 to L, then the target node is localized at (x, y), z = exp(−j 2πd sin θ / λ)

(11)

In (11), (x, y) is the estimated mobile target node position and (xl, yl) is the anchor-node l’s position in Cartesian coordinates. The approach introduced in (11) does not need the availability of LOS as the only important point is the capability of communication of target node and anchor nodes. Connectivity fusion does not need a priori information of the environment. But if anchor nodes are not uniformly distributed, the target node position would be biased toward the anchor node with the larger distance, because in (11), each anchor node’s position is assigned the same weight. To alleviate the biasing problem, a weighted

122

connectivity fusion is presented in (Shen, X. et al. 2005). Here, the RSS and the channel fading parameter (μ) are applied to assign fusion weight to each base node position, assuming the anchor nodes’ transmitting power is known. In this case, x=

L L 1 1 x , y = ∑ l =1 yl . ∑ l =1 l L L

(12)

In (12), Pl is the RSS of the lth anchor node at the target node, and μ is the channel fading parameter. In a rich scattering environment, NLOS reduces the accuracy of this technique. The technique presented in Equation (12) is sensitive to the availability of LOS, because if only NLOS is available, the received power, and the parameter μ in Equation (12) would change. However, the approach of Equation (12) would lead to a better performance compared to that of (11) when LOS is available. Next, we introduce signature mapping technique which needs a priori information of the environment. This technique is more suitable for indoor or urban areas where a rich scattering environment is available. In signature mapping (Konrad L. et al. 2005), a map (data base) of the environment RSS is prepared. In this map, a set of RSS (reference signature) of signals from multiple anchor-nodes (a signature) is measured at each reference point (xm, ym), m is the reference point index, and 1 ≤ m ≤ M, M is the total number of reference points. In other words, the signature (sm) of the reference point m, located at (xm, ym), corresponds to a set of RSS (RSS1,m, RSS2,m, … RSSl,m, … RSSL,m), i.e., sm = [RSS1,m RSS2,m … RSSl,m … RSSL,m]. The average of multiple measurements in a time period t at a reference point is taken as the reference signature of the reference point. One signature is shown in Figure 1. A mobile target node in the network listens to anchor-nodes’ beacon for a while (usually the same time period t that is used for shaping the RSS map is considered), then the RSS from the

Node Localization in Ad-hoc Networks

Figure 1. Signature mapping localization

number of anchor-nodes and they are uniformly distributed, using a fixed number α we can generate a reasonable accuracy. But if the number of anchor-nodes is small, and sparse, and they are not uniformly distributed in the ad-hoc network, the accuracy of this technique would be lower (the estimated target node position would be biased toward the anchor node with a larger distance). A reference signature that is the closest to the received signature can be found using x=

same anchor-node is averaged, and forms a received signature (sr = [RSS1,r RSS2,r … RSSl,r … RSSL,r]). The Manhattan distance1 or Euclidean distance2 can be applied to calculate the distance between the received signature (sr) and the reference signature (sm). If the Manhattan distance is applied, the distance between two signatures sr and sm is L

∑ x= L∑

µ

l =1 L l =1

pl x l µ

pl

L

∑ ,y = L∑

µ

l =1 L l =1

pl yl µ

pl

.

(13)

In (13), L is total number of anchor nodes in the network, RSSl,r and RSSl,m are the received signal strengths from anchor node l for the received signature and the mth reference signature, respectively. There are several ways to localize a mobile target node. A simple way is to find α (α is a predetermined number) reference signatures that are closer to the received signature than others. Then localize the mobile target node at the centroid of the positions of these reference signatures, i.e., L

d (sr , sm ) = ∑ l =1 RSSl ,r − RSSl ,m .

(14)

Where, (xm, ym) is the reference signature m’s position. In an ad-hoc network, if there are a large

a a 1 1 x , y = ∑ m =1 x m , ym . ∑ m =1 m α α

(15)

Then, a set of reference signatures that are closer to the received signature are selected for the calculation of mobile target node position. The set of close reference signatures is determined using s ∗ = arg min1≤m ≤M d (r, sm ).

(16)

In (16), c is a constant and can be empirically determined to be within 1.1 to 1.2 (Konrad L. et al. 2005). The reference signatures selection scheme in (15) enhances the localization performance by eliminating the problem of using the approach in Equation (14) which is the estimated position that is biased toward the node with the larger distance. The localization methods based on RSS measurement do not need extra hardware or complex software. The RSS can be measured in a receiver using a simple calculation. In addition, there is no requirement on the signal bandwidth, and can be easily implemented on the available protocols such as Bluetooth (Rodriguez, M. et al. 2005) and 802.11 (Harder, A. et al. 2005). The accuracy of RSSI localization depends on the number of anchor nodes (reference points), and the distance between anchor nodes (reference points). This category of localization methods would require a large number of anchor nodes (reference points) if higher accuracy is required. The signature mapping method is capable of local-

123

Node Localization in Ad-hoc Networks

izing in NLOS scenarios because the signatures in all LOS and NLOS areas are prepared in priori for a fixed position of anchor nodes.

Figure 2. TOA fusion localization

III.3 Localization Techniques Based on Range Measurement Localization techniques based on range measurement are composed of two methods. The first one is based on the measurement of the distance between mobile target node and multiple anchornodes, which is implemented via measuring TOA. Hence, we call it TOA fusion. The other is based on the measurement of the difference in the distance between mobile target node and multiple pairs of anchor-nodes, which is implemented via measuring the TDOA. Hence, we call it TDOA fusion. In both TOA and TDOA estimation techniques, it is assumed that LOS is available. In TOA fusion, the position of anchor-nodes is assumed to be known. The TOA of the signal traveling from mobile target node to anchor node (or from anchor node to mobile target node) can be directly measured using an active target node based on round-trip measurements (Tong, H. et al. 2007) or can be indirectly measured as this is the case in GPS (see details below). In general, three anchor-nodes would be required to localize a mobile target node in TOA fusion if TOA is directly measured similar to the one shown in Figure 2. When there are two anchor nodes, for example anchor node 1 and 2 in Figure 2, two circles are formed (represented by arcs 1 and 2). There are two crossing points on the two circles, one is at the target node position and the other is not. To eliminate the ambiguity, at least a third anchor node is needed, such as anchor node n in Figure 2. Based on Figure 2, the mobile target node’s position (x, y) is determined by the equation set d (r , sm ) d (r , s k )

124

≤ c.

(17)

In GPS, the TOA is indirectly measured via an extra parameter that is the user receiver’s local time tu, which is also an unknown parameter (Elliott D. Kaplan, 1996). Hence, in GPS there are four unknown parameters, the coordinate (x, y, z) of the mobile target node and the user receiver’s local time tu. Therefore, at least four equations are required to calculate these four unknown parameters. In other words, in GPS, user receivers should receive signals from at least four satellites to localize themselves. Figure 3 represents the TDOA fusion technique. In this method, the difference of the range between the target node and the pair of anchornodes (e.g., anchor nodes 1 and 2) is computed by measuring the TDOA. The calculated range difference and the two anchor nodes position determine a hyperbola (i.e., the one represented by Curve 1), and the target node is selected on the hyperbola. When three or more anchor-nodes are available, multiple hyperbolas are achievable, and the target node would be on the crossing point of these hyperbolas. In Figure 3, when the range differences (ΔR12, ΔR1n, ΔR2n) are calculated via TDOA measurement, we have

Node Localization in Ad-hoc Networks

Figure 3. TDOA fusion localization

R1 = (x 1(b ) − x )2 + (y1(b ) − y )2 , R2 = (x 2(b ) − x )2 + (y2(b ) − y )2 ,

(18)

Rn = (x n(b ) − x )2 + (yn(b ) − y )2 . In (18), the unknown parameters x and y can be calculated by solving the equation set.

III.4 Localization Techniques Based on Angle Measurement Due to the development of RF circuit and digital circuit, nowadays, smart antennas can be implemented at small scale static nodes (Hyeon, S. et al. 2009), and mobile nodes (Lucke, O. et al. 2007) in ad-hoc networks. The received signal’s DOA can be measured at these nodes that are equipped with smart antennas. There are two kinds of DOA fusion techniques. One is to measure DOA of the mobile target node at two or more anchor nodes, and then fuse these measured DOA to localize the target node. This method requires that all anchor nodes are aligned in the same direction, for example by using a compass (Niculescu, D. et al. 2003). The other DOA fusion method measures the relative DOA of two or more anchor nodes with respect to the target node’s reference direc-

tion at the mobile target node (Niculescu, D. et al. 2003), and then fuses these relative DOA to localize the target node itself. In the localization techniques based on the first kind of DOA measurement, anchor nodes are equipped with antenna arrays, and mobile target nodes are equipped with Omni directional antenna. Two or more anchor nodes measure the DOA of the mobile target node’s with respect to a direction that these anchor nodes are aligned with. The mobile target node is assumed to be on the sight lines of all anchor nodes to the mobile target node. Here, it is assumed that LOS is available. Hence, when multiple anchor nodes can observe the mobile target node, the target node is localized on the crossing point of the line of sights from anchor nodes to the target node. In Figure 4, all anchor nodes are aligned in the direction of east (i.e., x axis). Based on this figure, the target node position is determined by the equation set ∆R21 = (x 2(b ) − x )2 + (y2(b ) − y )2 , − (x 1(b ) − x )2 + (y1(b ) − y )2 , ∆Rn 1 = (x n(b ) − x )2 + (yn(b ) − y )2 − (x 1(b ) − x )2 + (y1(b ) − y )2 ,



(19)

∆Rn 2 = (x n(b ) − x )2 + (yn(b ) − y )2 − (x 2(b ) − x )2 + (y2(b ) − y )2 . In (19), θ1 = tan−1 tan

−1

y1(b ) − y

x 1(b ) − x yn(b)) − y . x n(b ) − x

, θ2 = tan−1

y2(b ) − y x 2(b ) − x

, θn =

,

θ1 and θ2 are measured angles, and θn , (x 1(b ), y1(b ) )

and (x 2(b ), y2(b ) ) are known anchor nodes’ position. Solving (19), the target node position (x, y) would be calculated.

125

Node Localization in Ad-hoc Networks

Figure 4. Absolute DOA fusion localization

Figure 5. Relative DOA fusion localization

In the methods that are based on relative DOA measurement, the target nodes would be equipped with antenna arrays. Here, it is not required to install antenna array at anchor nodes. They measure the DOA of anchor nodes with respect to its reference direction (e.g., the black thick arrow line in Figure 5). The reference direction can be the base line or the bearing line of the antenna array. This line is arbitrary and no direction alignment is needed. When DOA of each anchor node is measured, we calculate (x n(b ), yn(b ) ) , α1 and α2

with respect to the x axis shown in Figure 5. In addition, α0 , α1 and α2 are the estimated anchor nodes angles with respect to the target node reference direction. Using (20), we calculate αn and α12 , which correspond to α2n

(21)

Assuming that ψ is the set of anchor nodes that can be seen by the target node, calculating y (b ) − y

y (b ) − y

Based on Figure 5, we have αn and α12 = α2 − α1 When we estimate the target node position α2n = αn − α2 , we can calculate the angles of anchor nodes with respect to the x axis (the thinner arrow line in Figure 5), which corresponds to

up to α12 , an error function can be constructed

(x , y )

(20)

αn −1 n

,

The mobile target node position (x, y) is calculated via minimizing the error in (22), which is

, is

SE = ∑ i , j ∈ψ (aij − aij )2 .

a1′ = a 0 + a1 = tan

−1

In (20), a2′ = a 0 + a2 = tan

y (b ) − y 1

x 1(b ) − x y (b ) − y −1

an′ = a 0 + an = tan

2

(b ) 2 (b ) n (b ) n

x −x −y −1 y

. x −x the offset of the target node reference direction

126

a12 = a2 − a1 = a2′ − a1′ = tan−1 a 2n = an − a 2 = an′ − a 2′ = tan

2

x 2(b ) − x y (b ) − y −1 n

x n(b ) − x

− tan−1 − tan

1 , x 1(b ) − x y (b ) − y −1 2

x 2(b ) − x

.

(22)

(23)

Equations (17) – (19) and (23) are all nonlinear equations; thus, extended Kalman filter (EKF) or linear iteration technique can be applied to find the

Node Localization in Ad-hoc Networks

solution of (17) – (19) (Elliott D. Kaplan, 1996). Equation (23) can be solved using a linear iteration technique (Niculescu, D. et al. 2003). While, EKF is a precise technique, it involves with a risk of divergence. When EKF is applied to address nonlinear equations, its divergence should be avoided, in which, considerable error is generated. To avoid divergence, all system parameters, the process and measurement noise covariance matrix should be known (Liu, H.F. et al. 2003). In addition, the initial state should be as precise as possible (Shankararaman, R. et al. 2006).

III.5 Localization Techniques Based on Hybrid Measurement Multiple types of measurements can be integrated to achieve better localization performance. Techniques that are based on the integration of multiple types of measurements are called hybrid. In hybrid TOA-RSSI (Wann, C.D. et al. 2007), RSSI and a pre-established path loss model are applied to discriminate LOS and NLOS, and generate weights for TOA measurement in TOA fusion. Larger weights are assigned to TOAs that are generated by stronger signals (e.g., those received through LOS channel). These weights are generated by RSS. Applying these weights improves the localization accuracy in severe NLOS scenarios by decreasing the NLOS impact (smaller weights are assigned to weaker signals that are more likely from NLOS channel). In hybrid DOA-RSSI (Lee, J. et al. 2008) and joint TOA-DOA estimation (Tong, H. et al. 2007), both the angle and the range of a target node with respect to a base node are calculated via DOA and TOA/RSS measurements. Accordingly, the base node would be capable of localizing target nodes independently, while in DOA or TOA fusion scenarios at least three base nodes are needed for the process of localization. It should be mentioned that whenever multiple base nodes are available, DOA and TOA information measured by multiple base nodes could be ap-

plied across those base nodes to further improve the localization accuracy (Wang, Z. et al. 2009).

III.6 Localization without Base Node All the localization techniques introduced in subsections III.2-5 require base nodes, whose position can be computed prior to the localization of target nodes. In this sub-section, we introduce the localization method based on multidimensional scaling (MDS) technique (Yi, S. et al. 2004), (Jose, A. et al. 2006). In this method, the relative position of a node with respect to other nodes is calculated using the distances between itself and its neighboring nodes. Base nodes are not necessary in this method, but if three (for 2-D localization) or more (for 3-D localization) base nodes are available, the absolute position (in the coordinate that base nodes are located) of the objective node can be calculated. It is assumed that node i is located at xi (xi=[xiyi] T ) and the distance between any pair of nodes i and j is measurable via measuring TOA. Now, the real distance between nodes i and j is (ˆ, x yˆ) = arg min(x ,y ) SE .

(24)

We represent the measurement of dij with pij. The node position is computed by minimizing the summation of the squared difference between the measured (pij) and calculated (dij) distances, i.e., dij = (x i − x j )T (x i − x j ).

(25)

In a real system, there might be one-hop and multi-hop communication, by assigning a weight (wij) to each distances difference, we achieve the weighted MDS as following. xT − xT ′ −x R ′ − xT ′

=

yT − yT ′ −yR ′ − yT ′

.

(26)

127

Node Localization in Ad-hoc Networks

Table 1. Localization techniques comparison Methods based on Base node(s)

MDS

Requirement

Performance

Hardware

Software

SNR

Bandwidth

Installation

Small scale Low/High

RSS

Low

Low

Low

Low

Low/High

Range

Low

High

Low

High

Low

Angle

High

High

High

Low

Low

Low

High

Low

High

Low

In (Jose, A. et al. 2006), it is found that by assigning 1 to one-hop and ¼ to two-hop and 0 to those more than two-hop distance measurements, better localization accuracy is achievable. The localization method based on MDS does not need the availability of base nodes; hence, its installation is simpler than other methods. But it should be noted that there are multiple unknowns in the minimization process (25) and (26); hence, the computational complexity of the localization method based on MDS is much higher than those method having base nodes.

3

Large scale Low

4

High High

Low low

5

accuracy is needed, a huge map (large data base) should be made. This is a time consuming process. A computational methodology presented in (Skidmore R.R. et al. 2004) can simplify the RSS map making process by simulating the RSS in a given area when the system model is available. The system model includes:

IV. LOCALIZATION TECHNIQUES COMPARISON

1. Environment Physical Model, including transmitters’ position, buildings and furniture position, size (length, width and height), material, etc. 2. Signal Propagation Model. 3. Instruments Model, including transmitting antenna’s parameters, transmitting power, receiving antenna’s parameters, etc.

Based on the description in Section III, we can generate Table 1 to compare the localization techniques based on RSS, range and angle measurements. There are many kinds of combinations of hybrid measurements; hence, we do not compare the hybrid techniques with those based on single kind of measurement. Table 1 represents that RSSI method needs minimum requirement across all localization techniques (excluding the installation of the signature mapping method). The localization accuracy of the signature mapping is higher when a high resolution (fine grid) RSS map is created. A fine grid RSS map requires RSS measurement throughout the environment, i.e., the RSS should be measured at each grid point on the map. In a large scale application, if high localization

Using this method, an RSS map can be easily created if the accurate system model is available. But usually, we cannot attain accurate system model. If high accuracy is needed, onsite measured RSS map is still required. The installation of the connectivity fusion method only needs the availability of anchor nodes’ position, and it is much simpler. However, its localization accuracy decreases as the distance between anchor nodes increases. This technique is not suitable for applications in which anchor nodes are far away from each other. Localization techniques based on range measurement need complex ranging and localization software and high bandwidth. Its installation only needs anchor nodes position. Its accuracy is determined by the geometric dilution of precision

128

Node Localization in Ad-hoc Networks

(GDOP) and ranging error (Elliott D. Kaplan, 1996). GDOP is the impact of the nodes’ (including base/anchor nodes and target node) geometrical distribution on localization performance. A good anchor nodes and target nodes distribution (target nodes are surrounded by uniformly distributed anchor nodes) helps to achieve better performance like in GPS. In GPS, users are on the earth surface and the twenty four GPS satellites are uniformly distributed around the earth. The requirement of localization techniques based on angle measurement is the highest one within the three listed in Table 1. This kind of technique needs complex hardware (antenna array or ESPAR antenna) and implementation of DOA measurement and localization algorithms. As mentioned before although usually DOA estimation techniques such as MUSIC need high complexity in order to offer high performance; however, a fused DAS/MUSIC DOA technique such as the one in (Zekavat, S.A. 2007) can help reduce the complexity problem. In small scale applications, these methods can achieve high accuracy; while in large scale applications, their localization accuracy is low. The reason is that the GDOP of this kind of methods depends on angle measurement error, nodes distribution, and the distance between anchor nodes and target node (Dempster A.G. 2006). As the distance increases, the GDOP increases. The localization method based on MDS does not require the availability of base node. Its installation is much simpler than other method, but it can only provide the relative position of nodes. In the applications that absolute position is needed, base nodes are needed. A disadvantage of this method is its high computational complexity. Due to its complexity, this technique is not applicable to a system in which the relative speed of nodes is high. In NLOS scenario, the techniques discussed above usually generate considerable localization error except for the signature mapping method. While in real application, NLOS scenarios are inevitable; hence, LOS and NLOS should be

discriminated and techniques applicable in NLOS scenario should be designed.

V. LOS/NLOS DISCRIMINATION AND NLOS MITIGATION/LOCALIZATION A localization technique designed for LOS scenario would generate considerable error when LOS is not available, i.e., in NLOS scenario. The reason is that NLOS propagation increases RSS, TOA and DOA measurements errors (Güvenç, I. et al. 2008), (Li. X. 1998), and thus generates large localization error. To avoid the impact of NLOS on localization performance, a node should identify NLOS scenarios. In addition, techniques should be investigated to enable localization in NLOS scenarios. Accordingly, NLOS identification and NLOS mitigation methods have been investigated, and some NLOS localization techniques have been proposed. Here, we briefly review the available NLOS identification, NLOS mitigation, and NLOS localization techniques.

V.1 NLOS Identification To avoid the severe impact of NLOS, NLOS identification techniques have been proposed. Examples include methods based on the rootmean-squared delay spread (RDS) (Venkatesh, S. et al. 2007), the statistics of the measured range (Venkatraman, S. et al. 2002), and multi-antenna phase difference statistics (Wang, Z. et al. 2009). In the methods that are based on the RDS, the received ultra wideband (UWB) signal’s TOA and RSS are first calculated, and then the RDS is calculated and used to separate LOS and NLOS. Because UWB signal is needed, this method is not applicable to systems with wideband or narrowband signals. The method based on the statistics of the measured range, tests the normality of the measured range. If the signal is coming from LOS channel, the measured range should fall in normal or almost normal distribution; while

129

Node Localization in Ad-hoc Networks

if the signal is coming from NLOS channel, the measured range does not fall in normal distribution (Venkatraman, S. et al. 2002). This method involves with some latency as the full statistics of the estimated range should be computed. This requires considerable time. The method based on multi-antenna phase difference statistics applies the variance of the phase difference of the received signals from two co-installed synchronized receivers to discriminate LOS and NLOS. In NLOS conditions, the variance is large and it decreases as LOS power increases from zero. In LOS only condition, the variance is zero. A threshold based on application is selected to separate LOS and NLOS. The technique is not limited by signal bandwidth. In addition, its latency is small. In this method, only the full statistics of the received signal phase difference is needed, and that can be calculated in a within the time period that the channel remains unchanged (channel coherence time). This technique needs two co-installed synchronized receivers. However, it is a suitable technique for multi-input system applications, e.g., a smart antenna system, or the systems that are based on DOA estimation.

V.2 NLOS Mitigation The available NLOS mitigation techniques are divided into two categories. One category identifies the outliers (NLOS base nodes) and discards them, and use the information achieved via LOS base nodes’ to implement the localization (Xiong, L. 1998), (Chen, P.C. 1999). These mitigation techniques are only applicable in scenarios that the number of LOS base nodes is equal to or larger than the necessary number of base node to perform localization. In addition, the NLOS measurements are discarded, which may enhance the localization performance if used properly in the localization process. The second category is to identify NLOS base nodes and calibrate the data achieved via these NLOS base nodes using the statistics achieved in

130

NLOS channel models, e.g., the TOA statistics achieved in (Alavi, B et al. 2006), and then apply the data achieved in LOS and the calibrated data captured in NLOS to implement the localization (Güvenç, I. et al. 2008), (Cong, L. et al. 2005). These methods need the statistics of the localization information (RSS, TOA or DOA) in NLOS scenario to be achieved prior to the localization. Thus, these methods are applicable to situations where a priori information is available. But in some applications such as localizing fire fighters in a burning building, the prior NLOS information might not be available. Hence, these methods cannot be applied to these applications. Accordingly, localization techniques that directly apply NLOS measurement(s) are required in scenarios that LOS base nodes are not enough to perform localization or the priori NLOS measurement statistical information is not available. This motivates our discussion in the next subsection.

V.3 Localization with NLOS Measurements In urban environments (e.g., cellular networks), the number of base nodes (base stations) is limited. In addition, in many scenarios, the LOS between base nodes and target nodes (mobile user) is abstracted by buildings and objects (Anisetti, M. et al. 2008). Thus, NLOS mitigation methods introduced in subsection V.2 are not applicable to these scenarios, as they may generate large amount of error. If a priori information of the environment is not available, the RSSI based method (e.g., the one presented in (Anisetti, M. et al. 2008)) would fail, as well. In these scenarios, methods that directly apply the NLOS measurements are needed. One solution based on bidirectional TOADOA measurement is presented in (Seow, C.K. et al. 2008). In this method, antenna arrays are equipped with both base and target nodes; geometrical NLOS identification is applied to determine whether LOS channel is available and find which pairs of TOA-DOA measurements are

Node Localization in Ad-hoc Networks

coming from the same reflector (shared reflector) in NLOS channel. In NLOS channel, if two or more shared reflectors are available, the target node can be localized at the crossing point of multiple lines that are constructed with the TOADOA measurement pairs as shown in Figure 6. In Figure 6, the line T′TB′′ is constructed with the NLOS TOA-DOA measurement pair x i,i ∈{1,,M } = mini, j ∈{1,,M } wij (dij − pij )2 . a n d (R1(BT ), θ1(BT ) ) . The position of point T′ in the base node’s local coordinate is determined by the measurement (R1(TB ), θ1(TB ) ) . In the rectangular coordinate, it is (R1(BT ), θ1(BT ) )

(27)

The position of point B′ in the target node’s local rectangular coordinate is xT = R1(BT ) icos(θ1(BT ) ), yT = R1(BT ) isin(θ1(BT ) ). (28) The quadrangle BB’TB’’ is a parallelogram; thus, the position of B’’ in the base node B’s local coordinate is: x B ′ = R1(TB ) icos(θ1(TB ) ), yB ′ = R1(TB ) isin(θ1(TB ) ). (29) With the positions of points T’ and B’’, the function determining the line T′TB′′ is calculated by x B ′′ = −x B ′′ , yB ′′ = −yB ′′ .

(30)

When multiple shared reflectors are available, we can achieve a set of functions that determine multiple lines on which the target node T is localized. Solving the set of functions, the target node is localized in the base node’s local coordinate. NLOS localization is still an important issue in ad-hoc networks node localization. Research

Figure 6. Localization with NLOS measurements

should be conducted to enable single measurement NLOS identification. The methods that do not rely on priori information and enable direct localization using NLOS measurements should receive more attention.

VI. CONCLUSION This chapter introduces localization techniques that are applicable to ad-hoc networks. We reviewed wireless channels in ad-hoc networks. The basic theory of various localization techniques was presented and compared. We also introduced NLOS identification techniques and study the impact of NLOS on localization performance. In addition, some recent NLOS mitigation techniques were introduced.

REFERENCES Alavi, B., & Pahlavan, K. (2006). Modeling of the TOA-based distance measurement error using UWB indoor radio measurements. IEEE Communications Letters, 10(4), 275–277. doi:10.1109/ LCOMM.2006.1613745

131

Node Localization in Ad-hoc Networks

Anisetti, M., Ardagna, C.A., Bellandi, V., Damiani, E., & Reale, S. (2008). Advanced localization of mobile terminal in cellular network. International journal of communications, network and system sciences, 1, 95-103. Barabell, A. J. (1983). Improving the resolution performance of eigenstructure-based directionfinding algorithms. IEEE International Conference on Acoustics, Speech, and Signal Processing, Boston, MA, USA, (pp. 336-339). Bulusu, N., Heidemann, J., & Estrin, D. (2000). GPS-less low-cost outdoor localization for very small devices. IEEE Personal Communications, 7(5), 28–34. .doi:10.1109/98.878533 Chen, P. C. (1999). A non-line-of-sight error mitigation algorithm in location estimation. IEEE Wireless Communications Networking Conference, New Orleans, LA (pp. 316-320). Cong, L., & Zhuang, W. H. (2005). Non-line-ofsight error mitigation in mobile location. IEEE Transactions on Wireless Communications, 4(2), 560–573. .doi:10.1109/TWC.2004.843040 Dempster, A. G. (2006). Dilution of precision in angle-of-arrival positioning systems. IEEE Electronics Letters, 42(5), 291–292. doi:10.1049/ el:20064410 Fukuju, Y., Minami, M., Morikawa, H., & Aoyama, T. (2003, May). DOLPHIN: an autonomous indoor positioning system in ubiquitous computing environment. IEEE Workshop on Software Technologies for Future Embedded Systems, Hakodate, Hokkaido, Japan, (pp. 53-56). Gillette, M. D., & Silverman, H. F. (2008). A Linear Closed-Form Algorithm for Source Localization from Time-Differences of Arrival. IEEE Signal Processing Letters, 15, 1–4. .doi:10.1109/ LSP.2007.910324

132

Girod, L., Lukac, M., Trifa, V., & Estrin, D. (2006). The Design and Implementation of a Self-Calibrating Distributed Acoustic Sensing Platform. The 4th International Conference on Embedded Networked Sensor Systems, Boulder, Colorado. (pp. 71-84). Goud, P., Sesay, A., & Fattouche, M. (1991). A spread spectrum radiolocation technique and its application to cellular radio. IEEE Pacific Rim Conference on Communications, Computers and Signal Processing, Victoria, BC, Canada, (pp. 661–664). Güvenç, I., Chong, C.C., Watanabe, F., & Inamura, H. (2008). NLOS Identification and Weighted Least-Squares Localization for UWB Systems Using Multipath Channel Statistics. EURASIP Journal on Advances in Signal Processing, 2008, Article ID 271984, 14 pages, 2008. doi:10.1155/2008/271984. Haddad, E. C., Despins, C., & Mermelstein, P. (2003). Capacity gain of zone division for a position-based resource allocation algorithm in WCDMA uplink data transmission (pp. 597–601). Beijing, China: IEEE Proceedings on Personal, Indoor and Mobile Radio Communications. Harder, A., Song, L., & Wang, Y. (2005). Towards an Indoor Location System Using RF Signal Strength in IEEE 802.11 Networks. IEEE International Conference on Information Technology: Coding and Computing, Las Vegas, (pp. 228-233). Hyeon, S., Lee, C., Shin, C., & Choi, S. (2009). Implementation of a Smart Antenna Base Station for Mobile WiMAX Based on OFDMA. EURASIP Journal on Wireless Communications and Networking. Article ID 950674, 9 pages. Ingram, S. J., Harmer, D., & Quinlan, M. (2004). Ultra-wideband indoor positioning systems and their use in emergencies. Position Location and Navigation Symposium, Rome, Italy, (pp. 706715).

Node Localization in Ad-hoc Networks

Jose, A., Neal, P., & Alfred, O. (2006). Distributed Weighted-Multidimensional Scaling for Node Localization in Sensor Networks. [TOSN]. ACM Transactions on Sensor Networks, 2(1), 39–64. .doi:10.1145/1138127.1138129

Lucke, O., Pellon, A., Closas, P., & Fernandez-Rubio, J. A. (2007). Cost-Optimised Active Receive Array Antenna for Mobile Satellite Terminals. Mobile and Wireless Communications Summit, 16th IST, Budapest, Hungary, (pp. 1-5).

Kaplan, E. D. (1996). Understanding GPS: Principles and Applications (2nd ed.). Boston, MA: Artech House.

Mayorga, C. L. F., Della Rosa, F., Wardana, S. A., Simone, G., Raynal, M. C. N., Figueiras, J., & Frattasi, S. (2007), Cooperative Positioning Techniques for Mobile Localization in 4G Cellular Networks. IEEE International conference on Pervasive Services, Istanbul, Turkey, (pp. 39-44).

Karimi, H. A., & Krishnamurthy, P. (2001). Realtime routing in mobile networks using GPS and GIS techniques. System Sciences, Proceedings of the 34th Annual Hawaii International Conference. Maui, Hawaii, USA. Konrad, L., & Matt, W. (2005). Motetrack: A robust, decentralized approach to RF-based location tracking. International Workshop on Location-and Context-Awareness, Oberpfaffenhofen, Germany, (pp. 63-82). Lee, J., Kim, N. Y., Kim, S. J., Kang, J., & Kim, Y. (2008). Joint AOA/RSSI based multi-user location system for military mobile base-station. IEEE Military Communications Conference, San Diego, CA, USA, (pp. 1-5). Li, X. (1998). A selective model to suppress NLOS signals in angle-of-arrival (AOA) location estimation. IEEE International Symposium on Personal, Indoor and Mobile Radio Communications, Boston, MA, USA, (pp. 461-465). Liberti, J. C., & Rappaport, T. S. (1999). Smart Antennas for Wireless Communications. Prentice Hall. Liu, H. F., Wong, L. N., & Shi, P. C. (2003). Cardiac motion and material properties analysis using data confidence weighted extended Kalman filter framework. IEEE International Conference on Acoustics, Speech, and Signal Processing, Hong Kong, China, (pp. 465-468).

Niculescu, D., & Badrih, N. (2003). Ad hoc positioning system (APS) using AOA. Twenty-Second Annual Joint Conference of the IEEE Computer and communications Societies, San. Francisco, CA, USA, (pp.1734-1743). Ohira, T., & Gyoda, K. (2001, December). Handheld microwave direction-of-arrival finder based on varactor-tuned analog aerial beamforming. Microwave Conference, 2001, APMC 2001, 2001 Asia-Pacific, Taipei, Taiwan, China, (pp. 585-588). Peebles, P. Z. (1998). Radar Principles. John Wiely & Sons, Inc. Rodriguez, M., Pece, J. P., & Escudero, C. J. (2005). In-building location using Bluetooth. International Workshop on Wireless Ad Hoc Networks, Coruna, Spain, (pp. 1-5). Schmidt, R. O. (1986). Multiple emitter location and signal parameter estimation. IEEE Transactions on Antennas and Propagation, 34(3), 276–280. doi:10.1109/TAP.1986.1143830 Seow, C. K., & Tan, S. Y. (2008). Non-Line-ofSight Localization in Multipath Environments. IEEE Transactions on Mobile Computing, 7(5), 647–660. .doi:10.1109/TMC.2007.70780

133

Node Localization in Ad-hoc Networks

Shankararaman, R., & Lourde, R. M. (2006). Attitude control of a 3-axis stabilized satellite using adaptive control algorithm. IEEE/SMC International Conference on System of Systems Engineering, Los Angeles, CA, USA, (pp. 1-6).

Venkatraman, S., & Caffery, J., Jr. (2002). A statistical approach to non-line-of-sight BS identification. IEEE International Symposium on Wireless Personal Multimedia Communications, Honolulu, Ha, USA, (pp. 296-300).

Shen, X., Wang, Z., Lin, R., & Sun, Y. (2005). Connectivity and RSSI Based Localization Scheme for Wireless Sensor Networks. International Conference on Intelligent Computing, Hefei, China, (pp. 578-587).

Wang, Z., Xu, W., & Zekavat, S. A. (2009). A New Multi-Antenna Based LOS - NLOS Separation Technique. 13th IEEE Digital Signal Processing Workshop and 5th IEEE Signal Processing Education Workshop, Marco Island, FL (pp. 331-336).

Skidmore, R. R., Verstak, A., Ramakrishnan, N., Rappaport, T. S., Watson, L. T., He, J., et al. (2004). Towards integrated PSEs for wireless communications: experiences with the S4W and SitePlanner \ text registered projects. ACM SIGMOBILE Mobile Computing and Communications Review, 8(2), 20–34. Retrieved from http://www.cs.vt. edu/node/1437

Wang, Z., & Zekavat, S. A. (2009, September). A Novel Semi-distributed Localization via Multinode TOA-DOA Fusion. IEEE Transactions on Vehicular Technology, 58(7), 3426–3435. doi:10.1109/TVT.2009.2014456

Stojmenovic, I. (2002, July). Position-based routing in ad hoc networks. Communication Magazine, IEEE, 40(7), 128–134. doi:10.1109/ MCOM.2002.1018018 Tong, H., & Zekavat, S. A. (2007, May). A Novel Wireless Local Positioning System via a Merger of DS-CDMA and Beamforming: Probabilityof-Detection Performance Analysis under Array Perturbations. IEEE Transactions on Vehicular Technology, 56(3), 1307–1320. doi:10.1109/ TVT.2007.895499 Venkatesh, S., & Buehrer, R. M. (2007). Nonline-of-sight identification in ultra-wideband systems based on received signal statistics. IET Microwaves. Antennas & Propagation, 1(6), 1120–1130. Venkatesh, S., & Buehrer, R. M. (2008). Multiple-access insights from bounds on sensor localization. Elsevier Journal on Pervasive and Mobile Computing, 4(1), 33–61. doi:10.1016/j. pmcj.2007.09.003

134

Wann, C. D., & Chin, H. C. (2007). Hybrid TOA/RSSI Wireless Location with Unconstrained Nonlinear Optimization for Indoor UWB Channels. IEEE Wireless Communications and Networking Conference, Hong Kong, China, (pp. 3940-3945). Xiong, L. (1998). A selective model to suppress NLOS signals in angle-of-arrival AOA location estimation. IEEE International Symposium on Personal, Indoor and Mobile Radio Communications, Boston (pp. 461–465). Yi, S., & Ruml, W. (2004). Improved MDS-Based Localization. IEEE INFOCOM 2004, Hong Kong, China, (pp. 2640-2651). Zekavat, S. A., Kolbus, A., Yang, X., Wang, Z., Pourrostam, J., & Pourkhaatoon, M. (2007). A novel implementation of DOA estimation for node localization on software defined radios: achieving high performance with low complexity. IEEE International Conference on Signal Processing and Communications, Dubai, United Arab Emirates, (pp. 983-986).

Node Localization in Ad-hoc Networks

KEY TERMS AND DEFINITIONS

ENDNOTES

Node Localization: Finding node position in a coordinate or a system. RSSI: Received signal strength indication. TOA Fusion: Processing TOA (range) information to localize target node. DOA Fusion: Processing DOA (angle) information to localize target node. Multi-Node TOA-DOA Fusion: Localizing target node by processing multiple range (TOA) and angle (DOA) information. Multidimensional Scaling: Finding nodes’ relative position in a system with single kind of measurement, e.g., range measurement, in which no base node is needed. NLOS Identification: Identifying the data (e.g., TOA, DOA) is captured through LOS or NLOS wireless channel. NLOS Mitigation: Mitigating the severe localization error generated by the data achieved through NLOS channel. NLOS Localization: Localizing target node directly using NLOS measurements without severe NLOS error.

1



2



3



4



5



Manhattan distance is the distance between two points measured along axes at right angles. If there are two points at (x1, y1, z1) and (x2, y2, z2), the corresponding Manhattan distance is |x1 - x2| + |y1 - y2|+ |z1 - z2|. Euclidean distance is the ordinary distance between two points. If there are two points at (x1, y1, z1) and (x2, y2, z2), the corresponding Euclidean distance is xT − xT ′ y − yT ′ . = T −x R ′ − xT ′ −yR ′ − yT ′ Low installation requirement corresponds to the connectivity fusion, in which, only anchor nodes position should be available before implementation, no RSS map is needed. The high installation requirement corresponds to the signature mapping method, in which, an RSS map should be made before implementation. Low performance is applied to connectivity fusion, and high performance is applied to signature mapping. Here we say it low, because only relative node position is available.

135

136

Chapter 9

Wireless and Mobile Technologies Improving Diabetes Self-Management Eirik Årsand Norwegian Centre for Integrate, Norway Naoe Tatara Norwegian Centre for Integrated Care and Telemedicine, Norway Gunnar Hartvigsen University of Tromsø, Norway

ABSTRACT The technological revolution that has created a vast health problem due to a drastic change in lifestyle also holds great potential for individuals to take better care of their own health. This is the focus of the presented overview of current applications, and prospects for future research and innovations. The presented overview and the main goals of the systems included are to utilize information and communication technologies (ICT) as aids in self-management of individual health challenges, for the disease Diabetes, both for Type 1 and Type 2 diabetes. People with diabetes are generally as mobile as the rest of the population, and should have access to mobile technologies for managing their disease. Fortyseven relevant studies and prototypes of mobile, diabetes-specific self-management tools meeting our inclusion criteria have been identified; 27 publicly available products and services, nine relevant patent applications, and 31 examples of other disease-related mobile self-management systems are included to provide a broader overview of the state of the art. Finally, the reviewed systems are compared, and future research directions are suggested.

INTRODUCTION Type 1 diabetes, also called insulin-dependent diabetes, is typically diagnosed in people under 30 DOI: 10.4018/978-1-60960-042-6.ch009

years of age. In this type, the pancreas has stopped producing insulin, so that the patient needs insulin injections. In Type 2 diabetes, which is typically diagnosed in people over 40 years of age, the body stops responding correctly to the insulin produced by the pancreas. This type constitutes

Copyright © 2011, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited.

Wireless and Mobile Technologies Improving Diabetes Self-Management

more than 90% of diabetes cases, and treatment includes oral medication, dietary changes, increased physical activity, and sometimes insulin injections. Self-management of blood glucose can help to reduce long-term complications of diabetes. Important self-management strategies include healthy eating habits, physical activity, and appropriate medication. Long-term effects of diabetes include progressive development of retinopathy with potential blindness, nephropathy that may lead to renal failure and/or neuropathy with risk of foot ulcers, amputations, sexual dysfunction and substantially increased risk of cardiovascular diseases. Most of the existing self-management tools for chronically ill patients are designed to provide help through interaction with health care workers. Even though this is usually the kind of help that patients want most and that is also the most effective, e.g. (Calfas et al., 2002), (Martinson et al., 2008) and (Shishko, Mokhort, & Garmaev, 2006), it is resource-intensive. A specific change of focus expressed by the European Commission (EC) a few years ago in the Information Society Technologies (IST) programme (European Commission, 2005) was to orient R&D towards one process to integrate and use all relevant biomedical information for improving health knowledge and processes related to prevention, diagnosis, treatment, and personalization of health care. We first give an overview of mobile diabetesspecific self-management systems and tools, including publicly available systems, prototypes typically designed for research studies, and relevant patents and patent applications. Then, some systems for other chronic diseases are presented – but more briefly, in order to illustrate how technology is applied within the personalized health area in general. The objective of this chapter is to provide updated information about how mobile and wireless technologies are currently used in the context of eHealth, to describe the usability of the technologies in relation to people with a chronic disease such as diabetes, and to suggest

sound future directions for coming technologies that support mobile self-management.

BACKGROUND A search for patient-operated diabetes management software in general, including PC/Internet tools, shows that there are many systems available. No recently updated reviews were found, but a six-year-old study (Park & Daly, 2003) identified 47 Web-based or Windows-based programs for assisting people with diabetes in their self-help regimen, excluding educational and informational software. Few reviews of mobile diabetes systems were found. A search of the Cochrane Reviews database in June 2009 using the search words “diabetes” and “mobile” in all text fields yielded no relevant reviews, but some results with the status “Stage: Protocol”. Besides the review by Tatara, Årsand, Nilsen, and Hartvigsen (2009), two more general reviews covering the use of SMS in healthcare by Krishna, Boren, and Balas (2009) and Fjeldsoe, Marshall, and Miller (2009) were found, identifying some additional diabetesspecific mobile systems. Systems for and studies on self-management tools involving assistance by health care personnel are widespread, e.g. the DiasNet advisory system (Dinesen & Andersen, 2006), the system developed by Axon TeleHealthCare (PA Business, 2008), the telephone-linked care system (Glanz et al., 2002), the PARIS_Diabtel system (Rigla et al., 2006), the TeleObe programme (Schiel, Beltschikow, Radon, & Kramer, 2006), the Internet-based system BioDang (Kwon et al., 2004), and the Healthcare@Home monitoring framework (Subramanian et al., 2008). Our focus is however on patient-operated mobile self-management tools, an area that has exhibited relatively strong growth during the last three years, but is still rather immature. The main reason for the recent growth may be the evolution of mobile phones into small, programmable and function-

137

Wireless and Mobile Technologies Improving Diabetes Self-Management

rich computers. Several studies and prototypes as well as some publicly available products and services exist, directed at people with Type 1 and Type 2 diabetes, but few of these can actually be classified as self-management tools according to our definition: tools designed for personalized and patient-oriented use. This unfortunately means that despite the possibilities that the mobile terminals, miniaturization and wireless communication technologies provide, there are still relatively few efforts to create mobile systems intended for personalized use.

•

•

• • •

STATE OF THE ART: MOBILE SELFMANAGEMENT DIABETES SYSTEMS The state of the art in wireless and mobile technologies to improve diabetes self-management was explored by searching literature databases, searching the Internet, and searching patent databases.

Coverage in Scientific Publications To explore the extent to which mobile selfmanagement diabetes systems are addressed in scientific publications, peer-reviewed journals or conference papers were searched for combinations of the words “diabetes”, “mobile”, “PDA”, and “cellular”, mainly in the same data sources as used in Tatara et al. (2009); i.e. PsycINFO, EMBASE, CINAHL, Pubmed (includes MEDLINE, which includes JMIR), Cochrane Library, ISI Web of Science, INSPEC (Ovid), ACM Digital Library, IEEE Xplore. In addition, the ISI Web of Knowledge, Lecture Notes in Computer Science (LNCS), and the American Diabetes Association journals were searched. More relevant literature was found by checking the references of the identified relevant papers. The search was performed in May and June 2009, and the relevant tools, projects and trials that were identified are listed in Table 1. The inclusion criteria were:

138

mobile, patient-operated, self-help tools (in the general sense of the term) for people with diabetes systems suitable for use outside hospitals for fairly long periods, typically more than a month inclusion of an evaluation or a description of the system systems tested on at least one patient innovative diabetes self-help systems, including those without functions for management of blood glucose measurements (in contrast to the criteria for the systems included in Table 2, where the systems had to include at least functionality for blood glucose parameters)

The system may or may not involve interaction with the health care sector. We emphasize that there are many self-management systems for diabetes that are Web- and PC-based, as well as artificial pancreas systems, which are used only in hospitals, and are therefore not included in this section.

Publicly Available Systems There exist various relevant products and other publicly available systems, some of which are listed in Table 2. The systems were found by searching for prototype names, cooperative partners and companies mentioned in the cited literature listed in Table 1. Perhaps the best online overview of diabetes management software and systems in general is provided at David Mendosa’s website: www.mendosa.com/software.htm, which has also served as a source for finding the publicly available products and services listed in Table 2. The main criteria for inclusion were that the systems: • •

were mobile (usually based on mobile phones or PDAs); were publicly available;

Wireless and Mobile Technologies Improving Diabetes Self-Management

Table 1. Relevant studies and prototypes of mobile, diabetes-specific self-management tools, ↜sorted by the year of publication No.

Main disease

Name of tool, project and/or trial (year of publication)

No. of users*

Reference

1.

Type 1 diabetes

CARDS: Computerized Automated Reminder Diabetes System (SMS, e-mail) (2009)

22

(Hanauer, Wentzell, Laffel, & Laffel, 2009)

2.

Diabetes, type not reported

Using Zigbee and mobile phones for elderly patients (2009)

17

(H. J. Lee et al., 2009)

3.

Type 1 diabetes

The Diabetes Interactive Diary (DID) (2009)

41

(Rossi et al., 2009)

4.

Type 1 diabetes

Wireless Personal Assistant for telemedical diabetes care (2009)

10

(García-Sáeza et al., 2009)

5.

Type 1 and Type 2 diabetes

Evaluation of a mobile phone telemonitoring system for glycaemic control in patients with diabetes (2009)

72

(Istepanian et al., 2009)

6.

Type 2 diabetes

Mobile communication using a mobile phone with a glucometer (HealthPia) for glucose control – comparison with Internet-based glucose monitoring (2009)

38

(Cho, Lee, Lim, Kwon, & Yoon, 2009)

7.

Type 2 diabetes

Intervention study on the WellDoc’s Diabetes Manager system (Mobile phone and Web) (2009)

185

(C. Quinn et al., 2009)

8.

Type 2 diabetes

User-Involved Design of Mobile Self-Help Tools for People with Diabetes (Mobile phone) (2009)

12

(Årsand, Tatara, Østengen, & Hartvigsen, 2010)

9.

Type 2 diabetes

A short message service by cellular phone in type 2 diabetic patients (2008)

25

(Yoon & Kim, 2008)

10.

Type 2 diabetes

WellDoc: Mobile Diabetes Management (Mobile phone and PC) (2008)

15

(C. C. Quinn et al., 2008)

11.

Type 2 diabetes

Nurse intervention using SMS and Internet (PC) (2008)

18

(S. I. Kim & Kim, 2008)

12.

Type 2 diabetes

The NICHE pilot study (mobile phone and Internet) (2008)

15

(Faridi et al., 2008)

13.

Type 2 diabetes

Continuous glucose monitoring to change physical activity behavior (CGMS monitor, accelerometers) (2008)

27

(Allen, Fain, Braun, & Chipkin, 2008)

14.

Type 1 diabetes

Mobile Phones Assisting With Health Self-Care: a Diabetes Case Study (2008)

11

(Preuveneers & Berbers, 2008)

15.

Type 1 and Type 2 diabetes

MAHI (Mobile Access to Health Information) (2008)

25

(Mamykina, Mynatt, Davidson, & Greenblatt, 2008)

16.

Type 1 diabetes

The INCA System (PDA-based patient intervention, study prior to closed-loop test) (2008)

10

(Enrique J. Gomez et al., 2008)

17.

Diabetes and cardiovascular disease

MediNet: Personalizing the Self-Care Process for Patients with Diabetes and Cardiovascular Disease Using Mobile Telephony (2008)

Initial tests only

18.

Type 1 diabetes

Diab-Memory: Mobile Phone–Based Data Service for Functional Insulin Treatment of Type 1 Diabetes (2007)

10

(Kollmann, Riedl, Kastner, Schreier, & Ludvik, 2007)

19.

Type 1 diabetes

The HealthPia GlucoPack Diabetes (mobile) Phone (2007)

10

(Carroll, Marrero, & Downs, 2007)

20.

Type 1 diabetes

Using cellular phones (the GlucoNet system) in type 1 diabetic patients during insulin pump therapy: the PumpNet study (2007)

30

(Benhamou et al., 2007)

21.

Diabetes and other chronic diseases

MyMobileDoc - a Mobile Medical Application for the Management of Chronic Diseases (2007)

15

(Nischelwitzer, Pintoffl, Loss, & Holzinger, 2007)

22.

Type 1 and Type 2 diabetes

Combining digital photography and glucose data (2007)

7

(Smith, Frost, Albayrak, & Sudhakar, 2007)

(P. Mohan, Marin, Sultan, & Deen, 2008), (Permanand Mohan & Sultan, 2009)

continued on the following page

139

Wireless and Mobile Technologies Improving Diabetes Self-Management

Table 1. continued No.

Main disease

Name of tool, project and/or trial (year of publication)

No. of users*

Reference

23.

Type 1 and Type 2 diabetes

The ProWellness Self-Care System - Information technology supporting diabetes self-care (2007)

9

(Halkoaho, Kavllo, & Pietila, 2007)

24.

Type 1 and Type 2 diabetes, and other diseases

The Singapore health services experience (SMS and Internet/ PC) (2007)

n/a

(Seng et al., 2007)

25.

Type 1 diabetes

Recording of hypoglycaemic attacks. (SMS, Internet/PC and diary) (2007)

19

(Tasker, Gibson, Franklin, Gregor, & Greene, 2007)

26.

Type 1 diabetes

DiasNet Mobile (based on the DiasNet system) (2007)

1

(Jensen & Larsen, 2007)

27.

Type 2 diabetes

The CenTexNet Study: PDA Use in Diabetes Self-Care (“Diabetes Pilot” software) (2007)

42

(Forjuoh et al., 2007)

28.

Type 1 and Type 2 diabetes

Mobile Dietary Management Support Technologies for People with Diabetes (2007)

6

(Årsand, Tufano, Ralston, & Hjortdahl, 2008)

29.

Type 2 diabetes

Usability of a Mobile Self-Help Tool for People with Diabetes: the Easy Health Diary (2007)

32

(Årsand, Varmedal, & Hartvigsen, 2007)

30.

Type 1 diabetes

Sweet Talk: Text Messaging Support for Intensive Insulin Therapy for Young People with Diabetes (2006)

64

(Franklin, Waller, Pagliari, & Greene, 2003), (Franklin, Waller, Pagliari, & Greene, 2006)

31.

Type 1 and 2, visually impaired

eDiab: monitoring, assisting and educating people with diabetes (PDA or mobile phone) (2006)

n/a

(Fernandez-Luque et al., 2006)

32.

Type 1 diabetes

VIE-DIAB: reporting blood glucose, carbo-hydrate intake, insulin dosage) via mobile phone (2006)

36

(Rami, Popow, Horn, Waldhoer, & Schober, 2006)

33.

Type 1 diabetes

DiasNet-PN (based on the DiasNet system) (2006)

1

(Jensen, Pedersen, & Larsen, 2006)

34.

Type 1 diabetes

The telematic communication GlucoBeep system (2006)

20

(Jansa et al., 2006)

35.

Type 1 and Type 2 diabetes

Diabetic e-Management System (DEMS) (2006)

13

(Lutes, Chang, & Baggili, 2006)

36.

Type 1 diabetes

Diabetes education via mobile text messaging (2006)

11

(Wangberg, Årsand, & Andersson, 2006)

37.

Type 1 diabetes

Parent-Child Interaction Using a Mobile and Wireless System for Blood Glucose Monitoring (2005, 2007)

15

(Gammon et al., 2005), (Årsand, Andersson, & Hartvigsen, 2007)

38.

Type 1 diabetes

A real-time, mobile phone-based telemedicine system to support young adults with type 1 diabetes (2005)

93

(Farmer, Gibson, Hayton et al., 2005), (Farmer, Gibson, Dudley et al., 2005)

39.

Not specified

Mobile phone text messaging (SMS) in the management of diabetes (2004)

23

(Ferrer-Roca, Cardenas, DiazCardama, & Pulido, 2004)

40.

Type 1 and Type 2 diabetes

DiabNet: integration of handheld computer, mobile phone and Internet access (2004)

n/a

(Roudsari, Zhao, & Carson, 2004)

41.

Type 1 diabetes

The DAILY (Daily Automated Intensive Log for Youth) Trial: A Wireless, Portable System to Improve Adherence and Glycemic Control in Youth with Diabetes (2004)

40

(Kumar, 2004), (Kumar, Wentzell, Mikkelsen, Pentland, & Laffel, 2004)

42.

Type 1 diabetes

Cellular phone transfer for blood glucose self -monitoring – the WellMate system (2004)

100

(Vähätalo, Virtamo, Viikari, & Rönnemaa, 2004)

continued on the following page

140

Wireless and Mobile Technologies Improving Diabetes Self-Management

Table 1. continued No.

Name of tool, project and/or trial (year of publication)

Main disease

No. of users*

Reference

43.

Type 1 diabetes

Edutainment Tools for Initial Education of Type-1 Diabetes Mellitus: Initial Diabetes Education with Fun (2004)

58

(Aoki et al., 2004)

44.

Type 1, Type 2 and Gestational diabetes

The M²DM project, includes automatic generation of reminders and alarms, transmitted by SMS to patients (2003, 2006)

38

(Bellazzi et al., 2003), (Larizza et al., 2006)

45.

Type 1 diabetes

DiaBetNet: a handheld computer (guessing) game for young diabetics (blood glucose, carbs., insulin) (2003, 2004)

40

(Lane, 2003), (Pentland, 2004)

46.

Type 1 diabetes

DIABTel: evaluation of a Telemedicine system (PC and palmtop computer) (2002)

10

(E.J. Gomez et al., 2002)

47.

Not specified

Diabetes Monitoring System (DMS) based on a hand-held, touch-screen electronic diary (meals, blood glucose) (2001)

19

(Tsang et al., 2001)

* The number of users represents those who used the technology, not the total cohort, i.e. control groups are not counted.

• •

had at least a function to monitor blood glucose values, and provided more help or information than ordinary blood glucose meters.

However, there are many small applications, typically developed in student projects and made publicly available, which not were included in this overview. Blood glucose monitors (BGM) communicating with insulin pumps were also excluded, due to their generally limited circulation among people with diabetes. The search was done in May and June 2009.

Relevant Patents and Patent Applications For an overview of relevant systems filed in patent databases, searches were performed (June 2009) in the US-based patent database “Patentstorm” (PatentStorm LLC, 2009), which claims to be updated weekly with the U.S. Patent Office’s databases, and in the European patent database “esp@ cenet” (European Patent Office, 2009) (worldwide search). A mobile diabetes self-management tool might be considered too rich in functionality to allow filing of a patent. To provide further indications of whether this was the case, several patent searches were performed.

Diabetes, Mobile and Handheld: Searching Patentstorm using the criterion “diabetes and mobile” in the patent’s ‘Title’ field yielded no results, but the esp@cenet search found two (Henry, 2005),(Schildt, Franke, Kirmse, & Leitzmann, 2003). Both of these describe a remote monitoring system between a user with diabetes and the health care system or a database. Searching the same two patent databases using the search term “diabetes and handheld” in the ‘Title’ field returned no results in Patentstorm and two results in esp@ cenet. One of these concerned a pen-type injector and the other a device containing nutritional information. Full text searches for “Diabetes and Mobile and Patient-Operated”: A search in the patent databases’ ‘Full text’ field was expected to return very many more results, so more specific search terms were tested. The Patentstorm and esp@cenet had different search functionality for long search terms, so that we had to formulate the respective searches in two different ways. Patentstorm allowed long search terms in the ‘Full text’ field. A search for the criterion “diabetes and mobile and (self-help or self-management) and (tool or device or unit)” returned no patents. The shorter term “diabetes and mobile and patient-operated” returned 21 patents or patent applications. Most of these involved either “remote patient monitoring”,

141

Wireless and Mobile Technologies Improving Diabetes Self-Management

Table 2. Relevant and publicly available mobile diabetes-specific self-management systems No.

Main disease

1.

Systems for both Type 1 and Type 2 diabetes, or not specified

Name of vendor / product

URL

t+ Medical / t+ Diabetes

http://www.tplusmedical.co.uk/ information/01Patients--04tplus_diabetes.html

SugarStats LLC / SugarStats Mobile edition

http://www.sugarstats.com/

Digital Altitudes LLC / Diabetes Pilot

http://www.diabetespilot.com/

FutureWare / Personal GlucoseTracker

http://www.futurewaredc.com/products/FutureWare_PersonalGlucoseTracker.html

5.

Alive Technologies / Alive Diabetes Management System

http://www.alivetec.com/products.htm

6.

SINOVO Ltd. & Co. KG / SiDiary

http://www.sidiary.org/

7.

Diabetech LP / GlucoMON

http://mygluco.com/glucomonhowitworks

8.

HealthEngage / HealthEngage

http://www.healthengage.com/

9.

SymCare Personalized Health Solutions / SymCare

http://www.symcare.com/

2. 3. 4.

10.

MYLEstone Health / Glucose Buddy

http://beta.glucosebuddy.com/

11.

Ace t&t / DiabGo

http://www.ace-tt.com/

Elardo GbR / Elardo DiabetesProfiler

http://www.handango.com/catalog/ProductDetails.jsp?storeId=2218&productId=87853

13.

Mobile Diabetic Inc. / LogbookFX Diabetic Diary

http://www.mdiabetic.com/

14.

Glucose-Charter / Glucose-Charter Pro

http://glucose-charter.com/

GlucoControl / GlucoControl 3.0

http://www.brothersoft.com/glucocontrol-27960. html

16.

GlucoTools / GlucoTools

http://glucotools.sourceforge.net/

17.

UTS / UTS Diabetes

http://utracksys.com/software-diabetes/

18.

HMM Diagnostics GmbH / smartLAB genie

http://www.smartlab.org/genie/

19.

Confidant Inc. / CONFIDANT Diabetes Solution

http://www.confidantinc.com/

20.

Entra Health Systems Ltd / myglucohealth

http://www.myglucohealth.net/

21.

Apple Inc. / iPhone Diabetes applications (60 applications were found by searching iTunes for “Diabetes”, of which 16 supported manual input of BG data)

http://www.apple.com/no/ipodtouch/appstore

22.

GlucoseOne / GlucoseOne Palm Application

http://www.glucoseone.com/

23.

Polymap Wireless / The Polytel System

https://www.polymapwireless.com/

24.

LifeScan Inc. / OneTouch UltraSmart

http://www.lifescan.com/

HealthPia USA / GlucoPhone

http://healthpia.us/

Insulet Corporation / OmniPod with Personal Diabetes Manager

http://www.myomnipod.com/

WellDoc Communications / DiabetesManager

http://www.welldoc-communications.com/

12.

15.

25.

Juvenile diabetes (Type 1)

26.

Insulin dep. diabetes

27.

Type 2

or remote data storage or forwarding to a health care system. An interesting concept described in the patent application by Dicks et al. (Dicks, Kent, Tripp, Bartlett, & Crosley, 2008) from 2008 presents a smart way (introducing an additional device) to trigger automatic data transmission

142

from a medical device. It actually also exemplifies the use of the system using a blood glucose monitor device ((Dicks et al., 2008), Figure 4). A search in Patentstorm for “diabetes and mobile” alone returned 14275 results. The esp@ cenet database allows searches of the “abstract”

Wireless and Mobile Technologies Improving Diabetes Self-Management

and the “title” fields together as the option most similar to free text search, with a maximum of five keywords. Searching for the term “diabetes and mobile and patient-operated” yielded no results. Searching for the term “diabetes and mobile” returned 15 results. Of these, (Rao, Rao, & Rao, 2008),(Patel, Istoc, Lin, & Narang, 2008),(Rosen, 2007),(Ichikawa, 2006),(Henry, 2005),(Schildt et al., 2003) were relevant, and involved remote data monitoring, which might be relevant in future if self-management systems become more closely integrated with health care services.

Other Health Related Mobile Self-Management Systems Fitness: For fitness and physical activity purposes, a variety of mobile devices and plans for utilization of existing devices is available, e.g.: 1. the “TripleBeat” system (Oliver & FloresMangas, 2006); 2. the chest belt systems from Polar (Polar Electro Oy, 2008); 3. the “Nike+ SportBand” (Nike Inc., 2008); 4. the “Forerunner 301” system (Garmin Ltd., 2008); 5. the “PmEB” mobile-phone application for monitoring caloric balance (G. Lee, Tsai, Griswold, Raab, & Patrick, 2006); 6. the “Affective Diary” (Ståhl, Höök, Svensson, Taylor, & Combetto, 2009); 7. the “SenseWear” armband from BodyMedia (BodyMedia Inc., 2009); and 8. the “Mobile Health Diary” (Alpesh Tarapara Aristocrats, 2008). Heart Related: Generally, heart diseases involve more immediate risks than the various kinds of diabetes and other lifestyle-related diseases. The use of technology for this disease case reflects this: usually, the sensors are constantly active, and communication to health care actors is involved. Some examples are:

1. the electrocardiogram (ECG) sensor from Alive Technologies in combination with a smartphone (Lim, Chen, Ho, Tin, & Sankaranarayanan, 2007); 2. the ECG-sensor concept described by Fensli, Gunnarson, and Gundersen (2005); 3. the system based on mobile phones and pulse monitoring described by Lee, Hsiao, Chen, and Liu (2006); 4. the “eHit Health Gateway” proposed by Holopainen, Galbiati, and Voutilainen (2007); and 5. the “Wellness Diary” (Nokia, 2008). Asthma is a chronic disease which is generally treated by avoiding the things that trigger asthma attacks and taking one or more asthma medications. It sometimes involves use of a peak expiratory meter (PEF), and treatment varies from person to person. Examples of mobile systems and studies related to this disease include the following: 1. the mobile phone-based monitoring system for self-management of asthma described by Pinnock, Slack, Pagliari, Price, and Sheikh (2007); 2. the study by Ryan, Cobern, Wheeler, Price, and Tarassenko (2005), which also describes a system linking a mobile phone to an electronic spirometer/peak flow meter; 3. the Web-based mobile asthma management system (H. R. Lee, Yoo, Jung, Kwon, & Hong, 2005); and 4. the asthma management application for children and teens (Boland, 2007). Smoking: There are currently few studies that address use of tailored mobile self-help tools in smoking cessation. Some however describe use of the standard, built-in functionalities of mobile phones, typically SMS or MMS (multimedia messages), e.g.:

143

Wireless and Mobile Technologies Improving Diabetes Self-Management

1. the study of Rodgers et al. (2005); 2. the study described by Blake (2008); 3. the prototype text-messaging program and a Web-based program described by Riley, Obermayer, and Jean-Mary (2008); 4. smoking cessation counselling in a lowincome, HIV-positive population (Lazev, Vidrine, Arduino, & Gritz, 2004); 5. the concept of sending multimedia messages described by Whittaker et al. (2008); and 6. t h e mobile phone-based “QuitSmokingMobile” (QSM) application (QuitSmokingMobile.com, 2009). Other areas: 1. Mental health: e.g. the “Mobile Mood Diary” (Trinity College Dublin, 2008); 2. Chronic obstructive pulmonary disease (COPD): e.g. in the study described by Liu et al. (2008); 3. Cancer: e.g. for managing chemotherapyassociated side effects (Weaver et al., 2007); 4. Obesity: e.g. the TMS system (Morak et al., 2008), the Sensorphone system (Vlaskamp et al., 2007), and two SMS systems for overweight users (Patrick et al., 2009) and (Gerber, Stolley, Thompson, Sharp, & Fitzgibbon, 2009); 5. Eating disorders: e.g. the use of SMS as a therapeutic intervention (Hazelwood, 2008); 6. Dementia: the GSM- and GPS-based mobile “rescue locator” (Lin, Chiu, Hsiao, Lee, & Tsai, 2006); 7. Epilepsy: the PDA-based application for reporting side effects of treatment (Frings et al., 2008); and 8. Elderly people and healthcare in general: e.g. a mobile phone-based system for helping people with early dementia with everyday activities, described by Donnelly, Nugent, Craig, Passmore, and Mulvenna (2008), and the mobile medication adherence system

144

“SIMpill” (SIMpill, 2009), which sends SMS reminders about patients’ medication.

COMPARISON OF THE REVIEWED SYSTEMS Table 3 below shows the degree of completion and ease of use for the most function-rich systems for diabetes self-management with regard to the three cornerstone elements: healthy diet, blood glucose management, and exercise. The main criteria for including the systems in this comparison were that the systems were mobile (usually based on a mobile phone or PDA) and had at least the functionality to monitor blood glucose values. The systems included are mainly publicly available systems from Table 2, but there are also some relevant and promising prototypes from Table 1, indicated by “prototype” in the first column. Patents are not included in this comparison since they are more general and less concrete, contain no reference to clinical evaluations, and have restricted or no availability. As Table 3 shows, we found no systems that included all three cornerstone elements among the applications that are publicly available. The “MDoctor” (K. Kim et al., 2006) system is very similar to the “Few Touch” application (Årsand, 2009) in functionality, but little information about this system was found. The “Few Touch” application comprises fully automatic transfer of blood glucose data and step count data, and a system for recording food habits by only two touches on the mobile phone. The “Diabetes Interactive Diary” (Rossi et al., 2009) also looks promising in terms of functionality, but regarding usability it seems to be based on manual data input for most parameters. The “LogbookFX Diabetic Diary” (Mobile Diabetic Inc, 2009), the “SiDiary” (SINOVO Ltd. & Co. KG, 2009), the “WellDoc DiabetesManager“ (WellDoc Inc, 2007), and the “smartLAB genie” (HMM Diagnostics GmbH, 2008) were the publicly available systems that were closest in functionality, with automatic wire-

Wireless and Mobile Technologies Improving Diabetes Self-Management

Table 3. Diabetes-related functionalities included in function-rich mobile self-management systems Systems / Functions ↓→

BGM

Physical activity

Nutrition

Usability issues

Few Touch application (Årsand, 2009) (prototype)

X

X

X

Automatic wireless transmission of BG data and step-count data to mobile phone

MDoctor for DM (K. Kim, Han, Lee, Kim, & Ahn, 2006) (prototype)

X

X

X

“Full automatic recording of blood-glucose and exercise data”.

The Diabetes Interactive Diary (DID) (Rossi et al., 2009) (prototype)

X

X

X

“Automatic storage of blood glucose measurements”, manual physical activity registration.

LogbookFX Diabetic Diary (Mobile Diabetic Inc, 2009)

X

-

X

Wireless transmission of BG data.

SiDiary (SINOVO Ltd. & Co. KG, 2009)

X

-

X

Automatic wireless transmission of BG data to phone and desktop.

DiabetesManager (WellDoc Inc, 2007)

X

-

X

Automatic wireless transmission of BG data to health care web site.

smartLAB genie (HMM Diagnostics GmbH, 2008)

X

X

X

Automatic wireless transmission of BG data to mobile phone or computer.

OneTouch UltraSmart (LifeScan Inc., 2009)

X

X

X

A BGM with options for recording food and exercise data by using predefined choices using its keypad.

SymCare, the In-Touch Diabetes system (SymCare Personalized Health Solutions, 2009)

X

-

-

Automatic wireless transmission of BG data to health care web site.

The Polytel System (Polymap Wireless, 2008)

X

-

-

Automatic wireless transmission of BG data to phone and server.

Alive Diabetes Management System (Alive Technologies Pty. Ltd., 2009)

X

-

-

Automatic wireless transmission of BG data to health care web site or others.

GlucoMON (Diabetech LP, 2008)

X

-

-

Automatic wireless trans-mission of BG data to remote caregivers (manually initiated).

GlucoPhone (HealthPia USA, 2008)

X

-

-

Automatic wireless transmission of BG data to health care web site or others.

myglucohealth (Entra Health Systems, 2009)

X

-

-

Automatic wireless transmission of BG data to mobile phone or computer.

t+ Diabetes (t+ Medical, 2007)

X

X

X

Wirelessly transferred BG data, but manually initiated. Manual entry of other data.

Personal Assistant (García-Sáeza et al., 2009) (prototype)

X

-

X

Wirelessly transferred BG data, but manually initiated.

CONFIDANT Diabetes Solution (Confidant International LLC, 2007)

X

-

-

Wirelessly transferred BG data, but manually initiated, to mobile phone and remote server.

OmniPod with Personal Diabetes Manager (Insulet Corporation, 2009)

X

-

X

Blood glucose (CGM), food planning and insulin delivery in one unit.

Glucose Buddy (MYLEstone Health, 2009)

X

X

X

Manually initiated data transfer from phone to desktop.

Elardo DiabetesProfiler (Mündlein & Koch, 2007)

X

X

X

Manual entry of data, manually initiated data transfer from PDA to desktop.

Personal GlucoseTracker (Palm/Handspring companion program) (FutureWare, 2009)

X

-

-

Manually initiated data transfer from desktop to phone.

SugarStats Mobile Edition (SugarStats LLC, 2009)

X

X

X

Manual entry of data.

DiabGo (Ace t&t, 2009)

X

X

X

Manual entry of data.

145

Wireless and Mobile Technologies Improving Diabetes Self-Management

less transmission of blood glucose data. The blood glucose monitor “OneTouch UltraSmart” (LifeScan Inc., 2009) has predefined choices for recording food, exercise, medication and other health data, using the BGM’s keypad and menus. However, the operation of this system requires much manual entry using the keypad and it does not seem to be widely used. The Personal Assistant prototype (García-Sáeza et al., 2009) is rich in functions including insulin recording, and physical activity can be recorded as a general parameter, but its mobile terminal needs charging each 12 hours.

CONCLUSION Most of the mobile phone-based self-management solutions either use the very simple functionalities such as SMS, or more content-rich functionalities such as pictures and illustrations managed by the MMS services or mobile Web. A major problem with some of these solutions is that the data must be manually recorded using small keypads or softkeys or a stylus on the small touch-sensitive mobile-phone screen. Fragmented functionality is available across a variety of systems, but it is hard to find solutions incorporating sensors, analysis, feedback, and general information in a holistic way. Bluetooth as a short-range communication standard is spreading rapidly, but as long as the Medical Device Profile for Bluetooth (Bluetooth SIG, 2007) is not fully established, the sensor connections are neither easy enough to implement, nor adequately standardized. However, recent advances such as Microsoft´s HealthVault offer solutions for connecting various medical sensors to an online patient diary (Microsoft Corporation, 2009). Even though the typical mobile self-management system is designed to be used for secondary prevention, i.e. on people who have already been diagnosed with the disease, research shows that lifestyle intervention can also prevent conditions

146

such as Type 2 diabetes in high-risk subjects (Tuomilehto et al., 2001). In future plans, one should therefore consider use of such systems for primary prevention as well as secondary prevention. Many of the diabetes-related systems presented may be used also on heart disease patients, for obese patients, for fitness purposes, and even on children with various chronic diseases. People with a high risk of developing diseases such as diabetes or various kinds of heart diseases could for example use only the physical activity module and the nutrition habit registration module of some of applications presented. Future efforts should address whether and how mobile self-management systems could form part of the health care system, and how such systems can be designed to offer peer-feedback functionalities. There is also a vast medical research potential if these data are made available in databases for data mining and refined analysis. Sound algorithms for processing and presentation of data both for patients and for their health care helpers will further enhance the value of these systems. Mobile self-management systems will be even more valuable when implantable sensors become more common, accurate, and safe to use. Sensor systems will then be easier to use, making it more likely that patients will accept the total concept, e.g. using a mobile phone as their patient terminal. We strongly agree with the conclusion of Ballegaard, Hansen, and Kyng (2008) that healthcare technology involves much more than informing clinicians; it is also about supporting the collaboration between patients and clinicians – and we expect that wireless and mobile technologies will play an important role.

REFERENCES Ace t&t. (2009). DiabGo Version 2.5. Retrieved 14. May, 2009, from http://www.ace-tt.com/

Wireless and Mobile Technologies Improving Diabetes Self-Management

Alive Technologies Pty. Ltd. (2009). Alive Diabetes Management System. Retrieved 16. May, 2009, from http://www.alivetec.com/products.htm Allen, N. A., Fain, J. A., Braun, B., & Chipkin, S. R. (2008). Continuous glucose monitoring counseling improves physical activity behaviors of individuals with type 2 diabetes: A randomized clinical trial. Diabetes Research and Clinical Practice, 80(3), 371–379. doi:10.1016/j.diabres.2008.01.006 Alpesh Tarapara Aristocrats. (2008). Mobile Health Diary. Retrieved 15 July 2009, from http:// www-users.rwth-aachen.de/Alpesh.Tarapara/ MHD/index.html Aoki, N., Ohta, S., Masuda, H., Naito, T., Sawai, T., Nishida, K., et al. (2004). Edutainment Tools for Initial Education of Type-1 Diabetes Mellitus: Initial Diabetes Education with Fun. Paper presented at the Medinfo, Amsterdam, The Netherlands. Årsand, E. (2009). The Few Touch Digital Diabetes Diary - User-Involved Design of Mobile Self-Help Tools for People with Diabetes. Doctoral thesis, University of Tromsø, Norway. Årsand, E., Andersson, N., & Hartvigsen, G. (2007, 26.-28. November). No-Touch Wireless Transfer of Blood Glucose Sensor Data. Paper presented at the COGIS’07; Cognitive systems with Interactive Sensors, Stanford University, California, USA. Årsand, E., Tatara, N., Østengen, G., & Hartvigsen, G. (2010). Mobile Phone-Based Self-Management Tools for Type 2 Diabetes: The Few Touch Application. Journal of Diabetes Science and Technology, 4(2), 328–336. Årsand, E., Tufano, J., Ralston, J., & Hjortdahl, P. (2008). Designing Mobile Dietary Management Support Technologies for People with Diabetes. Journal of Telemedicine and Telecare, 14(7), 329–332. doi:10.1258/jtt.2008.007001

Årsand, E., Varmedal, R., & Hartvigsen, G. (2007, 22.-25 September). Usability of a Mobile Self-Help Tool for People with Diabetes: the Easy Health Diary. Paper presented at the IEEE CASE 2007, Scottsdale, Arizona, USA. Ballegaard, S. A., Hansen, T. R., & Kyng, M. (2008). Healthcare in Everyday Life - Designing Healthcare Services for Daily Life. Paper presented at the CHI 2008. Retrieved 21. May 2009, Bellazzi, R., Arcelloni, M., Bensa, G., Blankenfeld, H., Brugues, E., Carson, E., et al. (2003). Design, Methods, and Evaluation Directions of a Multi-Access Service for the Management of Diabetes Mellitus Patients. Diabetes Technology & Therapeutics, 5(4), 621-629. Benhamou, P.-Y., Melki, V., Boizel, R., Perreal, F., Quesada, J.-L., & Bessieres-Lacombe, S. (2007). One-year efficacy and safety of Web-based follow-up using cellular phone in type 1 diabetic patients under insulin pump therapy: the PumpNet study. Diabetes & Metabolism, 33(3), 220–226. doi:10.1016/j.diabet.2007.01.002 Blake, H. (2008). Innovation in practice: mobile phone technology in patient care. British Journal of Community Nursing, 13(4), 160–165. Bluetooth, S. I. G. (2007). Bluetooth SIG announces Medical Device Profile at Medica show. Retrieved 13. May, 2009, from http://www. bluetooth.com/Bluetooth/Press/SIG/BLUETOOTH_SIG_ANNOUNCES_MEDICAL_DEVICE_PROFILE_AT_MEDICA_SHOW.htm BodyMedia Inc. (2009). Know Your Body. Change Your Life. Retrieved 25. May, 2009, from http:// www.bodymedia.com/ Boland, P. (2007). The Emerging Role of Cell Phone Technology in Ambulatory Care. The Journal of Ambulatory Care Management, 30(2), 126–133.

147

Wireless and Mobile Technologies Improving Diabetes Self-Management

BrotherSoft.com. (2008). GlucoControl 3.0 Description. Retrieved 15. May, 2009, from http:// www.brothersoft.com/glucocontrol-27960.html Business, P. A. (2008). Diabetes telemedicine system tested. Retrieved 9. May, 2008, from http://www.hospitalhealthcare.com/default.asp? title=Diabetestelemedicinesystemtested&page= article.display&article.id=9301 Calfas, K. J., Sallis, J. F., Zabinski, M. F., Wilfley, D. E., Rupp, J., & Prochaska, J. J. (2002). Preliminary Evaluation of a Multicomponent Program for Nutrition and Physical Activity Change in Primary Care: PACE1 for Adults. Preventive Medicine, 34(2), 153–161. doi:10.1006/pmed.2001.0964 Carroll, A. E., Marrero, D. G., & Downs, S. M. (2007). The HealthPia GlucoPack Diabetes Phone: A Usability Study. Diabetes Technology & Therapeutics, 9(2), 158–164. doi:10.1089/ dia.2006.0002 Cho, J. H., Lee, H. C., Lim, D. J., Kwon, H. S., & Yoon, K. H. (2009). Mobile communication using a mobile phone with a glucometer for glucose control in Type 2 patients with diabetes: as effective as an Internet-based glucose monitoring system. Journal of Telemedicine and Telecare, 15(2), 77–82. doi:10.1258/jtt.2008.080412 Confidant International LLC. (2007). The CONFIDANT Diabetes Solution. Retrieved 4. June, 2009, from http://www.confidantinc.com/solution_applications.htm#diabetes Diabetech, L. P. (2008). GlucoMON - How it Works. Retrieved 18. May, 2009, from http:// mygluco.com/glucomonhowitworks Dicks, K., Kent, R., Tripp, R., Bartlett, T., & Crosley, T. (2008). USA Patent No. Digital Altitudes, L. L. C. (2009). Diabetes Software for iPhone and iPod Touch. Retrieved 16. May, 2009, from http://www.diabetespilot.com/ iphone/index.html

148

Dinesen, B., & Andersen, P. E. R. (2006). Qualitative evaluation of a diabetes advisory system, DiasNet. Journal of Telemedicine and Telecare, 12, 71–74. doi:10.1258/135763306776084329 Donnelly, M. P., Nugent, C. D., Craig, D., Passmore, P., & Mulvenna, M. (2008). Development of a cell phone-based video streaming system for persons with early stage Alzheimer’s disease. Paper presented at the 30th Annual International IEEE EMBS Conference. Retrieved 20 July 2009, from http://www.ncbi.nlm.nih.gov/entrez/utils/ fref.fcgi?PrId=6100&itool=AbstractPlus-def& uid=19163921&db=pubmed&url=http://dx.doi. org/10.1109/IEMBS.2008.4650418 Entra Health Systems. (2009). The New Way to Manage Your Diabetes. Retrieved 11. June, 2009, from http://www.myglucohealth.net/glucometer. asp European Commission. (2005, Feb.). Information Society Technologies, 2005-06 Work Programme. Retrieved 14. April, 2008, from http://www.cordis. lu/ist/workprogramme/fp6_workprogramme.htm European Patent Office. (2009). esp@cenet - Advanced Search. Retrieved 13. June, 2009, from http:// ep.espacenet.com/advancedSearch?locale=en_EP Faridi, Z., Liberti, L., Shuval, K., Northrup, V., Ali, A., & Katz, D. L. (2008). Evaluating the impact of mobile telephone technology on type 2 diabetic patients’ self-management: the NICHE pilot study. Journal of Evaluation in Clinical Practice, 14(3), 465–469. doi:10.1111/j.1365-2753.2007.00881.x Farmer, A. J., Gibson, O. J., Dudley, C., Bryden, K., Hayton, P. M., & Tarassenko, L. (2005). A Randomized Controlled Trial of the Effect of Real-Time Telemedicine Support on Glycemic Control in Young Adults With Type 1 Diabetes. Diabetes Care, 28(11), 2697–2702. doi:10.2337/ diacare.28.11.2697

Wireless and Mobile Technologies Improving Diabetes Self-Management

Farmer, A. J., Gibson, O. J., Hayton, P. M., Bryden, K., Dudley, C., & Neil, A. (2005). A real-time, mobile phone-based telemedicine system to support young adults with type 1 diabetes. Informatics in Primary Care, 13(3), 171–177. Fensli, R., Gunnarson, E., & Gundersen, T. (2005). A Wearable ECG-recording System for Continuous Arrhythmia Monitoring in a Wireless Tele-HomeCare Situation. Paper presented at the 18th IEEE International Symposium on Computer-Based Medical Systems. from http://portal.acm.org/ citation.cfm?id=1078022.1078147&coll=&dl= Fernandez-Luque, L., Sevillano, J. L., HurtadoNunez, F. J., Moriana-Garcia, F. J., del Rio, F. D., & Cascado, D. (2006). eDiab: A system for monitoring, assisting and educating people with diabetes. Computers Helping People with Special Needs. Proceedings, 4061, 1342–1349. Ferrer-Roca, O., Cardenas, A., Diaz-Cardama, A., & Pulido, P. (2004). Mobile phone text messaging in the management of diabetes. Journal of Telemedicine and Telecare, 10(5), 282–286. doi:10.1258/1357633042026341 Fjeldsoe, B. S., Marshall, A. L., & Miller, Y. D. (2009). Behavior Change Interventions Delivered by Mobile Telephone Short-Message Service. American Journal of Preventive Medicine, 36(2), 165–173. doi:10.1016/j.amepre.2008.09.040 Forjuoh, S. N., Reis, M. D., Couchman, G. R., Ory, M. G., Mason, S., & Molonket-Lanning, S. (2007). Incorporating PDA Use in Diabetes Self-Care: A Central Texas Primary Care Research Network (CenTexNet) Study. Journal of the American Board of Family Medicine, 20(4), 375–384. doi:10.3122/jabfm.2007.04.060166 Franklin, V., Waller, A., Pagliari, C., & Greene, S. (2003). “Sweet Talk”: Text Messaging Support for Intensive Insulin Therapy for Young People with Diabetes. Diabetes Technology & Therapeutics, 5(6), 991–996. doi:10.1089/152091503322641042

Franklin, V., Waller, A., Pagliari, C., & Greene, S. (2006). A randomized controlled trial of Sweet Talk, a text-messaging system to support young people with diabetes. Diabetic Medicine, 23(12), 1332–1338. doi:10.1111/j.14645491.2006.01989.x Frings, L., Wagner, K., Maiwald, T., Carius, A., Schinkel, A., & Lehmann, C. (2008). Early detection of behavioral side effects of antiepileptic treatment using handheld computers. Epilepsy & Behavior, 13(2), 402–406. doi:10.1016/j.yebeh.2008.04.022 FutureWare. (2009). FutureWare - Personal GlucoseTracker. Retrieved 16. May, 2009, from http://www.futurewaredc.com/products/FutureWare_PersonalGlucoseTracker.html Gammon, D., Årsand, E., Walseth, O. A., Andersson, N., Jenssen, M., & Taylor, T. (2005). ParentChild Interaction Using a Mobile and Wireless System for Blood Glucose Monitoring. Journal of Medical Internet Research, 7(5), e57 51-59. García-Sáeza, G., Hernandoa, M. E., MartínezSarrieguia, I., Riglab, M., Torralbaa, V., & Brugués, E. (2009). Architecture of a wireless Personal Assistant for telemedical diabetes care. International Journal of Medical Informatics, 78(6), 391–403. doi:10.1016/j.ijmedinf.2008.12.003 Garmin Ltd. (2008). Forerunner 301. Retrieved 15. May, 2008, from https://buy.garmin.com/shop/ shop.do?cID=142&pID=270 Gerber, B. S., Stolley, M. R., Thompson, A. L., Sharp, L. K., & Fitzgibbon, M. L. (2009). Mobile phone text messaging to promote healthy behaviors and weight loss maintenance: a feasibility study. Health Informatics Journal, 15(1), 17–25. doi:10.1177/1460458208099865

149

Wireless and Mobile Technologies Improving Diabetes Self-Management

Glanz, K., Shigaki, D., Farzanfar, R., Pinto, B., Kaplan, B., & Friedman, R. (2002). Participant reactions to a computerized telephone system for nutrition and exercise counseling. Patient Education and Counseling, 49, 157–163. doi:10.1016/ S0738-3991(02)00076-9 Glucose-Charter. (2009). Introduction to GC Pro. Retrieved 15. May, 2009, from http://glucosecharter.com/ GlucoseOne. (2009). Glucoseone for Palm OS. Retrieved 12. June, 2009, from http://www.glucoseone.com/productpalm.html GlucoTools. (2003). GlucoTools. Retrieved 15. May, 2009, from http://glucotools.sourceforge. net/ Gomez, E. J., Hernando, M. E., Garcıa, A., Pozo, F. D., Cermeno, J., & Corcoy, R. (2002). Telemedicine as a tool for intensive management of diabetes: the DIABTel experience. Computer Methods and Programs in Biomedicine, 69(2), 163–177. doi:10.1016/S0169-2607(02)00039-1 Gomez, E. J., Perez, M. E. H., Vering, T., Cros, M. R., Bott, O., & Garcıa-Saez, G. (2008). The INCA System: A Further Step Towards a Telemedical Artificial Pancreas. IEEE Transactions on Information Technology in Biomedicine, 12(4), 470–479. doi:10.1109/TITB.2007.902162 Halkoaho, A., Kavllo, M., & Pietila, A. M. (2007). Information technology supporting diabetes selfcare: A pilot study. European Diabetes Nursing, 4(1), 14–17. doi:10.1002/edn.70 Hanauer, D. A., Wentzell, K., Laffel, N., & Laffel, L. M. (2009). Computerized Automated Reminder Diabetes System (CARDS): E-Mail and SMS Cell Phone Text Messaging Reminders to Support Diabetes Management. Diabetes Technology & Therapeutics, 11(2), 99–106. doi:10.1089/ dia.2008.0022

150

Hazelwood, A. (2008, October 8). 2008). Using text messaging in the treatment of eating disorders. Nursing Times, 104, 28–29. HealthEngage. (2009). What is HealthEngage. Retrieved 14. May 2009, 2009, from http://www. healthengage.com/ HealthPia USA. (2008). How it Works. Retrieved 16. May, 2009, from http://healthpia.us/ GlucoPack-HowItWorks.html Henry, M. E. (2005). ZA - South Africa Patent No. The esp@cenet database - Worldwide. HMM Diagnostics GmbH. (2008). The allrounder for ambitious users Retrieved 21. May, 2009, from http://www.smartlab.org/genie/ Holopainen, A., Galbiati, F., & Voutilainen, K. (2007). Use of Smart Phone Technologies to Offer Easy-to-Use and Cost-Effective Telemedicine Services. Paper presented at the ICDS ‘07. First International Conference on the Digital Society, 2007. Retrieved 8. June 2009, from http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber= 4063766&isnumber=4063753 Ichikawa, A. (2006). Japan Patent No. Indigo Byte Systems, L. L. C. (2009). UTS Diabetes software for Palm OS PDA. Retrieved 16. May, 2009, from http://utracksys.com/softwarediabetes/ Insulet Corporation. (2009). System overview of the OmniPod. Retrieved 12. June, 2009, from http://www.myomnipod.com/about-omnipod/ system-overview/index.php Istepanian, R. S., Zitouni, K., Harry, D., Moutosammy, N., Sungoor, A., & Tang, B. (2009). Evaluation of a mobile phone telemonitoring system for glycaemic control in patients with diabetes. Journal of Telemedicine and Telecare, 15(3), 125–128. doi:10.1258/jtt.2009.003006

Wireless and Mobile Technologies Improving Diabetes Self-Management

Jansa, M., Vidal, M., Viaplana, J., Levy, I., Conget, I., Gomis, R., et al. (2006). Telecare in a structured therapeutic education programme addressed to patients with type 1 diabetes and poor metabolic control. Diabetes Research and Clinical Practice, 74, 26/32. Jensen, K. L., & Larsen, L. B. (2007, 10.-12. September). Evaluating the usefulness of mobile services based on captured usage data from longitudinal field trials. Paper presented at the The 4th International Conference on Mobile Technology, Applications, and Systems (Mobility 2007), Singapore. Jensen, K. L., Pedersen, C. F., & Larsen, L. B. (2006, 17-21 July 2006). Towards Useful and Usable Services in Personal Networks. Paper presented at the Third Annual International Conference on Mobile and Ubiquitous Systems: Networking & Services 2006, San Jose, California, USA. Kim, K., Han, S., Lee, S., Kim, J., & Ahn, H. (2006, Aug 27-Sep 01). MDoctor for DM: Development of Ubiquitous Healthcare System for Diabetes Self-Management. Paper presented at the World Congress on Medical Physics and Biomedical Engineering, Seoul, South Korea. Kim, S. I., & Kim, H. S. (2008). Effectiveness of mobile and internet intervention in patients with obese type 2 diabetes. International Journal of Medical Informatics, 77(6), 399–404. doi:10.1016/j.ijmedinf.2007.07.006 Kollmann, A., Riedl, M., Kastner, P., Schreier, G., & Ludvik, B. (2007). Feasibility of a mobile phone-based data service for functional insulin treatment of type 1 diabetes mellitus patients. Journal of Medical Internet Research, 9(5), e36. doi:10.2196/jmir.9.5.e36

Krishna, S., Boren, S. A., & Balas, E. A. (2009). Healthcare via Cell Phones: A Systematic Review. Telemedicine Journal and e-Health, 15(3), 231–240. doi:10.1089/tmj.2008.0099 Kumar, V. S. (2004). The design and testing of a personal health system to motivate adherence to intensive diabetes management. Unpublished M.D. Degree with Honors in a Special Field. Kumar, V. S., Wentzell, K. J., Mikkelsen, T., Pentland, A., & Laffel, L. M. (2004). The DAILY (Daily Automated Intensive Log for Youth) Trial: A Wireless, Portable System to Improve Adherence and Glycemic Control in Youth with Diabetes. Diabetes Technology & Therapeutics, 6(4), 445–453. doi:10.1089/1520915041705893 Kwon, H. S., Cho, J. H., Kim, H. S., Song, B. R., Ko, S. H., & Lee, J. M. (2004). Establishment of blood glucose monitoring system using the Internet. Diabetes Care, 27(2), 478–483. doi:10.2337/ diacare.27.2.478 Lane, C. (2003). Preventive Medicine - A computer game for diabetic kids [Electronic Version]. Spectrum, Summer 2003. Retrieved 20. May 2009, from http://spectrum.mit.edu/issue/2003-summer/ preventive-medicine/ Larizza, C., Bellazzi, R., Stefanelli, M., Ferrari, P., Cata, P. D., & Gazzaruso, C. (2006). The M²DM Project - The Experience of Two Italian Clinical Sites with Clinical Evaluation of a Multi-access Service for the Management of Diabetes Mellitus Patients. Methods of Information in Medicine, 45(1), 79–84. Lazev, A. B., Vidrine, D. J., Arduino, R. C., & Gritz, E. R. (2004). Increasing access to smoking cessation treatment in a low-income, HIV-positive population: The feasibility of using cellular telephones. Nicotine & Tobacco Research, 6(2), 281–286. doi:10.1080/14622200410001676314

151

Wireless and Mobile Technologies Improving Diabetes Self-Management

Lee, G., Tsai, C., Griswold, W. G., Raab, F., & Patrick, K. (2006, April 22 - 27, 2006). PmEB: a mobile phone application for monitoring caloric balance. Paper presented at the CHI’06 Conference on Human Factors in Computing Systems, Montréal, Québec, Canada. Lee, H. J., Lee, S. H., Ha, K. S., Jang, H. C., Chung, W. Y., & Kim, J. Y. (2009). Ubiquitous healthcare service using Zigbee and mobile phone for elderly patients. International Journal of Medical Informatics, 78(3), 193–198. doi:10.1016/j. ijmedinf.2008.07.005 Lee, H. R., Yoo, S. K., Jung, S. M., Kwon, N. Y., & Hong, C. S. (2005). A Web-based mobile asthma management system. Journal of Telemedicine and Telecare, 11, 56–59. doi:10.1258/1357633054461598 Lee, R. G., Hsiao, C. C., Chen, C. C., & Liu, M. S. (2006). A mobile-care system integrated with Bluetooth blood pressure and pulse monitor, and cellular phone. Ieice Transactions on Information and Systems. E (Norwalk, Conn.), 89d(5), 1702–1711. LifeScan Inc. (2009). OneTouch UltraSmart Blood Glucose Monitoring System. Retrieved 17. June, 2009, from http://www.onetouchdiabetes.com/ ultrasmart/index.html Lim, E., Chen, X., Ho, C., Tin, Z., & Sankaranarayanan, M. (2007, October). Smart Phone-Based Automatic QT Interval Measurement. Paper presented at the Computers in Cardiology 2007. Lin, C. C., Chiu, M. J., Hsiao, C. C., Lee, R. G., & Tsai, Y. S. (2006). Wireless health care service system for elderly with dementia. IEEE Transactions on Information Technology in Biomedicine, 10(4), 696–704. doi:10.1109/TITB.2006.874196

152

Liu, W.-T., Wang, C.-H., Lin, H.-C., Lin, S.-M., Lee, K.-Y., & Lo, Y.-L. (2008). Efficacy of a cell phone-based exercise programme for COPD. The European Respiratory Journal, 32(3), 651–659. doi:10.1183/09031936.00104407 Lutes, K. D., Chang, K., & Baggili, I. M. (2006). Diabetic e-Management System (DEMS). Paper presented at the The Third International Conference on Information Technology. Retrieved 30 March 2009, from http://portal.acm.org/citation. cfm?id=1128187 Mamykina, L., Mynatt, E. D., Davidson, P. R., & Greenblatt, D. (2008). MAHI: Investigation of Social Scaffolding for Reflective Thinking in Diabetes Management. Paper presented at the CHI 2008. Retrieved 21. May 2009, Martinson, B. C., Crain, A. L., Sherwood, N. E., Marcia Hayes, Pronk, N. P., & O’Connor, P. J. (2008). Maintaining Physical Activity Among Older Adults: Six Months Outcomes of the Keep Active Minnesota Randomized Controlled Trial. Preventive Medicine, 46(2), 111–119. Microsoft Corporation. (2009). Devices that connect with HealthVault. Retrieved October 19, 2009, from http://www.healthvault.com/personal/ devices.html?type=device Mobile Diabetic Inc. (2009). Products, LogbookFX Diabetic Diary. Retrieved 14. May, 2009, from http://www.mdiabetic.com/pct_products.html Mohan, P., Marin, D., Sultan, S., & Deen, A. (2008). MediNet: personalizing the self-care process for patients with diabetes and cardiovascular disease using mobile telephony. Conference Proceedings; ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Conference, 2008, 755–758.

Wireless and Mobile Technologies Improving Diabetes Self-Management

Mohan, P., & Sultan, S. (2009). MediNet: A Heath Care Management System Based on Mobile Telephony [Electronic Version]. Conferences@ UWI, Caribbean Regional ITS Conference 2009. Retrieved 10. June 2009, from http://ocs.mona. uwi.edu/ocs/index.php/its/2009/paper/view/25 Morak, J., Schindler, K., Goerzer, E., Kastner, P., Toplak, H., & Ludvik, B. (2008). A pilot study of mobile phone-based therapy for obese patients. Journal of Telemedicine and Telecare, 14(3), 147–149. doi:10.1258/jtt.2008.003015 Mündlein, & Koch. (2007). DiabetesProfiler. Retrieved 13. May, 2009, from http://www.handango. com/catalog/ProductDetails.jsp?storeId=2218&p roductId=87853 MYLEstone Health. (2009). Glucose Buddy. Retrieved 14. May, 2009, from http://beta.glucosebuddy.com/ Nike Inc. (2008). Nike+ SportBand. Retrieved 15. May, 2008, from http://nikeplus.nike.com/nikepl us/?ref=emealanding&sitesrc=emealanding#gear Nischelwitzer, A., Pintoffl, K., Loss, C., & Holzinger, A. (2007, Nov 22). Design and development of a mobile medical application for the management of chronic diseases: Methods of improved data input for older people. Paper presented at the 3rd Symposium of the Workgroup HumanComputer Interaction and Usability Engineering of the Austrian Computer Society, Graz, AUSTRIA. Nokia. (2008). Wellness Diary. Retrieved 16. July, 2009, from http://betalabs.nokia.com/betas/view/ wellness-diary Oliver, N., & Flores-Mangas, F. (2006, 12-15 September). MPTrain: A Mobile, Music and Physiology-Based Personal Trainer. Paper presented at the MobileHCI’06, Helsinki, Finland.

Park, J.-Y., & Daly, J. M. (2003). Evaluation of Diabetes Management Software. The Diabetes Educator, 29(2), 255–267. doi:10.1177/014572170302900216 PatelH.IstocE. S.LinJ. E.NarangA. S. (2008). PatentStorm LLC. (2009). What is Patentstorm? Retrieved 13. June, 2009, from http://www.patentstorm.us/faq.html Patrick, K., Raab, F., Adams, M., Dillon, L., Zabinski, M., Rock, C., et al. (2009). A Text Message–Based Intervention for Weight Loss: Randomized Controlled Trial [Electronic Version]. J Med Internet Res, 11, e1. Retrieved 17 July 2009, from http://www.jmir.org/2009/1/e1/ Pentland, A. (2004, May). 2004). Healthwear: Medical Technology Becomes Wearable. Computer, 37, 42–49. doi:10.1109/MC.2004.1297238 Pinnock, H., Slack, R., Pagliari, C., Price, D., & Sheikh, A. (2007). Understanding the potential role of mobile phone-based monitoring on asthma self-management: qualitative study. Clinical and Experimental Allergy, 37(5), 794–802. doi:10.1111/j.1365-2222.2007.02708.x Polar Electro Oy. (2008). Fitnes & Cross-Training. Retrieved 11 May, 2008, from http://www.polar. fi/en/products/fitness_crosstraining Polymap Wireless. (2008). Enhancing the delicery of healthcare information. Retrieved 16. June, 2009, from https://www.polymapwireless.com/ Preuveneers, D., & Berbers, Y. (2008). Mobile Phones Assisting With Health Self-Care: a Diabetes Case Study. Retrieved 20. May 2008, Quinn, C., Gruber-Baldini, A., Shardell, M., Weed, K., Clough, S., Peeples, M., et al. (2009). Mobile diabetes intervention study: Testing a personalized treatment/behavioral communication intervention for blood glucose control. Contemporary Clinical Trials,30(4), 334-346.

153

Wireless and Mobile Technologies Improving Diabetes Self-Management

Quinn, C. C., Clough, S. S., Minor, J. M., Lender, D., Okafor, M. C., & Gruber-Baldini, A. (2008). WellDoc mobile diabetes management randomized controlled trial: Change in clinical and Behavioral outcomes and patient and physician satisfaction. Diabetes Technology & Therapeutics, 10(3), 160–168. doi:10.1089/dia.2008.0283

Rossi, M., Nicolucci, A., Pellegrini, F., Bruttomesso, D., Bartolo, P., & Marelli, G. (2009). Interactive diary for diabetes: A useful and easyto-use new telemedicine system to support the decision-making process in type 1 diabetes. Diabetes Technology & Therapeutics, 11(1), 19–24. doi:10.1089/dia.2008.0020

QuitSmokingMobile.com. (2009). Quit smoking with mobile. Retrieved 18 July, 2009, from http:// www.quitsmokingmobile.com/

Roudsari, A., Zhao, N. Q., & Carson, E. (2004). A web-based diabetes management system. Transactions of the Institute of Measurement and Control, 26(3), 201–222. doi:10.1191/0142331204tm120oa

Rami, B., Popow, C., Horn, W., Waldhoer, T., & Schober, E. (2006). Telemedical support to improve glycemic control in adolescents with type 1 diabetes mellitus. European Journal of Pediatrics, 165(10), 701–705. doi:10.1007/ s00431-006-0156-6 Rao, R. K., Rao, S. K., & Rao, R. K. (2008). US Patent No. Rigla, M., Hernando, M., Gomez, E. J., Brugues, E., Torralba, V., & Garcia-Saez, G. (2006). Glycaemic control improvement with the weekly use of the Guardian together with a wireless assistant for asynchronic remote advice in pump treated type 1 diabetic patients. Diabetologia, 49(Supplement 1), 153–154. Riley, W., Obermayer, J., & Jean-Mary, J. (2008). nternet and Mobile Phone Text Messaging Intervention for College Smokers. Journal of American College Health, 57(2), 245–248. doi:10.3200/ JACH.57.2.245-248 Rodgers, A., Corbett, T., Bramley, D., Riddell, T., Wills, M., & Lin, R. B. (2005). Do u smoke after txt? Results of a randomised trial of smoking cessation using mobile phone text messaging. Tobacco Control, 14(4), 255–261. doi:10.1136/ tc.2005.011577 Rosen, H. (2007). Canada Patent No.

154

Ryan, D., Cobern, W., Wheeler, J., Price, D., & Tarassenko, L. (2005). Mobile phone technology in the management of asthma. Journal of Telemedicine and Telecare, 11, 43–46. doi:10.1258/1357633054461714 Schiel, R., Beltschikow, S., Radon, S., & Kramer, G. (2006). Identification of cardiovascular risk factors and long -term treatment of children and adolescents with overweight and obesity. Diabetologia, 49(Supplement 1), 459–460. Schildt, J., Franke, J., Kirmse, M., & Leitzmann, P. (2003). Germany Patent No. esp@cenet patent database: W. u. Kollegen. Seng, L. F., Foo, M., Kanagalingam, D., Lim, R., Bahadin, J. B., & Leong, T. K. (2007). Enhancing chronic disease management through telecare - the Singapore health services experience. Journal of Telemedicine and Telecare, 13, 73–76. doi:10.1258/135763307783247257 Shishko, E., Mokhort, D., & Garmaev, D. (2006). The role of lifestyle modification in preventing type 2 diabetes in subjects with impaired glucose homeostasis. Diabetologia, 49(Supplement 1), 457–458. SIMpill. (2009). The SIMpill® Medication Adherence System. Retrieved 19 July, 2009, from http:// www.simpill.com/

Wireless and Mobile Technologies Improving Diabetes Self-Management

SINOVO Ltd. & Co. KG. (2009). SiDiary - Version 5. Retrieved 14. May, 2009, from http:// www.sidiary.org/ Smith, B. K., Frost, J., Albayrak, M., & Sudhakar, R. (2007). Integrating glucometers and digital photography as experience capture tools to enhance patient understanding and communication of diabetes self-management practices. Personal and Ubiquitous Computing, 11(4), 273–286. doi:10.1007/s00779-006-0087-2 Ståhl, A., Höök, K., Svensson, M., Taylor, A. S., & Combetto, M. (2009). Experiencing the Affective Diary. Journal of Personal and Ubiquitous Computing, 13(5), 365–378. doi:10.1007/ s00779-008-0202-7 Subramanian, M., & Conley, C. E., F.Rana, O., Hardisty, A., Ali, A. S., Luzio, S., et al. (2008). Novel Sensor Technology Integration For Outcome-Based Risk Analysis In Diabetes [Electronic Version]. Healthcare@Home Webpage. Retrieved 25. May 2009, from http://www.healthcareathome. info/publications.html SugarStats LLC. (2009). SugarStats’ homepage. Retrieved 13. May, 2009, from http://www.sugarstats.com/ SymCare Personalized Health Solutions. (2009). SymCare Homepage. Retrieved 14. May, 2009, from http://www.symcare.com/ t+ Medical. (2007). Information for Customers t+ diabetes. Retrieved 16. May, 2009, from http:// www.tplusmedical.co.uk/information/01Patients-04tplus_diabetes.html Tasker, A. P. B., Gibson, L., Franklin, V., Gregor, P., & Greene, S. (2007). What is the frequency of symptomatic mild hypoglycemia in type 1 diabetes in the young?: assessment by novel mobile phone technology and computer-based interviewing. Pediatric Diabetes, 8(1), 15–20. doi:10.1111/j.1399-5448.2006.00220.x

Tatara, N., Årsand, E., Nilsen, H., & Hartvigsen, G. (2009, February 1-7, 2009). A Review of Mobile Terminal-Based Applications for SelfManagement of Patients with Diabetes. Paper presented at the 2009 International Conference on eHealth, Telemedicine, and Social Medicine (eTELEMED 2009), Cancun, Mexico. Trinity College Dublin. (2008). Mobile Mood Diary. Retrieved 16. July, 2009, from https:// www.cs.tcd.ie/TMH/project/mobile-mood-diary Tsang, M. W., Mok, M., Kam, G., Jung, M., Tang, A., & Chan, U. (2001). Improvement in diabetes control with a monitoring system based on a hand-held, touch-screen electronic diary. Journal of Telemedicine and Telecare, 7(1), 47–50. doi:10.1258/1357633011936138 Tuomilehto, J., Lindstrom, J., Eriksson, J. G., Valle, T. T., Hamalainen, H., & Ilanne-Parikka, P. (2001). Prevention of Type 2 Diabetes Mellitus by Changes in Lifestyle among Subjects with Impaired Glucose Tolerance. The New England Journal of Medicine, 344(18), 1343–1350. doi:10.1056/NEJM200105033441801 Vähätalo, M., Virtamo, H., Viikari, J., & Rönnemaa, T. (2004). Cellular phone transferred self blood glucose monitoring: prerequisites for positive outcome. Practical Diabetes International, 21(5), 192–194. doi:10.1002/pdi.642 Vlaskamp, F., Childs, R., Lemmens, B., Nelissen, H., Arnoldussen, E., & Soede, T. (2007). SensorPhone: Mobile Telephone Based Activity Monitoring System In J. M. A. Gorka Eizmendi, Gerald Craddock (Ed.), Challenges for Assistive Technology (Vol. 20): IOS Press. Wangberg, S. C., Årsand, E., & Andersson, N. (2006). Diabetes Education via Mobile Text Messaging. Journal of Telemedicine and Telecare, 12(Suppl. 1), 55–56. doi:10.1258/135763306777978515

155

Wireless and Mobile Technologies Improving Diabetes Self-Management

Weaver, A., Young, A. M., Rowntree, J., Townsend, N., Pearson, S., & Smith, J. (2007). Application of mobile phone technology for managing chemotherapy-associated side-effects. Annals of Oncology, 18(11), 1887–1892. doi:10.1093/ annonc/mdm354 WellDoc Inc. (2007). WellDoc Homepage. Retrieved 15. May, 2009, from http://www.welldoccommunications.com/ Whittaker, R., Maddison, R., McRobbie, H., Bullen, C., Denny, S., Dorey, E., et al. (2008). A Multimedia Mobile Phone–Based Youth Smoking Cessation Intervention: Findings From Content Development and Piloting Studies [Electronic Version]. Journal of Medical Internet Research, 10, e49. Retrieved 8. June 2009, from http://www. jmir.org/2008/5/e49 Yoon, K.-H., & Kim, H.-S. (2008). A short message service by cellular phone in type 2 diabetic patients for 12 months. Diabetes Research and Clinical Practice, 79(2), 256–261. doi:10.1016/j. diabres.2007.09.007

156

KEY TERMS AND DEFINITIONS BGM: Blood Glucose Monitor, a sensor device for testing the concentration of glucose in the blood. Diabetes: A disease called Diabetes mellitus, a condition where the body either does not produce enough, or does not properly respond to the hormone insulin produced in the pancreas. eHealth: A recent term for healthcare practice which is supported by electronic processes and communication ICT: Information and Communication Technologies, a broad range of technologies, methods for communication and techniques for storing and processing information. Self-Management: Ways that people with different chronic health problems help themselves. Medical Sensor System: A system in which sensor(s) are integrated in an application for a medical purpose. Wireless Short-Range Communication: technologies that enable data and voice communication between both mobile and stationary devices.

157

Chapter 10

Adaptive Multicarrier Frequency Hopping Spread Spectrum Combined with Channel Coding Othman Sidek Universiti Sains Malaysia, Malaysia Abid Yahya Universiti Malaysia Perlis, Malaysia Farid Ghani Universiti Malaysia Perlis, Malaysia R. Badlishah Ahmad Universiti Malaysia Perlis, Malaysia M. F. M. Salleh Universiti Sains Malaysia, Malaysia

ABSTRACT This chapter presents an adaptive Multicarrier Frequency Hopping Spread Spectrum (MCFH-SS) system employing proposed Quasi Cyclic Low Density Parity Check (QC-LDPC) codes instead of the conventional LDPC codes. A new technique for constructing the QC-LDPC codes based on row division method is proposed. The new codes offer more flexibility in terms of girth, code rates and codeword length. Moreover, a new scheme for channel prediction in MCFH-SS system is also proposed. The technique adaptively estimates the channel conditions and eliminates the need for the system to transmit a request message prior to transmitting the packet data. The proposed adaptive MCFH-SS system uses PN sequences to spread out frequency spectrum, reduce the power spectral density and minimize the jammer effects. DOI: 10.4018/978-1-60960-042-6.ch010 Copyright © 2011, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited.

Adaptive Multicarrier Frequency Hopping Spread Spectrum Combined with Channel Coding

INTRODUCTION The frequency hopping spread spectrum (FHSS) system with partial band interference requires appropriate combination of spread spectrum modulation, error correcting codes, diversity technique and decoding method in order to enhance the signal transmission (Berlekamp, 1980). The combination of a diversity technique and forward error correction codes for FHSS communications system offers the most reliable means of crossing the partial-band noise jammer (Berlekamp, 1980). In FHSS technique the transmission bandwidth W Hertz is divided into q non-overlapping frequency slots. After the signal is modulated to an intermediate frequency, the carrier frequency is hopped periodically according to some predesignated code (a pseudo-random sequence) (Don, 2005; Simon et al., 1995). A patent Hedy Lamarr and music composer George Antheil (Don, 2005) for a “Secret Communication System,” in 1942, is based on the frequency hopping concept, with the keys on a piano representing the different frequencies and frequency shifts used in music. In that year, the technology could not be realized for a practical implementation. Lemarr and Antheil incurred a patent for their idea soon after the expiry of the original patent. Then the U.S applied the FHSS technique for military communication systems onboard ships (Hoffman, 2002).The use of FHSS systems has then increased dramatically since 1962. The benefit of a frequency hopping signal is that it intercepts resistant. This feature is extensively used in military communications where the risk of signal jamming or intercept is higher. Nowadays, it is used in the mobile communication industry as a multiple access technique. The frequency hopping communication systems are utilized to handle high capacity data in an urban setting (Don, 2005; Simon et al., 1995). Frequency hopping communication systems play an important role in military communications strategy. FH communication systems offer an enhancement in the per-

158

formance when subjected by hostile interference. FH communication systems also reduce the ability of a hostile observer to receive and demodulate the communications signal. FH communication systems are susceptible to a number of jamming threats, such as noise jammers and narrowband, single or multitone jammers. If all frequency hops are to be jammed, the jammer has to divide its power over the entire hop band. Thus, it needs to lower the amount of the received jamming power at each hop frequency. Unfortunately, if the tone jamming signal has a significant power advantage, reliable communications will not be possible, even when the jamming tones experience fading (Robertson and Sheltry, 1996; Katsoulis and Robertson, 1997). If the FH signal has an ample hop range, received jamming power will be negligible. If a tone jammer is concentrated on a particular portion of the FH bandwidth, its power may adversely impact communications. A likely anti-jamming strategy is used as a narrow bandstop filter to remove the tone from the received signal spectrum (Don, 2005). Another method based on the undecimated wavelet packet transform (UWPT) isolates the narrowband interference using frequency shifts to confine it to one transform subband (Perez et al., 2002). This technique is a robust to avoid interference and is suitable for FHSS systems.

Frequency Hopping Over Direct Spread Spectrum The fundamental difference between direct spread Pseudonoise (PN) and frequency hop is that the instantaneous bandwidth and the spread bandwidth are identical for a direct spread PN system. While for a frequency hop system the spread bandwidth can be and is typically far greater than the instantaneous bandwidth. For the frequency hop system, all the anti-jamming (AJ) systems processing gain depends upon the number of available frequency slots. The spread bandwidth is generally equal to the frequency excursion from the lowest available

Adaptive Multicarrier Frequency Hopping Spread Spectrum Combined with Channel Coding

frequency slot to the highest available frequency slot. The instantaneous bandwidth, in turn, is determined by the hopping rate or symbol rate, whichever is greater. For the direct spread PN system, the spread bandwidth is limited to the instantaneous bandwidth or the rate at which the PN sequence is clocked. Theodore (2001) at Virginia Tech conducts a research project comparing the effects of interference on wireless LANs using Direct Sequence Spread Spectrum (DSSS) and FHSS technology with the 802.11 and 802.11b standards. The experimental results prove that FHSS is superior to DSSS in high interference settings. Experimental analysis shows that DSSS systems are suffered a performance degradation of 27-45% while FHSS systems are degraded only by 7-8%. The following characteristics show that FHSS overcomes DSSS (Baer, 1992; Cheun and Stark, 1995; Hoffman, 2002; Spasojevic et al., 2000). •

•

• • •

•

Throughput: Point-to-point throughput is variable between both DSSS and FHSS products. Protocols for DSSS throughput sacrifice mobility and roaming performance, but FHSS provides greater power, signal efficiency, mobility, and immunity from multipath interference. Interception: DSSS data is easier to intercept than FHSS data. Constant hopping of FHSS signals make it less susceptible to interception and interference. Power: FHSS radios use less power than DSSS. Efficiency: FHSS can provide up to four times more network capacity than DSSS. Mobility: FHSS products provide better mobility, are smaller, lighter, and consume less power. Unlike DSSS, FHSS incorporates roaming without sacrificing throughput and scalability. Immunity from Multipath Interference: Multipath interference is caused by signals that bounce off from the walls, doors, or

other objects so that signals arrive at the destination at different time. This problem is automatically avoided by FHSS. FHSS simply hops to a different frequency whenever the channel is attenuated. DSSS however, is not capable of overcoming this effect.

Low Density Parity Check Codes LDPC codes were neglected for a long time since their computational complexity for the hardware technology was high. LDPC codes have acquired considerable attention due to its near-capacity error execution and powerful channel coding technique with an adequately long codeword length (MacKay, 1999). LDPC codes have several advantages over Turbo codes. In the decoding of Turbo code it is difficult to apply parallelism due to the sequential nature of the decoding algorithm, while in LDPC decoding can be accomplished with a high degree of parallelism to attain a very high decoding throughput. LDPC codes do not need a long interleaver, which usually causes a large delay in turbo codes. LDPC codes can be constructed directly for a desired code rate. In case of turbo codes, which are based on Convolutional codes, require other methods such as puncturing to acquire the desired rate. DPC codes are in the category of linear codes. They cater near capacity performance on a large data transmission and storage channels. LDPC codes are rendered with probabilistic encoding and decoding algorithms. LDPC codes are designated by a parity check Hmatrix comprising largely 0’s and has a low density of 1’s. More precisely, LDPC codes have very few 1’s in each row and column with large minimum distance. In specific, a (N,j,k) low-density code is a code of block length N and source block length k. The number of parity checks is delimitated as m=N-k. The parity check matrix weight (number of ones in each column or row) or LDPC codes can be either regular or irregular. LDPC can be regular

159

Adaptive Multicarrier Frequency Hopping Spread Spectrum Combined with Channel Coding

if the number of ones is constant in each column or row and gets irregular with a variable number of ones in each column or row. A regular LDPC code is a linear block code whose parity-check matrix H constitutes exactly J1’s in each column 1’s in each row, with and exactly k = j N m the code rate R = 1 − j . k There is a demand to design and develop LDPC codes over a wide range of rates and lengths with efficient performance and reduced hardware complexity. Properly-designed LDPC codes execute very close to the Shannon’s theoretical limit and lower design complexity of the encoder and decoder. In view of this, LDPC codes are being projected as the channel coding solution for future modern digital communication systems. Quasi-Cyclic LDPC codes form a large class of codes with nice encoding and decoding properties, these have been comprehensively premeditated in the sense that their hardware is both cheap and easy to implement. Since both encoding and decoding require less memory, these codes have many advantages for hardware and software implementations. This memory advantage is catered by being able to illustrate the matrices employing a series of short polynomials. In this work, the new QC-LDPC codes flexible in terms of large girth, code rates and codeword length are developed. The method is derived from the basic idea of Bit-Filling (BF) and Progressive Edge-Growth (PEG) algorithms as proposed by Campello et al. (2001) and Hu et al. (2001) respectively. In order to keep the row weight to be dependent on the group size, the weight of any column or row is forced to be not smaller than 2 (greater or equal 2). In addition, a new scheme is also developed for determining the degree of variable-nodes distribution which affects the error correcting performance. Both spread spectrum methods carry large volumes of data, but FHSS is superior. FHSS is a very robust technology, it is scalable, mobile,

( )

160

secure, can accommodate overlapping networks, resistance to interference, and with little influence from noises, reflections, other radio stations or other environment factors(Cheun and Stark, 1995; Hoffman, 2002). Furthermore, the number of simultaneously active systems in the same geographic area is considerably higher than the equivalent number for DSSS systems. There have been limited research efforts to date, that incorporate FHSS system with LPDC channel coding scheme. It is revealed from the work (Deo, 1974; Gallager, 1963; Tanner, 1981; Mackay, 1999) that LDPC codes can be used as the alternative to Turbo codes for forward error correction scheme in FHSS system. This work employs MCFH-SS system, since the use of fast frequency hopping (FFH) system may not be feasible for high data rate systems, due to its high speed requirements. The remainder of this chapter is organized as follows. Section 2 presents the design of new QCLDPC codes. This section discusses the new codes in terms of girth, code rates and codeword length. Section 3 describes the design and development of the system employing MCFH-SS together with channel prediction scheme in order to set up the wireless modules, and to make them perform for communication. This section discusses how the PN sequences are used to spread out frequency spectrum, reduce the power spectral density and minimize the jammer effects. While in Section 4, overall proposed system setup and hardware implementation is presented. Conclusion is drawn in section 5. Glossary is provided at the end of chapter.

PROPOSED METHOD FOR GENERATING QC-LDPC CODES The BF algorithm designs a LDPC code by connecting rows and columns of a code one at a time without violating the girth condition. In the BF algorithm, the number of row connections is almost

Adaptive Multicarrier Frequency Hopping Spread Spectrum Combined with Channel Coding

uniformly distributed by first selecting randomly the rows with the least number of connections. The resultant codes are either having a fixed row or fixed column weight. The structure of the row-column connections in the BF algorithm is, therefore, almost random and hence increases the complexity of the decoder. In order to simplify the hardware implementation, the BF algorithm must be modified to incorporate some form of structured decoder interconnections. In the proposed algorithm, the restructuring of the interconnections is developed by splitting the rows into sub rows. Such a division reduces the hardware complexity. The proposed algorithm for the construction of the new QC-LDPC (N,j,k) code is as described in the following steps: The codeword can be encoded as follows: mod 2(Η ×C T ) = 0

(1.1)

where H represents parity check matrix and C denotes codeword. For an efficient encoding the codeword’s rows are splitted into sub rows with respect to group size. C = (b, p1, p2 .. … pk )

(1.2)

where b and p represent the information and parity check bits respectively. The constraint of group’s number on row weight size persists the row-column (RC) connections to generate variety of codes. k

ψ = ∑ i =1 χ

(1.3)

where, Ψ represents group size and χ stands for number of row. In each sub row the number of 1-component is selected in order to maintain the concentrated degree of distribution which results in random selection. Otherwise non-concentrated degree of distribution will appear. H Z  cir H H cir HQC=  cir  Z H cir   I  1 I  n Hcir=    I n j −1  

H cir Z Z H cir Z H cir

H cir H cir H cir

H cir Z  Z Z  Z H cir    (1.4)

… I mk −1           I mk −1n j −1   

(1.5)

where I ij = δij for i=1,2…m and j=1,2..n 0 fori ≠ j  δij =   1fori = j   

(1.6)

   0A … 0A  0A  0A   Z A =      (1.7)    0A  0A     A p×pHQC array is obtained from the aforesaid equations, where Hcir is the permutation matrix with the location vector of the field elements, cyclically shifting codeword by one position. Therefore, each row of Hcir is obtained from shifting the rows of the identity matrix to the left. Hence Hcir is p×p

161

Adaptive Multicarrier Frequency Hopping Spread Spectrum Combined with Channel Coding

Figure 1. Proposed QC-LDPC codes formation

circulant permutation matrix and HQC is p×p array of p×p circulant permutation matrix.

(b) Store shift values of the submatrices (c) UNION (‘rows’)

5. Computing the Eigenvalues of submatrix with respect to Eigenvector.

Referring to the above steps, the following facts are deduced;

Let the submatrix B=cir{b}, where b is the row vector in HQC, b= (b0,b1,……,bp-1)

1. Since HQC is p×p array of permutation and zero matrices, so no odd number of columns can be added to a zero column. The null space of HQC gives a (N,j,k) QC-LDPC codes, rate at least (k-j)/kwhich ignore zero matrix on the main diagonal of HQC. 2. Check nodes are used to form a cyclic graph which is then transformed to a parity-check matrix. Adopting the Algebraic method of Wang, et al., (2008) in determining the girth upper bounds and also finding the code dimensions in a graph given a base matrix for QC-LDPC codes. 3. The proposed QC-LDPC codes structure for more than two groups is shown in Figure 1. I and Ιsft are p×p identity and shifted submatrices, where 0 is p×p zero submatrix. The complexity of directing within groups computes on the transposition employed to connect rows and columns between groups. This modifies handling, when messages are communicated between functioning nodes.

p  p  det Β = ∏ ∑ γ iqbi   q =0   i =0

(1.8)

For each integer q, 0≤q≤p-1, where γ represents the primitive pth root of unity. Let Eigenvalue λ4â‹‹HQC be the transpose of the row vector 1 (1, γ q , γ 2q ,...γ ( p−1)q ) , then the Eigenvector Ψ4 p = bo + γ qb1 +  + γ ( p−1)qbp−1 . After matrix calculation it shows that λ4 is an Eigenvalue of B with Eigenvector Ψ4. Thus Equation (1.8) can be concluded as; p

det Β = ∏ λq q =0

(1.9)

It can be stated that Ψ4 is an Eigenvector corresponding to the Eigenvalue λ4 . 6. Select the rows and (a) Find the row with the least distance

162

Adaptive Multicarrier Frequency Hopping Spread Spectrum Combined with Channel Coding

Figure 2. (a) Transmitter and (b) Receiver of MCFH-SS system with LDPC Encoder and Decoder

MULTICARRIER FREQUENCY HOPPING SPREAD SPECTRUM MODEL The proposed adaptive MCFH-SS system block diagram is shown in Figure 4. The transmitter consists of LDPC encoder, inverse fast Fourier Transform (IFFT), channel interleaver, DPSK modulator and RF oscillator. The reverse opera-

tions are applicable at the receiver, except with the additional employment of a coherent detector. Referring to Figure 2, the transmitter initially sends a test packet to receiver prior to transmitting the rest of the packets. If the test packet is readable at receiver, that particular channel will be employed. Otherwise, the channel is considered to be unsuitable and is to be avoided. When the system is subjected to interference, then several test packets are sent over different channels in

163

Adaptive Multicarrier Frequency Hopping Spread Spectrum Combined with Channel Coding

Figure 3. Flowchart of adaptive channel prediction

order to find the good one. This selection process is handled by the channel prediction algorithm as mentioned in the following Section 3.1.

Adaptive Channel Prediction Scheme The performance of the MCFH-SS system can be enhanced by incorporating the adaptive channel prediction scheme to the system. This means the MCFH-SS system responds to the noise and fading by avoiding the channel that is unfit for

164

communication. Usually, a channel is banned only after it has been used to transmit data, which results in retransmission. The principle of blocking the poor channels is based on the prospect that these channels will remain in poor condition for transmission in quite some time. However, the banned channels may fit inadequately to the actual noise and fading. In a severe condition, the system may end up employing the low quality channels regularly and would ban the good ones too frequently and disturbs the performance. Apparently, it is attractive

Adaptive Multicarrier Frequency Hopping Spread Spectrum Combined with Channel Coding

Figure 4. Frequency and subband division of the proposed MCFH-SS system

to mitigate such undesirable consequences. If the system is designed in such a way that it attempts to forecast and ignore the poor channels, better performance can be incurred. This furnishes the need for attempting to forecast the quality of channels. In order to predict the fit channels the system transmits short test packets on channels. If the test packet arrives is readable, the channel will be occupied with a Pseudonoise (PN) code and use for transmission or else, it will be banned.

h

The channel prediction algorithm flow chart is shown in Figure 3 and explains as in the following steps: The probability of a poor channel p must be fixed. P = ∑ p(b0 i, bq ) i =1

q= number of test packets/channel test h= hopping table b= maximum number of packets in the receiver buffer b0= actual number of packets in the receiver buffer Determine the Value function for each possible channel test.

{

}

Eh ,q (b) = min P S + e −rT Eh ,q (b0 )

The Algorithm

h

where the parameters,

(1.10)

(1.11)

where, Eh ,q (b)represents the Value function, the other parameters are mentioned as; q= number of test packets/channel test h= hopping table b= maximum number of packets in the receiver buffer b0= actual number of packets in the receiver buffer

165

Adaptive Multicarrier Frequency Hopping Spread Spectrum Combined with Channel Coding

S= stochastic variable e-rT=discount factor, and 0≤γ≤1 3. Following step 2, check the receive buffer size and the number of best possible extra packets. 4. If all the q tested channels are poor, it is proposed that the radio shall simply use the next untested channel from the hopping table to transmit the frame. 5. Block the ready-to-use channels with PN code in order to keep them off from intentional and unintentional jammers. 6. Updates ready-to-use channels list.

Proposed Adaptive MCFHSS System Description The proposed adaptive MCFH-SS system utilizes a total bandwidth of WHz, and it is divided into L number of segments as shown in Figure 4. These small segments are called the subbands, and each subband has Lh frequency bins. In each signaling interval, there are Lh subcarriers chosen from the L frequency subbands. This is achieved by either using the inverse discrete Fourier transform (IDFT), or can easily be implemented using IFFT operation. Consider an independent identity distributed (i.i.d) transmitted bits sequence of length M. B={b(q)}

(1.12)

The sequence bit values are set to {0,1} with equal probabilities and q=0,…M-1. The system is assumed to employ DPSK modulation scheme. An information bit sequence b is first encoded by LDPC encoder to a code bit sequence c of length N. The encoded sequence then goes through a random channel interleaver. The performance of the LDPC decoder effected by burst errors due to correlated fading channel. The interleaver is therefore expended to eliminate the effect of fading correlation on consecutive bits. The base-

166

band equivalent of the transmitted signal can be written as; ∞

x (q ) = ∑ q =0

L

∑ l =1

2Es bqe

j 2 π( fl ,q +rq fd )m

ϕT m h

(1.13) where x(q) represents the transmitted power for each diversity transmission and E5is the energy of the signal. The parameter bq is the independent identity distributed (i.i.d) transmitted sequence bits. The sequence bit values are set to {0,1} since the system employs DPSK modulation scheme. The parameter fl,q is the hop frequency for each lth diversity transmission of qth symbol. The reciprocal of the bit duration (fd) separates the subchannels. The parameter rq is the i.i.d random variable uniformly distributed over the interval [1,L]. The parameter m is defined as m=t-qTh where Th is the hop duration of qth symbol and the variable ϕT represents the rectangular pulse of durah

tion Th. For DPSK modulation scheme, the phase difference of two consecutive signals is given by; υ(q ) = υ(q − 1) + ∆υ(q )

(1.14)

where, the phase difference ∆υ(q ) ∈ {0, τ } , corresponds to {0,1} of the coded sequence. At the receiver, the received signals undergo the reverse operations of those at the transmitter. First, the discrete Fourier transform (DFT) is used for reconstruction. The transmitted sequence passes through a slow flat fading channel with AWGN so that the received signal sequence is given as follows; Y={y(q)] where the value of q=0,…N-1.

(1.15)

Adaptive Multicarrier Frequency Hopping Spread Spectrum Combined with Channel Coding

The received modulation state y(q)is given by; y(q)=h(q)x(q)+N(q)

(1.16)

where, h(q)is the channel transfer factor with phase φ(q ) and N(q) is the AWGN term with power spectral density of 2σ 2 . The log likelihood ratios (LLR) are calculated based on demodulated received signal, which is then passed to the LDPC decoder after the values are deinterleaved. It is assumed that the h(q) is constant over one symbol interval and has average power equal to the following; Ε{h(q )h(q )*}

(1.17)

If the channel actually varies over the symbol interval TS, this assumption leads to the suboptimal sufficient statistics. For the moment, it is assumed that the channel coefficients h(q)are known to the receiver. The probability density function (PDF) of y(q) over Rayleigh Fading channel is given as; p(y(q ) x (q ), h(q )) =

=

1 e 2πσ 2

p(y(q ), x (q ), h(q )) (1.18) p(x (q ), h(q ))

2    y (q )−x (q )h (q )  −  2   2 σ    



1 e 2πσ 2

p(y(q ), y(q − 1) x (q ), x (q − 1), γ(q ),.γ(q − 1), φ(q ), φ(q − 1)) =

This can be written as (1.20)

p(y(q ), y(q − 1) γ(q ), ∆υ(q )) = p(y(q ), y(q − 1) γ(q − 1), υ(q − 1) = 0, .γ(q ), ∆υ(q ))p(γ(q − 1), υ(q − 1) = 0) +p(y(q ), y(q − 1) γ(q − 1), υ(q − 1) = τ,.γ(q ), ∆υ(q ))p(γ(q − 1), υ(q − 1) = τ )



(1.21) From Equation (1.21) the Log Likelihood ratio Λ(q ) is the given by;

Since,

(1.22)

p(y(q ), y(q − 1) γ(q ), ∆υ(q ))

2

 1   e =   2πσ 2 

(1.19)

When two consecutive independent symbols x(q)and x(q-1)with amplitude γ(q ) and phase difference ∆υ(q ) are transmitted over Rayleigh Fading channel, the conditional joint PDF of the two received consecutive symbols is given as;

2    y (q )−γ (q )x (q )e j φ (q )    −  2  2σ    

2    y (q )−γ (q )x (q )∆υ (q )  −   2 2σ    



(1.23)

Equation (1.22) can be simplified to Λ(q ) =

=

1 2σ 2

 4y(q ){γ(q )∆υ(q )} +     4y(q − 1){γ(q − 1)∆υ(q − 1)}   (1.24)

2  j ∆υ(q ) + y(q − 1)γ(q − 1) ES  y(q )γ(q ) ES e  σ2  (1.25)

167

Adaptive Multicarrier Frequency Hopping Spread Spectrum Combined with Channel Coding

Figure 5. Data packet

Λ(q ) =

2 ES σ2

Re y(q )γ(q ) + y(q − 1)γ(q − 1) (1.26)

Thus, the optimum decision rule that minimizes the bit error probability is given as; x = + ES if Λ(q ) > 0 (1.27) or x = − ES if Λ(q ) < 0 (1.28) Since the probability decreases with increasing Λ(q ) , the optimum selection rule to minimize the bit error probability is to select those pair of groups that provide largest Λ(q ) .

Implementation of MCFHSS Protocols The frequency hopping code hops among the 64 channels (channel 2 to channel 65) which are pseudo-randomly distributed in a 256-bytes constant table embedded in the code. In addition to the address and Cyclic Redundant Check (CRC), each packet consists of 0-25 bytes of data and one byte of control information as shown in Figure 5. In the first frame, default packets are transmitted. The transmitter and receiver test the channels in the hopping table, and find the suitable one or more channels among them for possible communication. Prior to the agreement between the receiver and transmitter upon which channel to be used, the transmitter must first be informed about the condition of test packets at receiver. Therefore, the receiver will include that informa-

168

tion in the ACK message that is sent after the frame has been received. The problem is that the receiver will not know whether the transmitted ACK message at the transmitter is in readable form or not. If the transmitter transmits an ACK message back to the receiver in order to verify that the acknowledgment was OK, the same problem happens, and resultantly a ceaseless series of ACK packets. In this work however, the transmission of multiple ACK messages can be avoided, since it reduces both the available transmission time and the number of channels that can be tested. A flowchart for a transmitter and receiver of the MCFH-SS system is shown in Figure 6. Referring to the flowchart of transmission packets, as shown in Figure 6 (a), each time the transmitter (of the transceiver) transmits a test packet through a channel; it waits for an acknowledgment (ACK) on that particular same channel. If the ACK message is received within a pre-defined time-out (3ms in this system), this channel is placed in the list of ready-to-use channels. The ready-to-use channels are blocked with PN codes in order to keep them away from intentional or unintentional jammers and only accessible to the proposed system. The transmitter and receiver update the ready-to-use and ban channels list after each pre-defined timeout. Before sending a data packet, the ready-to-use channels on the channel lists of transmitter and receiver are read out sequentially and then they are used for transmission. If the channels in the list are corrupted or busy, then the procedure of transmitting test packets as shown in Figure 6 (a) is repeated. There is no uncertainty about which

Adaptive Multicarrier Frequency Hopping Spread Spectrum Combined with Channel Coding

Figure 6. Flowchart of packets (a) Transmitter of the proposed MCFH-SS system (b) Reciever of the proposed MCFH-SS system

channel to be used for the next frame, and which channels to be tested next. After each received frame, the receiver transmits the ACK message to the transmitter as shown in Figure 6 (b). A copy of the receiver’s channels list, the number of packets in the receive buffer and receiver’s estimate values of p are enclosed in that ACK message.

OVERALL SYSTEM SETUP AND IMPLEMENTATION USING VECTOR SIGNAL ANALYZER Vector Signal Analyzer (VSA) provides flexibility in making so many kinds of measurements especially valuable for top-level (transmitter/ receiver characterization) system work. VSA software provides a complete vision to a user in

169

Adaptive Multicarrier Frequency Hopping Spread Spectrum Combined with Channel Coding

Figure 7. Overall system scheme

a perfect communication system to investigate different aspects of the same signal at the same time. Familiar tools such as spectrum analyzers with demodulation scheme indicate the existence of problem, but not able to detect the cause of it. The VSA provides the tools to identify the root cause of the problem and to examine frequent prospects such as changing phase, magnitude and frequency. There are several sources to get data such as multiple supported hardware platforms, recorded files or, from the Matlab software by means of general purpose interface bus (GPIB) and a signal generator. The VSA performance also depends on PC processing speed in order to know the functionality of a complete system. The VSA software provides an Application Programming Interface to its Component Object Model or (COM API). Herein the COM API will be used to program and control the VSA software and signal generator from Matlab. The overall system setup is shown in Figure 7. The workstation installed with Matlab program

170

and VSA software is connected to signal generator by means of a GPIB bus. It is then connected to a hardware platform and spectrum analyzer. The hardware platform output is connected to the spectrum analyzer RF in and serially connects to the workstation. The whole system is fully controlled by a workstation. Matlab is used to discover, identify, modify and control all the equipments and VSA software by means of Application Programming Interface to its Component Object Model (COM API). Once the signal is produced on Matlab, it is sent to the signal generator by means of GPIB commands. In order to send the signal, the load SCPI command is used, but it must be adapted to the dynamic range of the signal generator. First of all, identification for GPIB devices must be done in order to create an object in Matlab for possible communication with them. Later, the signal generator and spectrum analyzer alignment must be done. At the first one, aspects as the carrier power and frequency of the signal are firstly

Adaptive Multicarrier Frequency Hopping Spread Spectrum Combined with Channel Coding

adjusted. Then, the signal can be sent from Matlab by the esg_darb GPIB command. The ARB (arbitrary waveform) is then activated. It allows selecting the signal sent by the user. Finally, the RF output is changed to ON with the objective of allowing the Signal Generator to launch noise signal to the hardware platform. An important issue arises with the Automatic Level Control (ALC). When the ALC is set to ON, the internal level detector checks the output level and does not fluctuate much from the assigned amplitude value. The output level of ALC is directly proportional to the amplifier output gain. However, in some cases when the ALC is incapable to maintain the output level, the unlevel message appears to report the unleveled condition to the user. Thus, in order to allow the user to alter the amplitude from Matlab, the ALC should be OFF to see the correct amplitude value. The center frequency and span must be adjusted in order to observe the signal on spectrum analyzer. Otherwise, the signal will be seen on the PC with the help of VSA software.

SUMMARY This chapter presents an adaptive MCFH-SS system employing proposed QC-LDPC codes instead of the conventional LDPC codes. A new technique for constructing the QC-LDPC codes based on row division method is proposed. The new codes offer more flexibility in terms of girth, code rates and codeword length. In this method of code construction, the rows are used to form as the distance graph. They are then transformed to a parity-check matrix in order to acquire the desired girth. Moreover, a new scheme for channel prediction in MCFH-SS system is also proposed. The technique adaptively estimates the channel conditions and eliminates the need for the system to transmit a request message prior to transmitting the packet data. The proposed adaptive MCFH-SS system uses PN sequences to spread out frequency

spectrum, reduce the power spectral density and minimize the jammer effects. In the last section of this Chapter, the hardware implementation of the proposed system is described. The hardware platform consists of communication development kit that is interfaced with Xilinx Spartan-3E development board. The user can easily program the device with custom protocols for use in their end product or for general product development. The workstation installed with Matlab program and VSA software is connected to a signal generator by means of a GPIB bus. Then, it is connected also to the hardware platform and spectrum analyzer.

Future Work Some concrete ideas for future works are as the following; 1. The larger the system bandwidth, the shorter the pulses must be. Therefore, for direct spread spectrum system, WSS is restricted to few hundred MHz, by reason of hardware limitations. This means that the processing gain of the system is also inadequate. Additionally, the system bandwidth must be contiguous. The employment of FHSS system overcomes these restrictions. Implement the proposed system, presented in this thesis for underwater communications in shallow water. Thus, this reveals many in the challenges to determine the performance of the system, such as the propagation delays, very limited available bandwidth, the effect of Doppler spread and Doppler shift. 2. Design and implement the irregular QCLDPC codes in cooperative communication with more complex semi-parallel architecture, with reduced number of iterations and high decoding throughput. The mainly issues in cooperative communication are the cost of signaling overhead required for device cooperation, the design of cooperation

171

Adaptive Multicarrier Frequency Hopping Spread Spectrum Combined with Channel Coding

schemes with limited complexity, and exhibiting robustness to a lack of exact channel state information and the better evaluation of system-level gains.

REFERENCES Baer, H. P. (1992). Interference Effects of Hard Limiting in PN Spread-Spectrum Systems. IEEE Transactions on Communications, 5, 1010–1017.

Katsoulis, G., & Robertson, R. C. (1997). Performance Bounds for Multiple Tone Interference of Frequency-hopped Noncoherent MFSK Systems. Proceedings IEEE Military Communications Conference, p.307-312. Mackay, D. J. (1999). Good error-correcting codes based on very sparse matrices. IEEE Transactions on Information Theory, 45(2), 399–431. doi:10.1109/18.748992

Berlekamp, E. R. (1980). The technology of error correcting codes. Proceedings of the IEEE, 68, 564–593. doi:10.1109/PROC.1980.11696

Pérez, J. J., Rodriguez, M. A., & Felici, S. (2002). Interference Excision Algorithm for Frequency Hopping Spread Spectrum Based on Undecimated Wavelet Packet Transform. Electronics Letters, 38(16), 914–915. doi:10.1049/el:20020583

Campello, J., Modha, D. S., & Rajagopalan, S. (2001).Designing LDPC codes using bit-filling. Proceedings of IEEE International Conference on Communications (ICC ’01), Vol. 1, p. 55-59, Helsinki, Finland.

Robertson, R. C., & Sheltry, J. F. (1996). Multiple Tone Interference of Frequency Hopped Noncoherent MFSK Signals Transmitted Over Ricean Fading Channels. IEEE Transactions on Communications, 44(7), 867–875. doi:10.1109/26.508306

Cheun, K., & Stark, W. E. (1995). Performance of FHSS Systems Employing Carrier Jitter against One-Dimensional Tone Jamming. IEEE Transactions on Communications, 43(10), 2622–2629. doi:10.1109/26.469437

Simon, M. Omura, J. Scholtz, R., & Levitt. B. (1994). Spread Spectrum Communications handbook. McGraw–Hill Inc., revised edition.

Deo, N. (1974). Graph Theory with Applications to Engineering and Computer Science. Englewood Cliffs, NJ: Prentice Hall. Gallager, R. G. (1963). Low-Density Parity-Check Code. Cambridge, MA: MIT Press. Hu, X. Y. Eleftheriou, E. & Arnold, D. M. (2001). “Progressive edge-growth Tanner graphs,” IEEE Global Telecommunications Conference, p. 995–1001.

172

Spasojevic, Z., & Burns, J. (2000). Performance Comparison of Frequency Hopping and Direct Sequence Spread Spectrum Systems in the 2.4GHz Range (pp. 426–430). Aegis Systems Ltd. Tanner, R. M. (1981). A recursive approach to low complexity codes. IEEE Transactions on Information Theory, IT-27, 533–547. doi:10.1109/ TIT.1981.1056404 Wang, Y., Yedidia, J. S., & Draper, S. C. (2008). Construction of high-girth QC-LDPC codes. 5th International Symposium on Turbo Codes and Related Topics, p.180-185.

Adaptive Multicarrier Frequency Hopping Spread Spectrum Combined with Channel Coding

APPENDIX 1. Code Size The code size defines the dimensions of the parity check matrix(M × N ) . Occasionally the term code length is used referring to n. usually a code is defined employing its length and row column weight in the form(N , j, k ) . M can be deduced from the code parameters N , j and k . It has been determined that long codes execute better than shorter codes but need more hardware implementation resources. 2. Code Weights and Rate The rate of a code R , is the number of information bits over the total number of bits transmitted. Higher row and column weights result in extra computation at each node because of many incoming messages. Nevertheless, if many nodes contribute in estimating the probability of a bit the node accomplishes a consensus faster. Higher rate indicate fewer redundancy bits. Namely, more information data is transmitted per block resulting in high throughput. Though, low redundancy implies less protection of bits and thus less decoding performance or higher error rate. Low rate codes have more redundancy with less throughput. More redundancy results in more decoding performance. But, low rate may have poor performance with a small number of connections. LDPC codes with column-weight of two have their minimum distance increasing logarithmically with code size as compared to a linear increase for codes with column-weight of three or higher. Column weights higher than two are generally employed but carefully designed irregular codes could have better performance. 3. Code structure The structure of a code is ascertained by the pattern of connection between rows and columns. The connection pattern ascertains the complexity of the communication interconnect between check and variable processing nodes in an encoder and decoder hardware implementations. Random codes do not chase any predefined or known pattern in row-column connections. Structured codes on the hand have a known interconnection pattern. 4. Number of iterations The number of iterations is the number of times the received bits are estimated before a hard decision is made by the decoding algorithm. A large number of iterations may make sure decoding algorithm convergence but will increase decoder delay and power consumption. The number of corrected errors normally decreases with an increasing number of iteration. Binary-input memoryless symmetric channel, the term; Binary-input indicates that the data transmitted is a discrete symbol from Galois Field F2 i.e., {0; 1},

173

Adaptive Multicarrier Frequency Hopping Spread Spectrum Combined with Channel Coding

Memoryless implies that each symbol is affected independently by the noise in the channel and, Symmetric indicates the noise in the channel affects the 0s and 1s in the same way. As, there is no direct connection between any two variable nodes in the factor graph, the decoding of the code symbol can be mulled over on each variable node independently. It shows that the local decoding operations can be executed independently at both variable and check nodes side in the factor graph. If the factor graph is cycle free, then the independency supposition is valid while if the factor graph has cycles, then the assumption will be valid for few iterations until the messages have moved around the entire cycles. 5. Cycle A cycle in a distance graph is formed by a path edges or vertices starting from a vertex νx and ending at νx . No more than two vertices forming the cycle belong to the same column. 6. Girth The girth g, is the smallest cycle in the graph. A cycle of length of g in the graph corresponds to a cycle of length 2g in the matrix form. 7. Minimum distance The Hamming weight of a codeword is the number of 1’s of the codeword. The Hamming distance between any two code words is the number of bits with which the words differ from each other, and the minimum distance of a code is the smallest Hamming distance between two code words. The larger the distance the better the performance of a code. Very long and large girth LDPC codes tend to have larger minimum distances. A better code could be ascertained by employing minimum distance as a measure.

174

175

Chapter 11

Key Adoption Challenges and Issues of B2B E-Commerce in the Healthcare Sector Chad Lin Curtin University, Australia Hao-Chiang KoongLin National University of Tainan, Taiwan Geoffrey Jalleh Curtin University, Australia Yu-An Huang National Chi Nan University, Taiwan

ABSTRACT Although B2B e-commerce provides healthcare organizations a wealth of new opportunities and ways of doing business, it also presents them with a series of challenges. B2B e-commerce adoption remain poorly understood and it is also a relatively under-researched area. Therefore, case studies were conducted to investigate the challenges and issues in adopting and utilizing B2B e-commerce systems in the healthcare sector. The major aims of this study are to: (a) identify and examine main B2B e-commerce adoption challenges and issues for healthcare organizations; and (b) develop a B2B e-commerce adoption challenges and issues table to assist healthcare organizations in identifying and managing them.

INTRODUCTION Business-to-business (B2B) e-commerce has achieved considerable importance for the global economy in recent years. B2B e-commerce is an initiative which requires the participation of multiple firms making some sort of complementary DOI: 10.4018/978-1-60960-042-6.ch011

investment to enable one another’s EC strategy (Burn and Ash, 2005; Lin et al., 2007). Currently, the most popular electronic commerce applications used by organizations in their B2B relationships are EDI, web-forms, XML, and other varieties of Internet initiatives such as electronic marketplace and B2B web portals (Chan and Swatman, 2003; Standing and Lin, 2007). Depending on the

Copyright © 2011, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited.

Key Adoption Challenges and Issues of B2B E-Commerce in the Healthcare Sector

characteristics of the products, market variability, market volatility, and continuity of the business relationship between the channel partners the B2B e-commerce markets can take different forms, such as electronic hierarchies, cooperative arrangements, and spot markets. Although healthcare is the biggest service industry worldwide, it has historically lagged behind other industries in the use of information technology (IT) such as B2B e-commerce (Lee and Shim, 2007; Parente, 2000; Wickramasinghe et al., 2005). Despite high expectations for realizing benefits from B2B e-commerce in the healthcare, its use remains poorly understood (Bhakoo and Sohal, 2008; Davidson and Heslinga, 2007) and is also a relatively under-researched area (Chiasson et al., 2007). Little is known about why certain organizations within the healthcare industry have adopted and implemented IT such as B2B e-commerce successfully while many others have not (Heeks, 2006; Miller, 2003). Until recently, Australian healthcare supply chain has been still a mainly paper based system with manual processing (GS1, 2004). Effective utilization of B2B e-commerce in the healthcare industry has the potential to lead to an increased accessibility to healthcare providers, improved work-flow efficiency, a higher quality of healthcare services, improved inventory management, and reduction in healthcare costs and medical errors (Bhakoo and Sohal, 2008; Heeks, 2006; Lin et al., 2008a, 2008b).

BENEFITS OF B2B E-COMMERCE IN HEALTHCARE The adoption of B2B e-commerce by healthcare organizations includes online activities such as biotechnology online transactions between, for example, hospitals and their suppliers and the sales of medical products and services via emarketplace (Parente, 2000; Standing et al., 2008). It enables healthcare organizations to minimize

176

their procurement costs and assists their suppliers to sell via an efficient marketing channel (Parente, 2000). B2B e-commerce allows health organizations’ business partners to access their internal business systems via the Internet. Some of the major benefits of B2B e-commerce specifically to healthcare organizations are: (1) it gives medical suppliers access to huge number of new customers and suppliers (Suomi et al., 2001); (2) it can reduce healthcare organizations’ costs such as procurement costs, inventory holding, and search costs (Bhakoo and Sohal, 2008); (3) it allows healthcare organizations to compare the partners effectively and easily the change traditionally rigid procurement contracts (Suomi et al., 2001); (4) it provides an efficient and effective channel for medical information exchange and sharing among healthcare organizations (Suomi et al., 2001); (5) it helps to provide accurate and timely business information and streamline orders and payments (CR Group, 2002); and (6) it offers an effective channel and platform for suppliers and customers to trade and communicate (Suomi et al., 2001).

GOVERNMENT/INDUSTRY B2B E-COMMERCE INITIATIVES IN HEALTHCARE There are several subtle as well as obvious differences between healthcare industry and other industries. For instance, the healthcare industry, unlike other industries, does not generally face stiff competition. In addition, organizations in other industries expect to maximize their profits whereas many stakeholders (e.g hospital patients and governments) in the healthcare organizations expect better and efficient products and services (Suomi et al., 2001). Therefore, healthcare organizations must look at B2B e-commerce from a strategic perspective and measure its contribution because it can assist them in developing and controlling strategic, tactical, and operational plans that define the appropriate role of B2B e-

Key Adoption Challenges and Issues of B2B E-Commerce in the Healthcare Sector

commerce in the organization (McGaughey, 2002). However, the difficulty of dealing with challenges and benefits are some of the major problems in B2B e-commerce adoption (Teo and Ranganathan, 2004) among healthcare organizations. In the past, a large number of healthcare organizations have been conducting business via EDI systems and these systems often link customers to only one healthcare supplier. In addition, these proprietary and restrictive EDI systems are expensive to develop and maintain. Driven by concerns about escalating procurement costs in the industry, several B2B e-commerce initiatives have been launched by Australian Federal and State governments during the last decade to develop better ways of, among other things, ordering and procuring medical supplies electronically within the Australian healthcare industry – the HUGLIT group, the Monash Project (GS1, 2005), National E-Health Transition Authority (NEHTA, 2004), National Supply Chain Reform Task Force (NSCRTF, 2008), Project Electronic Commerce and Communication (PECC) and the Pharmaceutical Extranet Gateway (PEG) (More and McGrath, 2002), and Queensland Health pilot e-Procurement project (Morgan, 2004). The main objectives of these initiatives were to reduce costs of supply chain, save time, increase market share, and improve business relationships for the healthcare organizations (GS1, 2005; More and McGrath, 2002; NEHTA, 2004; NSCRTF, 2008). These initiatives attempted to assist in supply chain management as well as to set standards (e.g. numbering structures and information exchange protocols) for B2B e-commerce in the healthcare industry in Australia. According to Bushell (2000), some of the benefits for having in place standards for the supply chain management process within the healthcare industry include: •

establishing a single identifiers for all transaction types that need to be shared and communicated across borders;

•

•

•

having mandatory IT operability for sharing information inside and outside a healthcare organization; integrating supply chain functionality seamlessly into the other aspects of healthcare delivery; and creating an effective billing and claims systems that is also able to meet the wider requirements of both the clinical environment as well as the administrative one.

MAJOR PROBLEMS OF B2B E-COMMERCE IN HEALTHCARE Despite the best intention and the substantial amount of resources being spent by the government and to a lesser extent, the healthcare industry, these projects/initiatives had only achieved partial success. Organizations in the healthcare industry are still under increasing pressure to adapt and leverage B2B e-commerce to manage their medical supply chains. This is due to the fact that potential for realizing significant benefits for implementation of B2B e-commerce does not always appear to have been enough to provide impetus for extensive adoption by healthcare organizations (Bhakoo and Sohal, 2008). Although B2B e-commerce provides healthcare organizations a wealth of new opportunities and ways of doing business, it also presents them with a series of challenges (Laudon and Laudon, 2004; Lin et al., 2007; Lynch and Beck, 2001) such as: •

•

•

Many new B2B e-commerce business models are difficult to evaluate and therefore, have yet to prove enduring sources of profit; Web-enabling business processes for B2B e-commerce requires far-reaching organizational change; The legal environment for B2B e-commerce has not yet solidified and organiza-

177

Key Adoption Challenges and Issues of B2B E-Commerce in the Healthcare Sector

•

tions must be vigilant about establishing trust, security and consumer privacy; and Many healthcare organizations often take the short-term view of examining their electronic commerce success by only looking at the potential advantages of IT use while at the same time are unaware of the factors that may hinder the benefits attainment in the long term such as managing the relationship between the justification of the B2B e-commerce initiatives to stakeholders.

Moreover, it appears that our current understanding of B2B e-commerce adoption and management processes in healthcare has little impact on organizational practice, making factors such as power and politics and resource constraints extremely difficult to identify and measure (Howcroft and McDonald, 2004; Lin et al., 2007). Failure to identify and manage challenges and issues can have detrimental consequences on healthcare organizational performance. Some of the major problems associated with B2B e-commerce in healthcare are: (a) it is difficult to integrate different databases among different suppliers and buyers in order to create new data resources and to increase productivity and effectiveness (CR Group, 2002; Parente, 2000); (2) healthcare is different from other industries due to the high level of government regulations and investments. Compared with other industries, the healthcare industry as a whole generally possesses a relatively underdeveloped IT infrastructure and expertise such as B2B e-commerce (Herzlinger, 2006; Porter and Teisberg, 2004); (3) healthcare organizations are less accustomed to work with new technologies and there is a need to obtain not only technologies but also other relevant expertise and skills to handle them (Parente, 2000); and (4) organizations within the healthcare industry need to cooperate with each other to comply with the requirement of both internal (e.g. pharmacists, healthcare staff, employees) and external stake-

178

holders (governments, suppliers, insurers). Thus, business relationships among healthcare organizations are often subject to a variety of factors.

RESEARCH METHODS Case studies were conducted with participants (i.e. IT managers, IT procurement managers, and/or CIOs) from several Australian healthcare organizations. The questions asked related to their B2B e-commerce investments, B2B e-commerce adoption challenges and barriers, and key factors and issues in B2B e-commerce adoption and implementation. Other data collected included contract documents, planning documents and minutes of relevant meetings. Transcripts were coded and analyzed. Semi-structured interviews were used to gain a deeper understanding of issues. Qualitative content analysis by Miles and Huberman (1994) was used to analyze the data from the case study. The analysis of the case study results was conducted in a cyclical manner and the results were checked by other experts in the field. Finally, the guidelines set out by Klein and Myers (1999) for conducting and evaluating interpretive case studies were also followed in an attempt to improve the quality of this research.

RESEARCH FINDINGS Several challenges and issues arose from the analysis of case study data and they are presented below. These are presented in Table 1.

Intra-organizational and Resources Top Management Support Obtaining top management commitment throughout the adoption stage was found to be critical to the success of the IT investments in e-commerce (Power, 2004). Most organizations interviewed

Key Adoption Challenges and Issues of B2B E-Commerce in the Healthcare Sector

Table 1. Types of key challenges and issues for healthcare organizations adopting and implementing B2B e-commerce Types of Challenges

Key Issues

Intra-organizational and Resources

• Top management support • Alignment of organizational goals with business objectives • User involvement & participation • Financial & managerial resources

Industry Inter-organizational and Supply Chain

• Government regulations & standardized B2B protocols • Role of Medicare Australia • Supply chain management • Interoperability

indicated that their senior executives had provided sufficient management leadership as well as obtained necessary organizational commitment towards the adoption of B2B e-commerce within their organizations. Most senior managers were very enthusiastic about their IT investments in B2B e-commerce during the initial adoption stage. However, most organizations also indicated that while senior managers had signed off on the B2B e-commerce projects, there was then little in the way of follow-up support. They perceived it simply as an IT-enabled cost-cutting mechanism rather than as a central strategic issue for their business.

Alignment of Organizational Goals with Business Objectives Alignment with stated organizational goals has a key bearing on how investment is organized and conducted, and the priorities that are assigned to different IT investment proposals (Mirani and Lederer, 1993). There appeared to be a lack of obvious linkage between stated organizational goals the expected outcomes of business objectives for B2B e-commerce. Business objectives for adopting and implementing B2B e-commerce systems by organizations varied greatly. The objectives mentioned by most healthcare organizations were related to the improved supply chain process, cost reduction, decreased administrative errors, and improved work efficiency. Other e-

commerce studies conducted in other countries (e.g. Enterprise Ireland, 2004 in Ireland; Levy et al., 2001 in UK; Locke and Cave, 2002 in New Zealand) suggested that many organizations simply failed to establish a linkage between the reasons for adopting an e-commerce system and their organizational goals. These systems were often installed without linking the benefits to their organizational goals. Other healthcare organizations did not see B2B e-commerce as a tool that would enable to increase their profits and revenue in the short term. Rather, it was adopted and implemented because some of their competitors were already using it, instead of using it to gain competitive advantages or achieve organizational and business objectives. Moreover, for many healthcare organizations interviewed, B2B e-commerce systems were still being regarded as IT focused rather than as a part of the business objectives and/or strategies. Furthermore, most healthcare organizations interviewed had set business goals such as achieving certain percentages of market share or reducing operating costs by certain percentages. But yet they generally did not focus on what strategic directions they were going and how B2B e-commerce can be strategically utilized to assist them to achieve their business goals. They need to be treated as a cost saving mechanism, as an effective way to compete in the healthcare

179

Key Adoption Challenges and Issues of B2B E-Commerce in the Healthcare Sector

industry, and as a strategic tool to improve their supply chain.

User Involvement /Participation The relevant literature has stressed that there is a direct relationship between user involvement and success of any IT systems (e.g. Davidson, 2002; Lin and Shao, 2000). However, the adoption and use of the B2B e-commerce systems by the healthcare organizations interviewed generally were forced upon the employees by the senior management. Data collected from the case study indicated that employees, particularly users of B2B e-commerce systems, were not consulted extensively beforehand and were not involved in the designing and adoption of these systems. Those organizations which kept their employees in the dark tended to have lower usage for their systems. Moreover, many benefits expected from the adoption of these systems were tailored mainly for the senior managers, not the lower ranked employees. Lack of user involvement often results in distrust between the affected employees, the senior management, and external B2B e-commerce systems vendors/suppliers. Employees felt that their requirements were not solicited before the adoption of B2B e-commerce projects. Healthcare organizations need to alignment of incentives with their organizational strategies. Establishment of appropriate incentives for their employees and stakeholders is critical in successful adoption of B2B e-commerce projects.

Financial and Managerial Resources Responses from the case study participants revealed that there was a general lack of financial and managerial resources in adopting and implementing B2B e-commerce. All of the organizations interviewed had Internet access and agreed that the further adoption of B2B e-commerce systems would be an important factor for the future success

180

of their organizations. The participating organizations indicated that they had used emails to communicate with their customers and suppliers and to increase internal efficiencies. However, more than half of these organizations failed to utilize their B2B e-commerce to conduct business effectively with their customers and suppliers. Many organizations indicated that they did not have sufficient financial and managerial resources to adopt and maintain their B2B e-commerce systems in order to conduct their business transactions online effectively.

Industry Government Regulations and Standardized B2B E-Commerce Protocol The healthcare industry often is subject to government regulations with regard to the adoption of the innovative B2B e-commerce technologies. As revealed by the case study interviewees, the healthcare industry had been subject to the pressures exerted by the government. The pharmaceutical companies, in particular, had to comply with both national and international regulations (e.g. bar coding, Sarbanes-Oxley Act). However, standards alone do not necessarily translate into ongoing improved performance. Many healthcare organizations stated that they need to work towards standardization of a universal product numbering system and for data exchange and information flows in order to facilitate B2B e-commerce. This is vital given that the transactions among healthcare organizations spans many parties and geographical dimensions. This is one important factor influencing the adoption of B2B e-commerce for these organizations in the healthcare industry. Given the substantial efforts and costs of the standardized policies, protocols, and procedures required for the healthcare industry, government role is crucial to facilitate successful B2B e-commerce initiatives. However, a significant

Key Adoption Challenges and Issues of B2B E-Commerce in the Healthcare Sector

number of healthcare organizations interviewed were quite hesitate to participate in government B2B e-commerce initiatives such as PECC. They were also partially influenced by some hardware and software vendors/suppliers fearing that these initiatives might have an impact on their profit margins and thus had no incentives to support these healthcare organizations to participate in these initiatives.

Role of Medicare Australia Medicare Australia is a statutory agency that provides public-funded universal health care system in Australia. It delivers services and payments for, among other things, Medicare, Aged care, and, Private Health Insurance Rebate, the Pharmaceutical Benefits Scheme. In recent years, it has encouraged its customers (e.g. hospitals, health providers, and patients) to engage in electronic transactions via its new Medicare Online, Aged Care Online Claiming, and Medicare Easyclaim. According to its 2007-2008 annual report (Medicare Australia, 2009), electronic claiming of bulk bill transactions, for example, has increased from 33% to more than 67% in June 2008. In addition, 69% of all residential aged care service providers had registered for online claiming in June, 2008 (Medicare Australia, 2009). At the moment, it is developing the Unique Healthcare Identifier service to the National E-Health Transition Authority which will be used to uniquely identify individuals as well as healthcare providers (Medicare Australia, 2009). This will have an impact on how healthcare organizations adopt and implement their B2B e-commerce systems.

Inter-Organizational: Supply Chain Effective Supply Chain Management Reliability and consistency of data integrity also affects the effectiveness of B2B e-commerce supply chain management and all other related business

functions which underpin business processes and transactions (Morgan, 2004). Inaccuracies and delays are likely to occur throughout the B2B ecommerce supply chain if there is a problem with data integrity (Morgan, 2004). Hence, effective B2B e-commerce supply chain management in the healthcare industry is critical as it affects the transactions between different stakeholders such as medical manufacturers, pharmacies, wholesalers, retailers, and hospitals. According to several participating health organizations, new technologies such as RFID enhanced supply chain integrity but they were still too costly to implement throughout the entire supply chain for most organizations. Several healthcare organizations interviewed had raised the concerns about whether the B2B e-commerce investments would bring about an equitable benefits for all stakeholders. Those organizations that had fewer problems in adopting and implementing B2B e-commerce systems were those who had better communication with key stakeholders throughout the entire supply chain and had listened to their cost concerns. The other issue mentioned by some participating healthcare organizations related to disaster recovery. They emphasized the importance of having a backup/ alternative B2B e-commerce system and IT disaster recovery and contingency plans in case of system failure. Several interview participants had also mentioned that having an effective supply chain management in their organizations were not good enough. Benefits would only come about if these management processes can be extended into other organizations.

Interoperability Problems Interoperability problems reduce organization’s IT systems to exchange information. It is one of the most cited reasons for B2B e-commerce failure (Clay and Strauss, 2002). Most participating healthcare organizations had some difficulties or simply failed to integrate their B2B e-commerce system with other functions throughout the supply

181

Key Adoption Challenges and Issues of B2B E-Commerce in the Healthcare Sector

chain. Most did not have an IT strategy to integrate their B2B e-commerce with other systems. For example, many healthcare organizations such as hospitals and pharmaceutical companies had purchased their own IT/B2B e-commerce systems, it was therefore not surprising to see that their IT/B2B e-commerce systems within the same organizations was unable to communicate, let alone between hospitals or different healthcare organizations across the entire supply chain. Moreover, many healthcare organizations implemented B2B e-commerce just to obtain gains promised by the vendors but did not believe that their B2B e-commerce systems could be seamlessly integrated with other functions within their organizations. Proper integration of B2B ecommerce and other functions of organizations clearly required a lot of managerial, financial, and technical resources as well as organizational capabilities. Only those healthcare organizations with higher level of IT maturity had more sophisticated B2B e-commerce systems and had been using it for a while had seen the integration of various functions as a main benefit. Lack of standardization polices cause problems for interoperability of systems and the consistency of data exchange and transfer protocols are essential for universal data access across the supply chain within the healthcare industry. The Key B2B E-Commerce Adoption Challenges and Issues Table was developed to assist senior managers to better manage their B2B ecommerce adoption as well as to ensure that the benefits expected are actually realized in order to improve their organization’s long term profitability (Table 1). This table has been developed from literature review and case study results. The table is consist of various major challenges that healthcare organizations need to overcome in order to realize B2B e-commerce benefits. The eight main are the important issues which healthcare organizations should pay attention in order to overcome the B2B e-commerce valuation and conversion barriers. Intra-organizational and resources key

182

challenges and issues are consist of top management support, alignment of organizational goals with business objectives, user involvement, and financial and management resources. Industry key success factor includes of government regulation and standardized B2B e-commerce protocol. Inter-organizational and supply chain key challenges and issues are comprised of supply chain management and interoperability issue.

CONCLUSION Case studies were conducted in healthcare organizations which had adopted and implemented B2B e-commerce systems. Several major B2B e-commerce adoption challenges and issues were identified for the healthcare organizations. The findings have the potential to assist healthcare organizations in enhancing the understanding of their B2B e-commerce adoption challenges and issues. One key contribution of the research is the development of the Key B2B E-Commerce Challenges and Issues Table for healthcare organizations. The table suggests that the eight main issues should be considered by healthcare organizations in the process of overcoming their valuation and conversion challenges. Paying close attention to these key challenges and issues will enable healthcare organizations to reap benefits from their B2B e-commerce investments. The Key B2B E-Commerce Adoption Challenges and Issues Table proposed for healthcare organizations in this paper should form part of an overall organizational strategy to guide healthcare organizations towards a balanced approach to manage their IT/B2B e-commerce investments adoption processes. It is also imperative for the senior executives of the healthcare organizations to carefully examine the reasons behind having these B2B e-commerce challenges and then focus their attention on the key challenges and issues identified in the table in order to obtain expected B2B e-commerce benefits. Other behavioral,

Key Adoption Challenges and Issues of B2B E-Commerce in the Healthcare Sector

attitudinal and organizational issues should not be neglected. From a practical standpoint, understanding the reactions of their employees toward the newly acquired B2B e-commerce systems and their subsequent behavior can help healthcare organizations to devise appropriate intervention strategies and programs to maximize their use and their effects on the organizations. This is important given that most healthcare firms, from our own observation, still pay little attention toward the effective utilization of any IT including B2B e-commerce. According to Davidson and Heineke (2007), stakes are too high to be content with gradual diffusion of the healthcare IT and therefore, deliberate steps must be identified to achieve desirable outcomes and increase the pace of dissemination. One way of doing it is to identify the relevant opinion leaders who can serve to reduce the uncertainty of others in adopting new IT resources. Finally, despite large investments in B2B ecommerce over many years, it has been difficult for healthcare organizations to determine where benefits have occurred, if indeed there have been any. Little published work has been conducted in adoption of B2B e-commerce in the healthcare industry and hence there is still a lot to be learned in the area. Through the case study results presented in this paper it is hoped that better B2B e-commerce adoption and implementation strategies may be developed by healthcare organizations.

FUTURE RESEARCH DIRECTIONS The very first electronic commerce was conducted over electronic data interchange (EDI) networks but the technology was expensive and extremely restrictive (Standing and Lin, 2007). Later, some major organizations (e.g. Wal-Mart) created their own Intranets and private networks to link to their purchasers, distributors, and suppliers. Currently, we see the rise two different electronic commerce entities - e-Marketplaces and B2B e-commerce

web portals. E-Marketplaces allow online purchasers and suppliers to find each other and avoid the inefficiencies of traditional markets. B2B ecommerce web portals, as mentioned earlier, are industry focused and often are initiated by the government and the industry itself. They allow industry players (e.g. hospitals and pharmaceutical companies) to access a wider range of purchasers, distributors, and suppliers and to sell their products and services to each other. Unlike many other industries, B2B e-commerce web portals within the healthcare industry generally are not mature enough. Due to the limited IT sophistication between healthcare organizations and their purchasers, distributors, suppliers, and business partners in the past, they are lagging behind the bigger players in other industries in using web based commerce applications to create competitive business networks. However, these applications are transforming industry by shifting business focus from physical stores to the virtual business. It is creating a more level playing field for all industry players. It is expected that future Internet applications will bring about faster and more efficient business to healthcare organizations by providing customers convenience and variety. To take advantages of these new highly efficient and cost effective channels, getting all healthcare organizations online is critical to forming the critical mass that will make the B2B e-commerce web portals successful. Forming the critical mass will also help make this new channel an industry standard. As discussed in the chapter, many challenges remain for the healthcare organizations to overcome. In particular, healthcare organizations need to pay attention to the supply chain challenges such as lack of standardized B2B protocols and interoperability issue. Industry/governmentsponsored B2B protocols and standards do exist but they need to be improved in order for the healthcare organizations to better interact within and between multiple B2B e-commerce web portals. In addition, healthcare organizations need not only to resolve the interoperability problems

183

Key Adoption Challenges and Issues of B2B E-Commerce in the Healthcare Sector

between their e-commerce systems and the B2B e-commerce web portals but also their ability to participate in multiple B2B e-commerce web portals. This issue will grow in importance in the near future as more and more B2B e-commerce web portals reach critical mass in not just the healthcare industry but also in other industries. Healthcare organizations are likely to have to work out how to access different B2B e-commerce web portals without incurring the costs of installing new systems as well as membership fees.

REFERENCES Bhakoo, V., & Sohal, A. (2008) An Assessment into the Drivers of E-Business Adoption within the Australian Pharmaceutical Supply Chain, 39th Annual Meeting of the Decision Sciences Institute at Baltimore, Maryland, USA, 22-25 November. Burn, J. M., & Ash, C. G. (2005). A Dynamic Model of e-Business Strategies for ERP Enabled Organisations. Industrial Management & Data Systems, 105(8), 1084–1095. doi:10.1108/02635570510624464 Bushell, S. (2000) PeCCing Order, CIO, 15 November. Chan, C., & Swatman, P. M. C. (2003). International Examples of Large-Scale Systems – Theory and Practice IV: B2B E-Commerce Implementation in the Australian Context. Communications of the Association for Information Systems, 11, 394–412. Chiasson, M., Reddy, M., Kaplan, B., & Davidson, E. (2007). Expanding Multi-disciplinary Approaches to Healthcare Information Technologies: What does Information Systems offer Medical Informatics. International Journal of Medical Informatics, 76s, s89–s97. doi:10.1016/j. ijmedinf.2006.05.010

184

Clay, K., & Strauss, R. P. (2002). Institutional Barriers to Electronic Commerce: An Historical Perspective. Advances in Strategic Management, 19, 247–273. doi:10.1016/S0742-3322(02)19008-5 Davidson, E., & Heslinga, D. (2007). Bridging the IT Adoption Gap for Small Physician Practices: An Action Research Study on Electronic Health Records. Information Systems Management, 24(1), 15–28. doi:10.1080/10580530601036786 Davidson, E. J. (2002). Technology Frames and Framing: A Social-cognitive Investigation of Requirements Determination. Management Information Systems Quarterly, 26(4), 329–358. doi:10.2307/4132312 Davidson, S. M., & Heineke, J. (2007). Toward an Effective Strategy for the Diffusion and Use of Clinical Information Systems. Journal of the American Medical Informatics Association, 14(3), 361–367. doi:10.1197/jamia.M2254 Enterprise Ireland. (2004). IT/eBusiness Status and Issues of Small and Medium Sized Irish SMEs, BSM Ltd Report, available on-line at: http://www.enterprise-ireland.com/ebusiness/ eBIT_ICTissues.htm. GS1. (2004) Hospital Pharmaceutical eCommerce Looks Healthy, Media Press Release, GS1 Australia, Australia. CR Groups (2002) Community Pharmacy: Pathways in an E-commerce World, CR Group Pty Ltd, December. Heeks, R. (2006). Health Information Systems: Failure, Success and Improvisation. International Journal of Medical Informatics, 75, 125–137. doi:10.1016/j.ijmedinf.2005.07.024 Herzlinger, R. E. (2006). Why Innovation in Health Care is so Hard. Harvard Business Review, 84(5), 58–66.

Key Adoption Challenges and Issues of B2B E-Commerce in the Healthcare Sector

Howcroft, D., & McDonald, R. (2004) An Ethnographic Study of IS Investment Appraisal, The 12th European Conference on Information Systems (ECIS2004), Turku, Finland, June 14-16. Klein, H. K., & Myers, M. D. (1999). A Set of Principles for Conducting and Evaluating Interpretive Field Studies in Information Systems. Management Information Systems Quarterly, 23(1), 67–94. doi:10.2307/249410 Laudon, K., & Laudon, J. (2004). Management Information Systems: Managing the Digital Firm. New Jersey: Pearson Education, Inc. Lee, C., & Shim, J. P. (2007). An Exploratory Study of Radio Frequency Identification (RFID) Adoption in the Healthcare industry. European Journal of Information Systems, 16(6), 712–724. doi:10.1057/palgrave.ejis.3000716 Levy, M., Powell, P., & Yetton, P. (2001). SMEs: Aligning IS and the Strategic Context. Journal of Information Technology, 16, 133–144. doi:10.1080/02683960110063672 Lin, C., Huang, Y., & Burn, J. (2007). Realising B2B e-Commerce Benefits: The Link with IT Maturity, Evaluation Practices, and B2B e-Commerce Adoption Readiness. European Journal of Information Systems, 16(6), 806–819. doi:10.1057/ palgrave.ejis.3000724 Lin, C., Huang, Y., & Jalleh, G. (2008a) Improving Alliance Satisfaction: The Resource Alignment of IT Competency in Small Healthcare Centers. International Technology Management Review, November, 1(2), 25-42. Lin, C., & Pervan, G. (2003). The Practice of IS/IT Benefits Management in Large Australian Organizations. Information & Management, 41(1), 13–24. doi:10.1016/S0378-7206(03)00002-8

Lin, C., Pervan, G., Lin, H.-C., & Tsao, H. (2008b). An Investigation into Business-to-Business Electronic Commerce Organizations. Journal of Research and Practices in Information Technology, 40(1), 3–18. Lin, W. T., & Shao, B. B. M. (2000). The Relationship Between User Participation and System Success: A Simultaneous Contingency Approach. Information & Management, 37, 283–295. doi:10.1016/S0378-7206(99)00055-5 Locke, S., & Cave, J. (2002). Information Communication Technology in New Zealand SMEs. Journal of American Academy of Business, 2(1), 235–240. Lynch, P. D., & Beck, J. C. (2001). Profiles of Internet Buyers in 20 Countries: Evidence for Region-Specific Strategies. Journal of International Business Studies, 32(4), 725–748. doi:10.1057/ palgrave.jibs.8490992 McGaughey, R. E. (2002). Benchmarking Business-to-business Electronic Commerce, Benchmarking. International Journal (Toronto, Ont.), 9(5), 471–484. Medicare Australia. (2009) Medicare Australia Annual Report 2007-2008, Canberra, Australia. Melville, N., Kraemer, K., & Gurbaxani, V. (2004). Review: Information Technology and Organizational Performance: An integrative Model of IT Business Value. Management Information Systems Quarterly, 28(2), 283–322. Miles, M. B., & Huberman, A. M. (1994). Qualitative Data Analysis: An Expanded Sourcebook. California: Sage Publications. Miller, J. (2003). Measuring and Aligning Information Systems with the Organization. Information & Management, 25(4), 217–228. doi:10.1016/03787206(93)90070-A

185

Key Adoption Challenges and Issues of B2B E-Commerce in the Healthcare Sector

Mirani, R., & Lederer, A. L. (1993). Making Promises: The Key Benefits of Proposed IS Systems. Journal of Systems Management, 44(10), 10–15. More, E., & McGrath, M. (2002). An Australian Case in e-Health Communication and Change. Journal of Management Development, 21(7/8), 612–632. Morgan, L. (2004) Standards Underpin Ecommerce in Healthcare, Standards Australia Limited, October. NEHTA. (2004) The Future is e-Health, National E-Health Transition Authority, [Online] http:// www.nehta.gov.au/, Australia. NSCRTF. (2008) Australian Health Care Reform, National Supply Chain Reform Task Force, [Online] http://www.healthsupplychain.gov.au/. Parente, S. T. (2000). Beyond the hype: A Taxonomy of e-Health Business Models. Health Affairs, 19(6), 89–102. doi:10.1377/hlthaff.19.6.89 Porter, M. E., & Teisberg, O. (2004). Redefining Competition in Health Care. Harvard Business Review, 82(6), 64–76. Power, D. (2004). The Comparative Importance of Human Resource Management Practices in the Context of Business to Business (B2B) Electronic Commerce. Information Technology & People, 17(4), 380–406. doi:10.1108/09593840410570302 Standing, C., & Lin, C. (2007). Organizational Evaluation of the Benefits, Constraints and Satisfaction with Business-To-Business Electronic Commerce. International Journal of Electronic Commerce, 11(3), 107–153. doi:10.2753/ JEC1086-4415110304 Standing, S., Standing, C., & Lin, C. (2008) A Framework for Managing Knowledge in Strategic Alliances in the Biotechnology Sector, Systems Research and Behavioral Science, November, 25(6), pp783-796.

186

Suomi, R., Tahkapaa, J., & Holm, J. (2001) Organizational and Information System Metaphors in the Health Care Sector – From Harmonized Value Chain to Realistic Market Models, The 9th European Conference on Information Systems, Bled, Slovenia, June 27-29. Teo, T. S. H., & Ranganathan, C. (2004). Adopters and Non-adopters of Business-to-business Electronic Commerce in Singapore. Information & Management, 42, 89–102. Wickramasinghe, N. S., Fadlalla, Geisler, E., & Schaffer, J. L. (2005). A Framework for Assessing E-Health Preparedness. International Journal of Electronic Healthcare, 1(3), 316–334. doi:10.1504/IJEH.2005.006478

KEY TERMS AND DEFINITIONS B2B E-Commerce: Business-to-business electronic commerce. Business conducted through the Internet between companies. Benefits: The tangible and intangible returns or payback expected to be obtained from a systems investment or implementation. E-Commerce: A business model that is conducted over the Internet in which clients are able to participate in all phases of a purchase decision. Electronic commerce can be between two businesses transmitting funds, goods, services or between a business and a customer. Information Technology (IT): Any computerbased tool that users use to work with information and support the information needs of an organization. Healthcare Industry: The health profession industry which offers services in relation to the preservation of health by preventing or treating illness. Supply Chain Management (SCM): It involves coordinating and integrating the network of retailers, wholesalers, distributors, transporters,

Key Adoption Challenges and Issues of B2B E-Commerce in the Healthcare Sector

storage facilities, and suppliers that involve in the sale, delivery and production of a particular product or service, both within and among different organizations. It covers all movement and storage of raw materials, work-in-process inventory, and finished goods within and across different organizations.

User Involvement: User participation. It is an act or a process for users to actively participate and/or or share their expertise, thoughts, and experience during a system development life cycle or project.

187

188

Chapter 12

A Pervasive Polling SecretSharing Based Access Control Protocol for Sensitive Information Juan Álvaro Muñoz Naranjo Universidad de Almería, Spain Justo Peralta López Universidad de Almería, Spain Juan Antonio López Ramos Universidad de Almería, Spain

ABSTRACT This chapter presents a novel access control mechanism for sensitive information which requires permission from different entities or persons to be accessed. The mechanism consists of a file structure and a protocol which extend the features of the OpenPGP Message Format standard by using secret sharing techniques. Several authors are allowed to work in the same file, while access is blocked for not authorized users. Access control rules can be set indicating the minimum number of authors that need to be gathered together in order to open the file. Furthermore, these rules can be different for each section of the document, allowing collaborative work. Non-repudiation and authentication are achieved by means of a shared signature. The scheme’s features are best appreciated when using it in a mobile scenario. Deployment in such an environment is easy and straight. DOI: 10.4018/978-1-60960-042-6.ch012 Copyright © 2011, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited.

A Pervasive Polling Secret-Sharing Based Access Control Protocol for Sensitive Information

INTRODUCTION Protection of sensitive information is an everpresent concern which is gaining more and more attention as digitalization of information and use of Internet increase. Numerous privacy protocols, standards and applications exist in order to keep information away from unauthorized persons, as well as to authenticate its author. Most of them regard privacy while information is being transmitted through the web and provide authentication (OpenSSH) (Dierks&Rescorla, 2008) (Atkinson, 1995) (Kohl, 1989) (Paterson &Yau, 2006). GnuPG (GnuPG) and PGP (PGP) applications also protect information while stored in a device. Both implement the OpenPGP Message Format standard (Callas, Donnerhacke, Finney, Shaw, & Thayer, 2007). Those protocols and applications usually work on an individual basis, that is, a single user manages the privacy of its own information. Some examples are: communicating in a private and/or authenticated way with a Web server, protecting personal information and keeping a private and/ or authenticated email conversation. Some scenarios may require, in addition to protection and authentication of information, some kind of access control measures. That is the case of governmental classified documents (defense, foreign affairs, historical, etc.) or high-value information in private companies. Controlling the access to this kind of documents is critical, and may require the approval of third-party entities or individuals, or even a set of them. Security restrictions will be even higher when modifying the classified information. On a different matter, advances in smart devices and connectivity have given us the chance to access Internet from almost anywhere and at any time. Internet access is no longer confined to static devices that cannot be taken with us. Now that the technology already exists, it is time to develop new mechanisms and applications that take advantage of it.

Having all this in mind we have designed a polling-based file access control mechanism that is presented in this chapter. This mechanism includes an extension of the OpenPGP Message Format and a protocol: access to the file is granted only under the approval of a minimum number of authorized users, and modifications are signed for authenticity and integrity verification. The first feature is achieved by using secret sharing techniques; the later by using a shared signature. Section Background explains and discusses some technologies that keep some similarity, along with the OpenPGP Message Format, the secret sharing techniques and the shared signature. Section Our Proposal introduces our scheme with some mobility and security considerations, and finally the last section shows the conclusions of the chapter.

BACKGROUND Alike Technologies: Publish/ Subscribe Systems and Version Control Systems There are two technologies that keep some relation to the one proposed here. They are (1) publish/ subscribe systems and (2) version control systems. State-of-the-art publish/subscribe systems consist of an infrastructure that provides communication capabilities for large-scale, wide-area distributed systems. They nicely fit multi-domain, heterogeneous environments. The communication pattern is based on a per-event, asynchronous basis, either one-to-many or many-to-many, depending on the scenario. Such a system is presented in (Pietzuch & Bacon, 2002), along with a wider bibliographic review. A typical case-of-use appears in (Pesonen, Eyers, & Bacon, 2007): a multi-domain network that handles car plate recognition for fee-charging purposes in the London metropolitan area. Additionally, several works and proposals have been presented in order to cover

189

A Pervasive Polling Secret-Sharing Based Access Control Protocol for Sensitive Information

the security aspects, such as (Wang, Carzaniga, Evans, & Wolf, 2002) (Fiege, Zeidler, Buchmann, Kilian-Kehr, Mühl, & Darmstadt, 2004)(Pesonen, Eyers, & Bacon, 2006). Version control systems are (usually) distributed applications that facilitate the management of projects and documents, especially when there are several authors involved. Typical features of these systems include: • • •

•

Different users involved. Document versions management and publication (milestones, branches, etc.). Document history tracking: it is possible to follow a document’s history, and to recover a previous state or version. Event logging.

Well known version control systems are CVS (CVS), Subversion (Subversion), Microsoft SourceSafe (SourceSafe), JEDI Version Control System (JEDI), etc. Each technology has its own scenario of application. The goals of our mechanism, though related, differ slightly from those mentioned above. A typical case of use for ours is management of high-value information that is stored in one or several trusted servers. A set of authorized users may access the protected data, and modify them only after the approval of a sufficiently large number of (different) users. The nature of the high-value information might be commercial (the firm’s major stakeholders being the authorized users), personal (laws protect several kinds of citizens’ personal information in many countries) or even military.

The OpenPGP Message Format The OpenPGP Message Format was designed to offer a combination of privacy and authentication when storing (on physical devices) and sending (via Internet) information in files. Compression can be also applied. The GnuPG and PGP appli-

190

cations provide more functionality, such as key and certificate management, but the OpenPGP Message Format can be seen as the core of both applications. This section covers the basic structure of the format. A wider description can be found in (Callas, Donnerhacke, Finney, Shaw, & Thayer, 2007) and (Lucas, 2006). First of all, let us have a quick look at how GnuPG works. The application allows a user to encrypt a piece of information (such as a message or a file) and send it to one or more recipients. Only the specified recipients (including the creator himself) will be able to decrypt the information. Additionally, the creator can prove its identity by adding a digital signature, no matter whether the information is encrypted or not. The message can also be compressed is desired. More technically, each GnuPG user owns one or more public/private key pair. As the reader probably knows, an information piece encrypted with a public key can only be decrypted with its private counterpart, and vice versa. This elemental property of Public Key Cryptography is smartly exploited in the OpenPGP Message Format. One of the features GnuPG offers is the use of a contacts agenda: it allows storing and managing the information related to our contacts, such as name, email and, most important, public key. They will be useful when establishing the recipients of a message. In the same way, our contacts will probably have our data too. Now we can focus on the Message Format itself. The basic idea is to encrypt the information with a symmetric key, namely session key, and then attach it to the message in a way it is protected from outsiders. The problem is solved by encrypting the session key with the public key of the recipient. The result is attached to the message. There can be more than one recipient: the session key is encrypted separately with every recipient’s public key, and all the results are packed together. There are, hence, three main sections in a

A Pervasive Polling Secret-Sharing Based Access Control Protocol for Sensitive Information

complete message: the encrypted information, the signature and the encrypted session keys section. Let D be the information to be signed and encrypted. D is signed in first place, and the signature attached. If decided to, D is compressed in this moment. Afterwards, a new session key is generated, and D is encrypted (the session key is created according to the symmetric algorithm to be used). Now, let us assume there are n recipients for the message. The session key is encrypted n times, each one with the public key of the respective recipient. All the encrypted session keys are attached at the beginning of the message. The decryption process for each recipient consists of finding a header that can be decrypted, obtaining the session key, then decrypting the information and verifying the signature with the sender’s public key. Note that the information can also be encrypted for personal use only, to keep information private when stored in a physical device. In that case the recipient is the same as the creator, and only he/ she will be able to decrypt the file, if desired. An inconvenient of the headers approach is the need for encrypting the same piece of info several times, hence having redundant information. The good news is that it is a direct and efficient approach that keeps the file structure simple.

Secret Sharing An intuitive approach to Secret Sharing might be the following: “Secret Sharing techniques allow dividing a secret into several ‘pieces’, in a way that the secret can be recovered by gathering some of them (all of them in some cases).” Actually, this affirmation is just a way of seeing Secret Sharing. Usually, these techniques do not “slice” the secret: they generate pieces of information that depend on it. These pieces, called

shares, can be processed together to obtain the original piece of info. The most popular family of techniques within the Secret Sharing pool is that of threshold schemes. A (t,n) threshold scheme can be described as: “computing n shares that depend on a secret S, so that t≤n shares allow the generation of S, being this impossible with a number of shares below the threshold t”. A perfect (t,n) threshold scheme is one in which knowing only t-1 or fewer shares provides no advantage to an attacker over knowing no pieces (Menezes, van Oorscht, & Vanstone, 1996). The initial motivation for secret sharing was backing up cryptographic keys in such a way that no participant had a full knowledge of them. More recent uses have emerged: shared control for activities that require the collaboration or supervision of many entities or persons (opening bank vaults, military orders) and even logical and physical access control (the protocol presented in this chapter is a logical access control).

Shamir’s Threshold Scheme Probably the most famous threshold scheme is the one from Shamir (Shamir, 1979). It shows some desirable features: 1. It is perfect (as defined previously). 2. The size of each share does not exceed that of the secret. 3. New users are accepted: it is possible to compute new shares without affecting those existing. 4. Its security is demonstrated mathematically and it is not based on the unfeasibility of brute force attacks. Let us see now a mathematical description of this elegant method, taken from (Pastor Franco

191

A Pervasive Polling Secret-Sharing Based Access Control Protocol for Sensitive Information

& Sarasa López, 1998). Assume S is the secret and that at least t shares will be needed for its reconstruction. First, a prime p such that 0 ≤ p > S and a polynomial P (x ) with degree t-1 are chose. We can note P (x ) as t −1

P (x ) = S + ∑ Pi x i (mod p) i =1

Where 0 ≤ Pi > p are random integers such

that 0 ≤ Pi > p for i = 1,...,(t − 1) . A share

consists of a couple [x i , P (x i )] for i = 1,..., n . Recalculating S with t couples (shares) is possible thanks to the fact that interpolation of a (t-1)-th degree polynomial requires t points at least. The interpolation is calculated by solving the following system of equations: P (x 1 )= S + P1x 1 + ... + Pt −1x 1t −1 (mod p)  P (x t ) = S + P1x t + ... + Pt −1x tt −1 (mod p)

where(S , P1,..., Pt −1 ) are the unknown values. A system of this kind has always a solution, which can be efficiently found thanks to the Fast Fourier Transform (Blahut, 1984)(McEliece & Sarwate, 1981). A system created from less than t points (hence with less equations than unknown values) would be underdetermined and would have an infinite number of solutions. To make things even better, the scheme can be easily modified to establish hierarchies within the set of shares, in a way that some shares have more importance than others. An example would be a special share such that the secret could not be computed without it. Even several hierarchy levels can be set: forcing the presence of a top level share, two second level shares and three regular shares would be an example.

192

Shamir’s scheme has been extensive studied and some attacks have been conceived, always under the assumption of a many share-owners alliance with the intention of illegally recovering the secret. There are also other secret sharing schemes (Brickell, Some ideal secret sharing schemes, 1989)(Brickell & Davenport, 1991)(Chen, Gollmann, Mitchell, & Wild, 1997)(Numao, 1999) (Srinathan, Tharani Rajan, & Pandu Rangan, 2002) (Wang & Wong, 2008). Secret sharing is used in our protocol to force a rendezvous of at least a given number of users (each user is provided with a single share).

Shared Signature The utility of digital shared signature techniques is to bind two signers together when signing a document or piece of information, proving that the action was carried together by both. It was first introduced in (Radu, Govaerts, & Vandewalle, 1997), though other schemes exist (Luo, Si, Liu, & Li, 2008). They have been extensively used in electronic cash. Shared signature techniques work by computing a pair of keys (public and private) from signers’ own pairs. A hash of the info is encrypted with the new private key. The verifier can use the new public key to verify the signature. It must be unfeasible to obtain the signers’ private keys from the new pair. Signers’ private keys must remain secret: each signer must have no information about the other’s private key. Our scheme uses a shared signature to leave evidence of the last edition on the document, thus providing non-repudiation. The document is signed with a shared private key derived from the author who made the last changes and an Authorization Center (which will be introduced later). The resultant shared public key can be used later to verify the signature.

A Pervasive Polling Secret-Sharing Based Access Control Protocol for Sensitive Information

OUR PROPOSAL

•

Proposed File Structure The proposed scheme is based in a file structure which is described in this section. This structure is managed by an Authorization Center, which will be introduced in the following subsection. The file contains the encrypted information and other sections at the beginning and the end of it. The symmetric key used to encrypt the information is called session key, to keep similarity with the OpenPGP Message Format Standard. Session keys should be generated using a cryptographically secure pseudo-random number generator, and they should be different for each file (ANSI X9.17 standard) (Zhuo, Zhengwen, & Nan, 2008). Shares are generated for the session key at the moment of file creation. The number of shares should be equal to the number of authors considered. Each author is assigned a share, and a header is built for him/her. The Authorization Center also assigns itself a header in which it stores the whole key. The reason for this is explained in the Security Considerations Section. The following information is included in each header: •

Author’s Id/name.

• •

Encryption parameters: symmetric algorithm used, block size (if applicable), key length (if applicable), encryption mode (ECB, CBC ...), padding mode (PKCS7, ISO 10126 ...) and any other information required for decryption (Menezes, van Oorscht, & Vanstone, 1996). Minimum number of authors required for reconstruction of the session key. The share itself.

The header is then encrypted by using a public key algorithm and the corresponding author’s public key. This ensures privacy of the share. Note that the “encryption parameters” and the “minimum number of authors” values are similar for all headers. All headers are grouped together in a headers section and included at the beginning of the file, before the encrypted information. Figure 1 shows the process. An Encrypted metadata section is placed after the headers. It contains: • • • •

Content type/format of encrypted info. Version of info. Any other information required. Whether the info is compressed or not (and the compression algorithm if necessary).

Figure 1. Generation of headers and headers section

193

A Pervasive Polling Secret-Sharing Based Access Control Protocol for Sensitive Information

The metadata section is encrypted with the session key. It is followed by the encrypted info section, which contains the original information encrypted using the encryption parameters detailed in every header and the session key. Finally, a shared signature is placed at the end of the file: this signature consists of a hash of all sections (headers, metadata and encrypted info) encrypted using a shared signature technique, as mentioned above in the Shared Signature Subsection. The signature is replaced after every modification of the file. The complete file structure is depicted in Figure 2.

File Management: Proposed Access Protocol Management and hosting of the file is carried out by an Authorization Center, a third part trusted by all authors. We call this entity The Host. The Host, therefore, must keep the file secure and manage the process of access and/or edition to the file when required. In case an Authorization Center does not exist then an author can take the role, if trusted by all others. The Host determines the authors in the system. Additionally, it keeps and manages a list with useful information from the authors, such as their email address and their public keys.

Figure 2. File structure

Files are created outside The Host. A new file must be uploaded to the Host by an author or by other kind of authorized user, such as a Host administrator. The policy that determines which authors are assigned to the file is not discussed in this chapter, since different scenarios may require different strategies. For the sake of simplicity let us assume that The Host decides which authors might have access to the file (and, therefore, have a header attached to it). After the file is uploaded and the header section is created, a shared signature is created. This signature binds the up loader and The Host. A protocol for read and edition access is presented next. Communications are made under a secure channel (using SSL, TLS or similar). Each file in The Host is assigned a unique identifier (FileID). To assure version consistence the edition access imposes a lock for other actions of the same type. This means that only one edition access at a time is permitted. Nevertheless, many read accesses can be carried out concurrently, even if an edition access is being performed. In that case, the reader is aware of the fact that the file is being edited at that moment, since all edition actions are notified. Each transaction is assigned a unique identifier (TransID). Only one transaction of any kind is allowed per author at a time. Let us assume there are n authors and a minimum of t are required to decrypt the file. Every author and The Host are assigned a header in the file, as mentioned in the Proposed File Structure Subsection. Assume also that author A is the one who performs the read or edit action. For read access: 1. A sends The Host a ReadFileID message. 2. The Host assigns the transaction a new TransID and sends every other author the message ReadFileIDTransIDA, which indicates that author A wants to access the file for reading purposes only. This message

194

A Pervasive Polling Secret-Sharing Based Access Control Protocol for Sensitive Information

includes the corresponding header for every contacted author. 3. Every author excepting A replies with a message Allow TransID (giving permission) or Deny TransID (denying permission), depending on the author decision. Messages Allow TransID also include the header content, decrypted with the author’s private key. 4. i. If The Host receives at least t affirmative responses then it reconstructs the session key, decrypts the file and sends it to A along with the message Read Access FileID Granted. The Host also informs every author with the message Read Access TransID Granted. ii. If The Host does not receive at least t affirmative responses then it informs A with the message Read Access FileID Denied, and every other author with Read Access TransID Denied. 5. The Host encrypts the file again with a new random session key and generates new headers. For edition access: 1. A sends The Host an Edit FileID message, along with a text description of the changes he/she wants to make to the document. 2. The Host sends every other author the message Edit FileIDTransID A, along with the description of the changes and the corresponding header for the contacted author. 3. Every author excepting A replies with a message Allow TransID (giving permission) or Deny TransID (denying permission), depending on his/her decision. Messages Allow

TransID also include the header content, decrypted with the author’s private key. 4. a. If The Host receives at least t affirmative responses then it reconstructs the session key, decrypts the file and sends it to A along with the message Edit Access FileID Granted. The Host also informs every author with the message Edit Access TransID Granted. Go to step 5. b. If The Host doesn’t receive at least t affirmative responses then it informs A with the message Edit Access FileID Denied, and every other author with the message Edit Access TransID Denied. The process is aborted. Go to step 9. 5. A receives the decrypted file, edits it and sends back the changes (the result of a “diff” operation) to The Host along with the message Changes FileID. 6. The Host forwards the changes to all other authors along with the message Changes TransID (including those who replied with a denial). 7. Every author except for A replies with a message Allow Changes TransID (giving its approval to the changes) or Deny Changes TransID (disapproving the changes), depending on his/her decision. 8. a. If The Host receives at least t affirmative responses then it commits the changes to the file and informs A with Changes FileID Committed, and every other author with Changes TransID Committed. b. If The Host does not receive at least t affirmative messages then it informs A with Changes FileID Rolled Back, and every other

195

A Pervasive Polling Secret-Sharing Based Access Control Protocol for Sensitive Information

author with Changes TransID Rolled Back. 9. The Host encrypts the file again with a new random session key and generates new headers. A new shared signature binds A and The Host. The old file is deleted. Sending a description of the intended changes at the beginning of the edition process allows for an early stop if changes are not going to be approved by the minimum number of authors required (steps 1-4). Two timeout values may be set (users should be aware): •

•

An author response timeout: if an insufficient number of authors do not reply on time then the transaction is rolled back. An edition timeout: if A does not send back the changes made to the file in time then the process is aborted by The Host. This prevents authors from holding the document for too long. Good values for the timeout can vary from several hours to many days, depending on the scenario.

Figure 3 and Figure 4 show a read and edition attempt by author A, respectively. For the sake of simplicity only one author is required to give his/ her approval. Two possible results are depicted in each figure.

Dividing the Information into Parts This scheme can be extended in the case of documents that can be divided into parts, such as text files. We will call those parts document sections. Depending on the scenario it can be convenient that a given author cannot access all document sections. This can be easily achieved by using different session keys for different document sections. Each header will then contain a share for the session key of every document section he/she has permission. The Host’s header will contain a complete copy of every session key used. Note

196

that the metadata section must be accessed each time a read/edit action is performed. This section must be therefore encrypted with a different key, which will be included in The Host’s header too. The reconstruction process of any session key is carried out in the same way as explained before. For proper authentication purposes, a different shared signature should exist for every document section. The corresponding signature should be updated after modification of a section.

Mobility Considerations Application of the scheme to a mobile scenario is straightforward, as long as client devices support public key cryptography and document edition. The scheme does not change: documents are stored at a central Host, with mobile clients accessing them as usual. Obviously clients must be connected to Internet at the time of access. Note that when asking for approval after a change on a document has been proposed, clients only receive the differences with the original instead of the whole document. This feature nicely fits mobile scenarios, since wireless connections are usually slower. What’s more, the use of mobile technologies along with the scheme will clearly improve users experience, since the approval process can be managed on real time. Offline periods are a common problem when facing the mobile scenario: clients are suddenly cut off and cannot be reached. Let us consider three interesting situations. 1. An online client is granted edit access to a document while online, but it is offline at the time of sending the changes back (step 5 in the edition access procedure): a smart implementation should allow the user to edit the document as desired and keep the result until connectivity is recovered. The changes should be sent then to The Host in a transparent way. That is similar to how email clients work.

A Pervasive Polling Secret-Sharing Based Access Control Protocol for Sensitive Information

Figure 3. Read access example: (a) Access granted (b) Access denied

2. A client is offline when a change notification is sent by The Host (step 6 in the edition access procedure): little can be done in this case. The client should get the notification when going online again. 3. An online client receives a notification asking for approval (either for access or changes commit), but it is offline at the time of emitting its verdict: as in case 1, the application

should remember the verdict until it can be sent to the Host. As can be learnt from these situations, it is important to manage connectivity issues in a transparent way, so user’s activity is interfered the least possible.

197

A Pervasive Polling Secret-Sharing Based Access Control Protocol for Sensitive Information

Figure 4. Edit access example: (a) Access granted (b) Access denied

198

A Pervasive Polling Secret-Sharing Based Access Control Protocol for Sensitive Information

Security Considerations Note that after each access attempt The Host file reencrypts the file and generates new headers. This is possible thanks to The Host’s header present in the headers section. The reason for re-encryption is preventing the exchange of shares among authors. To explain this, suppose that, for a threshold t, t+1 authors join forces to gain illegal access to the file (without notifying The Host). One of the authors might request access, while the other t might deny it. After the process, these authors already own their share and can communicate them in order to reconstruct the session key. Establishing a new session key and new shares avoids the problem. To this end, The Host is required to generate a header with a complete copy of the current session key for further re-encryption. Authenticity of the file is assured by the shared signature, as long as one of the signers is The Host (if The Host’s private key wasn’t compromised, of course). If an author modifies its local copy without permission then the hash in the signature will not match that of the file content. Privacy in communications is achieved by using SSL or TLS, which also guarantee authenticity of the messages. Any incoming message from a finished transaction should be immediately discarded by the receiver in order to prevent malicious actions. To run such protocols the client device should support public key cryptography, which should not be an issue given the state of the art in commercial devices.

Possible Applications Implantation of the scheme may be desirable on scenarios in which several organizations or individual must collaborate altogether towards a common goal, but there does not exist a mutual sense of trust, whatever the reason and, furthermore, the information handled should not be made public. We now show two examples, both settled in a commercial scenario.

In the first example we have two enterprises that collaborate on one or many shared projects, like the release of a new product, or even a fusion. There are many commercial documents, such as strategic plans, action lines, etc., that should only be known by some key members of the two firms (both boards of directors, advisers, delegates, etc.). In addition, any modification of those documents should only be done after agreement or ballot. Mutual mistrust, especially among firms that compete in the same strategic market, might lead to the adoption of a mechanism like ours. The second example uses the hierarchical version of Shamir’s Threshold Scheme. It was mentioned in Section Background that one or many shares can be given more value than others: the secret could only be reconstructed then if a given number of those important shares were used. Let us consider a single firm. The president, or the most important stakeholders, might be given each one important share, thus controlling the private document(s) evolution. The leader or supervisor of a workgroup might also take advantage of our scheme.

CONCLUSION This work introduces a novel access control mechanism for documents that need high privacy, such as classified governmental files, medical files (Lederman, 2004) (Peterson, 2005) (Tang, 2008) or financial information. The proposed mechanism grants access to the protected information only when a minimum number of authorized users agree on the action. This is achieved by using secret sharing techniques. A header for every user is added at the beginning of the document. Each header contains a share for the symmetric key (session key) used to encrypt the private information. In order to keep shares private each header is encrypted using an asymmetric algorithm along with the public key of the respective user, so only he/she can obtain the share.

199

A Pervasive Polling Secret-Sharing Based Access Control Protocol for Sensitive Information

Decryption of the document requires a virtual rendezvous of a given minimum number of users, avoiding the necessity of a physical meeting to authorize the access to the private information. A protocol for this is presented. It uses SSL, TSL or an equivalent protocol to protect communications. A shared digital signature is attached at the end of the document to keep record of the users responsible for the last edition. The mechanism can be extended so access permissions to individual sections of the document can be given by encrypting every section with a different session key. The scheme nicely fits mobile scenarios, since all participating users can receive every notification on real time, making the process more dynamic.

ACKNOWLEDGMENT We would like to thank the reviewers for their valuable ideas and suggestions. This work has been partially funded by grants from the Spanish Ministry of Science and Innovation (TIN2008-01117), Junta de Andalucía (P06-TIC-1426, P08-TIC-3518, FQM 0211 and TEC2006-12211-CO2-02), and in part financed by the European Regional Development Fund (ERDF).

REFERENCES Anton, A. I., Earp, J. B., He, Q., Stufflebeam, W., Bolchini, D., & Jensen, C. (2004). Financial privacy policies and the need for standardization. Security & Privacy, IEEE, 2(2), 36–45. doi:10.1109/MSECP.2004.1281243 Atkinson, R. (1995). IP Encapsulating Security Payload. ESP. Blahut, R. E. (1984). A Universal Reed-Solomon Decoder. IBM Journal of Research and Development, 28(2). doi:10.1147/rd.282.0150

200

Brickell, E. F. (1989). Some ideal secret sharing schemes. J. Combin. Math. Combin. Comput., 9, 105–113. Brickell, E. F., & Davenport, D. M. (1991). On the Classification of Ideal Secret Sharing Schemes. CRYPTO’89 Proceedings. Callas, J., Donnerhacke, L., Finney, H., Shaw, D., & Thayer, R. (2007). OpenPGP Message Format. RFC 4880. Chen, L., Gollmann, D., Mitchell, C. J., & Wild, P. (1997). Secret sharing with reusable polynomials. In Information Security and Privacy. New York: Springer. doi:10.1007/BFb0027925 Dierks, T., & Rescorla, E. (2008). The Transport Layer Security (TLS) Protocol, Version 1.2. IETF. Fiege, L., Zeidler, A., Buchmann, A., Kilian-Kehr, R., Mühl, G., & Darmstadt, T. (2004). Security Aspects in Publish/Subscribe Systems. Third International Workshop on Distributed Eventbased Systems (DEBS’04). JEDI. https://jedivcs. sourceforge.net/ GnuP.G. http://www.gnupg.org/ Kohl, J. T. (1989). The use of Encryption in Kerberos for Network Authentication. In Advances in Cryptology - CRYPTO’ 89 Proceedings. Springer Berlin / Heidelberg. Lederman, R. (2004). The Medical Privacy Rule: Can Hospitals Comply Using Current Health Information Systems? In Proceedings of the 17th IEEE Symposium on Computer-Based Medical Systems. Lucas, M. W. (2006). PGP and GPG. US: No Starch Press. Luo, Y.-A., Si, Y.-L., Liu, W.-Y., & Li, F. (2008). Shared Signature based on Elliptic Curve and its Application In Electronic Cash.

A Pervasive Polling Secret-Sharing Based Access Control Protocol for Sensitive Information

McEliece, R. J., & Sarwate, D. V. (1981). On Sharing Secrets and Reed-Solomon Codes. Communications of the ACM, 24(9). doi:10.1145/358746.358762 Menezes, A., van Oorscht, P., & Vanstone, S. (1996). Handbook of applied cryptography. CRC Press. Numao, M. (1999). Periodical Multi-secret Threshold Cryptosystems. ASIACRYPT’99. OpenSSH. (n.d.). Retrieved from http://www. openssh.org/ Pastor Franco. (1998). José & Sarasa López, Miguel Ángel. Criptografía Digital: Fundamentos y Aplicaciones. Publicaciones Universitarias Universidad de Zaragoza. Paterson, K. G., & Yau, A. K. (2006). Cryptography in Theory and Practice: The Case of Encryption in IPsec. Springer-Verlag. Pesonen, L. I., Eyers, D. M., & Bacon, J. (2007). Encryption-enforced access control in dynamic multi-domain publish/subscribe networks. ACM International Conference Proceeding Series, 233, pages 104-115. Pesonen, L. W., Eyers, D. M., & Bacon, J. (2006). A Capability-based Access Control Architecture for Multi-Domain Publish/Subscribe Systems. SAINT, 2006, 222–228. Peterson, M. G. (2005). Privacy, Public Safety, and Medical Research. Journal of Medical Systems, 29(1). doi:10.1007/s10916-005-1106-y PGP. http://www.pgpi.org/ Pietzuch, P. R., & Bacon, J. M. (2002). Hermes: A Distributed Event-Based Middleware Architecture. 22nd International Conference on Distributed Computing Systems, (pages 611-618). Radu, C., Govaerts, R., & Vandewalle, J. (1997). Efficient Electronic Cash with Restricted Privacy.

Schneier, B. (2003). Practical Cryptography. Wiley & Sons. Shamir, A. (1979). How to share a secret. Communications of the ACM, 22(11), 612–613. doi:10.1145/359168.359176 SourceSafe. http://msdn.microsoft.com/en-us/ library/3h0544kx%28VS.80%29.aspx Srinathan, K., Tharani Rajan, N., & Pandu Rangan, C. (2002). Non-perfect Secret Sharing over General Access Structures. INDOCRYPT 2002. Subversion. http://subversion.apache.org/ Tang, X. (2008). Personal health care and medical treatment information and protection of privacy right. Frontiers of Law in China, 3(3), 408–422. doi:10.1007/s11463-008-0019-3 Wang, C., Carzaniga, A., Evans, D., & Wolf, A. L. (2002). Security issues and requirements for Internet-scale publish-subscribe systems. Proceedings of the Thirtyfifth Hawaii International Conference on System Sciences (HICSS-35). Wang, H., & Wong, D. S. (2008). On Secret Reconstruction in Secret Sharing Schemes. 54 (1). Zhuo, C., Zhengwen, Z., & Nan, J. (2008). A Session Key Generator Based on Chaotic Sequence. Computer Science and Software Engineering, 2008 International Conference on, 3.

KEY TERMS AND DEFINITIONS GnuPG: A GNU software application that provides privacy and authentication in communications between two persons. It is mainly used as a complement for email applications. It uses the OpenPGP Message Format. It is the open-source equivalent to PGP. OpenPGP: Message Format: A standard that defines a file structure to be handled by GnuPG and similar software applications.

201

A Pervasive Polling Secret-Sharing Based Access Control Protocol for Sensitive Information

PGP: A commercial software application similar to GnuPG. It is compatible with the OpenPGP Message Format standard. Public Key Cryptography: A cryptographic paradigm in which every individual owns two bound keys: a public and a private key. What is encrypted with a key can only be decrypted with its counterpart. Public Key Cryptography has helped to evolve security and privacy in the digital era, especially on Internet. Secret Sharing: A cryptographic technique that allows regenerating a secret by computing separate pieces of information.

202

Shared Signature: A cryptographic technique that allows signing a piece of information by binding two different signers together. SSL: A protocol that provides privacy and authentication in communication between two hosts. Its last version is 3.0, which later evolved into TLS 1.0. Symmetric Cryptography: A cryptographic paradigm in which the same key is used to encrypt and decrypt information. TLS: A protocol similar to SSL. In fact it is an evolution of SSL 3.0. The current version is TLS 1.0.

203

Chapter 13

Mobile Location-Based Recommender: An Advertisement Case Study Mahsa Ghafourian University of Pittsburgh, USA Hassan Karimi University of Pittsburgh, USA

ABSTRACT Mobile devices, including cell phones, capable of geo-positioning (or localization) are paving the way for new computer assisted systems called mobile location-based recommenders (MLBRs). MLBRs are systems that combine information on user’s location with information about user’s interests and requests to provide recommendations that are based on “location”. MLBR applications are numerous and emerging. One MLBR application is in advertisement where stores announce their coupons and users try to find the coupons of their interests nearby their locations through their cell phones. This chapter discusses the concept and characteristics of MLBRs and presents the architecture and components of a MLBR for advertisement.

INTRODUCTION With the exponential increase of cell phone users in the past several years, more specifically cell phones with location-aware capabilities, the parameter of “location” has become an integral component of mobile applications. (Bellotti, et al., 2008) conducted a survey and reported that mobile Internet is permeating into different location-based applications such as train schedule, weather report, and restaurant finding. Among DOI: 10.4018/978-1-60960-042-6.ch013

the available location-based applications on cell phones, maps are the most popular means of user interface (Meng & Relchendbacher, 2008). Current mobile phones, which support a higher bandwidth and localization (Baus, Cheverst, & Kray, 2005) are paving the way for the emergence of a new class of systems, which we call mobile location-based recommenders (MLBRs). MLBRs combine information on user’s location with information about user’s interests and requests to provide useful recommendations based on location via mobile devices. Many diverse applications can benefit from MLBRs; these include health (e.g.,

Copyright © 2011, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited.

Mobile Location-Based Recommender

recommending nearby hospitals), education (e.g., recommending nearby libraries), and e-commerce (e.g., recommending clothing stores with special offers). The goal of MLBRs is to recommend mobile users with information which is of their interest and needs using their current location. MLBRs are beneficial to individuals in that they save their time in exploring the needed locationbased information by receiving automatic recommendations that meet their needs and preferences. For instance, for a user who is interested in the stores that have items of his/her interest on sale, an MLBR can recommend such stores while the user is driving and getting close to each. In this chapter, we describe a MLBR for advertisement (mainly coupons search) and present a prototype for recommending coupons. The objectives of the paper are: (a) to understand the infrastructure and components of MLBRs, (b) to become familiar with the range of technologies appropriate for MLBRs, and (c) to understand important issues for designing and developing MLBRs. The structure of the paper is as follows. The structure of the paper is as follows. In Section 2, related works are discussed. Section 3 overviews the MLBR concept and its relationship to recommender, navigation, and mobile services. Section 5 and 6 describes an application of MLBR in m-commerce and a scenario for MLBR-Coupon. In section 7 and 8 MLBR-Coupon architecture and components and a prototype MLBR-Coupon for advertisement are presented. Section 9 provides the conclusion and future research.

RELATED WORKS Among the various available MLBR applications, some are extensions of current mobile applications with the parameter of “location” as another way of injecting data into the decision process. (Yu & Chang, 2009) presented a personalized MLBR for tour planning. Sightseeing spots, hotels, restaurants, and other points of interest (POIs) to tour-

204

ists are recommended based on tourist’s location, time, and personal preferences and needs. (Yang & Wang, 2009) developed an architecture using WEB2.0 services for restaurant recommender. In this research restaurants are recommended to users based on their location that is obtained via Global Positioning System (GPS). (Hinze & Buchanan, 2006) presented a MLBR for tourists called Trip Information Provider (TIP). TIP provides a user with general information based on their location, personal profile, and their travel history once they have entered a museum. Moreover, users are informed of scheduled events such as opening hours of a museum. (Rashid, Coulton, & Edwards, 2008) presented a system which provides location-based information/advertisement for mobile users. By implementing the system in a supermarket, nearby customers are provided with the latest information on products as well as special offers, using Bluetooth. SMMART (Kurkovsky & Harihar, 2005) is another context-aware system which provides users with recommendations or promotions in a given retail store, considering user’s preferences. (Bellotti, et al., 2008) presented Magitti, a leisure guide, which automatically recommends its user a leisure activity. It predicts user’s future activity based on context and their patterns of behavior, and then recommends a useful activity considering user’s preferences. (Park, Hong, & Cho, 2007) developed a map-based personalized recommendation system, which collects context information, location, time, weather upon a mobile user request, and provides the user with a proper service on a map. The POI recommender presented by (Horozov, Narasimhan, & Vasudevan, 2006) is another mobile recommender which provides its users with recommendations on POIs (e.g., restaurant) considering their location and preferences. In this chapter, we present the architecture of MLBR-Coupon and discuss a prototype MLBRCoupon where stores coupon promotions are recommended to users. MLBR-Coupon facilitates accessing to stores with coupons on products/

Mobile Location-Based Recommender

Table 1. MBLR characteristics from application perspective Applications/Services

Features

Users

Scale

Advertisement Tourist Guide Education Navigation/POI Health

Localization Recommendation Personalization Contextawareness Adaptation

General populations Special needs Youth Students Tourists

Outdoor Neighborhood Campus City County State Country Indoor Buildings

services of interests using “location” as the main search criterion.

MOBILE LOCATION-BASED RECOMMENDER (MLBR) MLBR encompasses three main services: recommender services, navigation services, and mobile services. MLBR recommends to mobile users location-based information relevant to the application, such as location of stores, real estates, restaurants, learning resources, and recreation facilities and provides them with navigation services, i.e., routes and step-by-step directions to reach the recommended location. Figure 1 shows the relationships among the recommender, navigation, mobile, and MLBR services. Recommender services provide users with recommendation on different information. An example recommender service is store recommender, restaurant recommender, and tourist guide. Mobile services provide mobile users with information relevant to their location. For example, smart phones provide users with their current location. Navigation services provide users with navigation assistance, such as locating destinations, computing an optimal route (e.g., shortest distance or fastest time) between the origin and destination, and providing step-by-step directions to reach the destination. MLBR services are at the intersection of recommender, navigation, and mobile services. This

implies that MLBRs are services that provide mobile users with location-based recommendations, e.g., nearest shoe store to user’s current location, and navigation services, e.g., fastest route to the nearest store. A MLBR employs user’s location, obtained through either GPS or user entry on mobile devices (increasingly smart phones) to provide a variety of location-centered functions. In MLBRs, both a mobile platform and a web-browser platform can be considered for delivery of content to the user; the web-browser platform would be of interest when the user prefers not to deal with the constraint nature of mobile phone screens.

MLBR CHARACTERISTICS MLBR can used in different applications and provide users with several features. Table 1 illustrates the characteristics of MLBR from application perspective. Applications/Services. There are currently many applications that are benefiting from MLBRs. Examples of these applications include advertisement (e.g., recommending nearby stores that have items of user’s interest and needs), tourist guide (e.g., recommending nearby tourists attractions), education (e.g., recommending nearby libraries), navigation/POI (e.g., recommending POIs and routes to reach them), health (e.g., recommending nearby hospitals), and social

205

Mobile Location-Based Recommender

Table 2. MLBR characteristics from implementation perspective Technologies

Positioning

Communication

Outdoor: GPS Bluetooth DR Cell-based Indoor: RFID A-GPS Bluetooth Cell-based

Short range: Bluetooth Medium range: Cell Large range: Wi-Fi

Decision Spatial Analysis

Outdoor: GIS Indoor: CAD

Platform

Android iPhone WHERE

Mobile Device

Cell phone Smart phone PDA Laptop

networking (e.g., recommending to meet nearby friends). Features. MLBRs may include a variety of features to support the specific requirements of applications. However, regardless of the application requirements, common features in MLBRs include localization, recommendation, personalization, context-awareness, and adaptation. Localization refers to the ability to determine current location of the user. Recommendation is the ability to automatically send location-based information to the user. Personalization refers to the ability to prepare location-based information based on user’s preferences and interests. Context-awareness is the ability to recommend information based on the context, such as location (indoor or outdoor), time (day or night), and weather (rainy or sunny). Adaptation is the ability to adapt to user’s changes of behavior/preferences/interests. Users. MLBRs can provide recommendations to users with different needs and preferences. Examples users are general population using incar navigation systems, people with special needs using specialized navigation systems, youth using friend finder, students using a system for recommendations on learning resources, and tourist using tourist guide. Scale. MLBR can provide services at different scales based on the application requirements. In outdoors, MLBRs can be at the scale of a neighbor-

206

Data Outdoor: Road Sidewalk POIs Indoor: Hallway POIs

Common Functions

Geocoding Map matching Buffering Proximity Routing

Architecture

Issues

Centralized: Stand-alone Decentralized: Client/Server Distributed

Scalability Extensibility Interoperability Privacy Screen size

hood (e.g., recommending nearby stores), campus (e.g., recommending libraries within a campus), city (e.g., recommending houses within a city), county (e.g., recommending tourist attractions within a county), state (e.g., recommending cities with certain stores within a state), and country (e.g., recommending universities within a country). On the other hand, in indoors MLBRs are confined to the scale of buildings. Examples of such MLBRs are mostly limited to navigation systems for indoors. MLBR characteristics from implementation perspective are illustrated in Table 2. These characteristics include technologies, data, common functions, and architecture. Technology. Developing a MLBR requires combination of different technologies, including positioning, communication, decision spatial analysis, platform, and mobile device. Positioning technology can be categorized into two groups: those that operate in outdoors and those operate in indoors. Positing technologies for outdoors mainly include GPS, Assisted GPS (A-GPS), cellular network, Bluetooth, dead reckoning, and IP address and positioning technologies for indoor include Radio Frequency Identification (RFID), A-GPS, Bluetooth, cellular network, and IP address. Communication technology facilitates transition of information between the different com-

Mobile Location-Based Recommender

ponents of a MLBR, such as client and server(s). Examples of communication technologies are Wi-Fi, cell, and Bluetooth. Spatial/Decision analysis technology facilitates checking “locations”, “attributes”, and “relationships of features” in spatial data via several techniques to address spatial queries. Spatial/ Decision analysis technologies are categorized into indoor and outdoor. Geographic information system (GIS) is used for outdoors and is a system that captures, stores, retrieves, analyzes, and displays geographical information (Cowen, 1988). Computer-Aid Design (CAD) is used for indoors and is a system for drawing and updating maps (Curry, 2004). LBS Platform is used for implementing MLBRs. Examples of such platforms include WHERE, and Android, and iPhone. WHERE is a location-based platform developed by uLocate that supports several programming languages such as PHP, Cold Fusion, and Ruby. While WHERE provides developers with an easy and opensource implementation environment, it has slow performance and its functionalities are limited to navigation and POI finder. Android is another open-source platform for developing locationbased applications (Meier, 2009) and can be used via different platforms, i.e., Windows, Mac OS X, and Linux (i386). iPhone, which has been emerged in the past couple of years, allows connection to and utilization of third party location-based systems vendors. Examples of such vendors are Google Maps, Yahoo Maps, and Skyhook Wireless (Sadun, 2009). Despite the advantages of iPhone, such as supporting more than one programming languages and enabling access to several external geospatial resources, it suffers from some limitations, such as non-interoperability and high expenses for both developers and users. Mobile device. Since MLBR provides recommendations for mobile users, it needs to be accessible via a mobile device, such as cell phone, smart phone, PDAs (personal digital assistant), and laptops.

Data. MLBRs require two types of data: spatial and non-spatial. Spatial data for outdoors include geometry and topology of roads and sidewalks, POIs (e.g., geographic position of a restaurant), and spatial data for indoors include geometry and topology of hallways and POIs (e.g., location of a classes). Non-spatial data for outdoors include street names, road and sidewalk segments’ length, road segments’ speed limit, number of lanes, and users’ profile/preferences. For indoors nonspatial data include hallway segment’s length, POIs’ type (e.g., restrooms, drinking fountain, and room), accessibility information, and users’ profile/preferences. Functions. Common MLBR functions include geocoding, map matching, buffering, proximity, and routing. Geocoding is “the conversion of analog maps into computer-readable form” (Clarke, 2003) which is used for representing addresses on the map. Map matching is the process of finding the road/sidewalk segment (in outdoor) or hallway/ corridor segment (in indoor) on which the user is located, once his/her position through GPS or the complementary positioning service is determined. Buffering is “the process of generating a buffer zone from a point, line, or, area feature by offsetting a user-defined distance from these features” (Yeung & Lo, 2007). Routing is the process of computing a route between a pair of origin and destination based on criteria such as shortest path, fastest time, less number of intersections, and/or toll free. Architecture. MLBR’s architecture can be either centralized or decentralized. Centralized MLBRs are stand-alone systems where all computations are performed on a single device. Decentralized MLBRs are based on either client/server model, where a server is responsible for providing recommendations to the client, or distributed architecture in which more than one server supports the client with recommendations. Since in MLBRs, the client is usually a handheld

207

Mobile Location-Based Recommender

device, such as a smartphone, which has limited computational capabilities, storage, and battery life, most MLBRs are based on client/server architecture. In such an architecture, the client is responsible for presentation tasks and the server is responsible for computations and database processing. Issues. Implementing MLBRs, with aforementioned characteristics, requires addressing several issues such as scalability, extensibility, interoperability, privacy, and screen size. Scalability refers to the capability of the MLBR to handle more operations as the number of clients increases. In client/server architecture, where different clients request to MLBR for receiving location-based recommendation, simultaneously, the response time is of importance. Clients expect to receive services (recommendations) immediately. Extensibility refers to the provision of being able to include additional functions in the MLBR as they deem necessary. Interoperability refers to the ability of different components of different MLBRs to work together. For instance, MLBR should be available for different clients regardless of the platform. Interoperable MLBRs must consider adherence to standards such as OpenLS (Open Location-based Services) defined by OGC (Open Geospatial Consortium). Privacy is another issue that needs to be addressed where user’s personal information, especially location, could be available in the database on the server. Screen size of mobile devices where information such as recommendations, maps, and routes is communicated with the user is another issue to be addressed when developing MLBRs.

MLBR IN M-COMMERCE MLBRs play a pivotal role in m-Commerce (mobile commerce). m-Commerce is defined as “the activities of consumers shopping using a mobile device such as a cell phone, Personal Digital Assistant or a combination cell phone-PDA device

208

from anywhere any time connected to the Internet through a wireless network” (Zenebe, Ozok, & Norcio, 2005). MLBRs in m-Commerce are recommender systems that recommend products/ services which meet user’s preferences with respect to location and proximity to the mobile user (Horozov, et al., 2006). One prevalent application of MLBRs in m-Commerce is advertisement, where stores/vendors announce their products/ services through the Internet and consumers can access them through mobile devices equipped with geo-positioning technologies such as GPS. The mobile device (e.g., a smart phone) computes user’s location in real time and facilitates user’s location-based queries including nearby stores, routing, and directions to the stores. The advantage of MLBRs in advertisement is twofold. One, it is beneficial to consumers due to its “mobility”, “broad reach”, “ubiquity”, and “convenience” in localizing products and services of their interest (Zenebe, et al., 2005). Second, it is beneficial to companies since it allows them to broadcast their advertisements only once without incurring expenses of flyers and their distribution. An example MLBR in advertisement is MLBRCoupon which is coupons announcement with the goal of facilitating access to coupons on products/ services of interests using “location” as the main search criterion. While maps are the most popular means of submitting queries and visualizing results, screen sizes of cell phones hinder effective overlaying and presenting map and text data, simultaneously. To address this issue, the mapping and location-based nature of the service should be strictly behind the scenes unless the user explicitly requests map or direction data. On the web-browser side, a dynamic mapping interface is a suitable method of displaying data to the user, and the map could be the central delivery point for the data, both advertisement data, as well as the originating location of that advertisement data. A dual panemethod of displaying data would be required for the web-browser platform, in which the available

Mobile Location-Based Recommender

stores (with distance from the user’s location) are listed separate from the individual store data (with relevant advertisements) on the map. On the cell phone side, there should be a simple interface to display the requested information. The location information is obtained through the GPS on the cell phone, or the address is entered manually when the GPS signal is not strong enough or obscured, and then the search keyword(s) are entered. Once user’s location along with search keyword(s) are submitted, stores with matched advertisement keyword(s) are returned. A fine tuned display for each store and subsequently each advertisement is then enabled from that point, as well as an affordance for the user to go back, or to exit.

SCENARIO Mary is a subscriber of a MLBR-Coupon that provides advertisement services on her cell phone. One of the services is searching for coupons on items of her interest. The following is her experience with the MLBR-Coupon for advertisement services. Upon selection of the MLBR-Coupon for advertisement services on her cell phone, Mary requests for receiving information on coupons of the nearby stores as she drives away from her home. The MLBR-Coupon searches for the stores in her proximity and upon finding one explores it for available coupons. Once a coupon is found, the MLBR-Coupon sends the relevant information to her cell phone and displays the store on the map along with the shortest path (a preferred routing criteria pre-selected by Mary) to reach the store.

ARCHITECTURE AND COMPONENTS In this section, we present an architecture for MLBR-Coupon that provides personalized location-based recommendation on coupons The

MLBR-Coupon architecture consists of four main components: client, server, databases, and web mapping services. Figure 2 shows the architecture of MLBR-Coupon. Client- is either of mobile type or web-based type. Mobile client is specialized location-based software installed on cell phones which can obtain position data using the GPS receiver on the phone and passes it on to the server to be used in other functions. In this way, all functions would be stored on the phone, and transmission times would be short, as the only transmission needed would be short text and small images. The other client type is a browser environment where location information is simply placed in a standard HTTP request, and the entire code for displaying the information would be returned to the device, in a form similar to HTML, Cascading Style Sheets (CSS) and JavaScript on webbrowsers. However, with this client, the amount of exchanged data for each request is large, even though some elements of the page such as images (logos, separation lines, or background images) are cached. Its advantages are that the entire application resides on the server and any changes to it would simply be realized by the client the next time they are accessed. This is in contrast to the mobile client where the client software must be updated constantly to be consistent with the standards set by the service provider. The overhead differences between the two clients are significant, and there may be some advantages or disadvantages in using one client type over the other. Server- is the main component of MLBRCoupon which consists of three major modules: the web server, which provides standard static markup text to the web client or mobile client; the serverside scripting language; and the database(s). The database module stores all of the non-presentation oriented data and parses data from the database along with presentation data into a cohesive package (i.e., HTML, XML or otherwise) delivered to either the web client or the mobile client. Whether the client is a web client or a mobile client, the

209

Mobile Location-Based Recommender

pages or structured text are created on-the-fly when requested from the user, upon searching the system. The web server along with the server-side scripting language play an important role, since it is the conduit to retrieve data from the database before it is sent to the user. In addition to using a server-side scripting language to retrieve data and display it in a meaningful way, applying a framework system which would aid with the data extraction, as well as displaying information in a proper style is essential. As mentioned earlier, in addition to the client-side software that can access the service, a web-based version is also needed. This is similar to the web-based platform for delivery in the mobile device, where the web application would be downloaded anew, in standard web style, upon the user’s request. When there is no other option, other than a traditional compiled application which resides on the client’s device, this method is preferable considering the speed of Internet and modern computers in general. In this situation, as was mentioned earlier, the user would be prompted to identify their location, or a location in which they would like to search around. The system, through geolocation via IP address, would estimate the user’s closest metropolitan area, which may be beneficial for some users, who choose to search around their closest area. Once a location is determined, the user’s queries are sent to the server. The server sends the query to the spatial database, which resides on a web mapping service’s database (e.g., Google Map), for geo-computations, if it is deemed necessary. Then, it sends the query to the local database in order to retrieve the requested coupon. Upon completion of all computations and retrieval, a webpage with a dual pane, one showing a list of stores and the other showing marked stores on a map, is displayed to the user. Web Mapping Service contains a map database and supports functions such as mapping, geocoding, overlaying, routing/directions, and proximity. Google Maps is an example Web Mapping Service.

210

A scalable database structure is imperative to handle increase in data volumes. There are two options to handle geospatial data, which is an important aspect of this database. One option is by using the server-side scripting language, which takes time to calculate complex computations, and another option is by using geospatial plugin for databases. The latter option has a higher performance than the former option, which is extremely important considering numerous concurrent requests.

A PROTOTYPE MLBR-COUPON A prototype of the MLBR-Coupon for advertisement was developed and tested. The database schema of the prototype is shown in Figure 3. The logic of how the prototype searches coupons based on user’s location is shown in Figure 4. The prototype has three main interfaces: one for the mobile user on a cell phone, one for the web user, and one for the administration side. Each interface has a typical use-case associated with it. The mobile user would be in the field, looking for advertised coupons. They would activate the phone, search for deals nearby their location using a keyword, find the deals, and utilize them at the point of sale, or store them for later use. The web user would be interested in deals around a certain location, perhaps their current location which they would enter into the system. They would search for deals that match their interests by entering a search keyword, and the deals would be returned in a hybrid map interface, involving both map and coupon data. The administration interface is for businesses to enter and manage their data in order for it to be searchable on the client-side interfaces. For the prototype, we chose rapid development rather than a best-case scenario for the end user. This allowed us to focus on the structure of the system rather than the underlying necessity of learning a mobile device programming framework, such as J2ME, Java’s mobile framework

Mobile Location-Based Recommender

running on many consumer-grade cellular phones. A framework called WHERE, presented itself as the framework of choice, considering that it fit all of the requirements we had identified for rapid prototyping. WHERE is a platform developed as a mobile web browser and compared to common mobile web browsers, it supports a location parameter injected into the POST data by the application (“WHERE Developer”, 2008). This location data is obtained by the GPS receiver built into the phone, and delivered using a programming API to the application. The steps of how WHERE could be used for developing MLBR applications/services are as follows. Once the WHERE application is started, the phone begins to obtain a GPS signal in the background. The user is then displayed an index of applications that they have added, which are essentially links to specialized WHERE URLs. Upon selection of the MLBR-Coupon, the client phone checks to ensure that the current location has been obtained by the GPS. If no GPS data is available, the client phone will prompt the user to enter an address. Once a location is obtained, the coordinates (e.g., latitude and longitude) are passed through POST data in the HTTP request to the server for the WHERE application. The server takes the POST data and processes it using the server-side scripting language to deliver content relevant to the user’s location. Then, the location is sent in the form of POST data upon any new page request. If the user requests for information on a store’s coupon, WHERE explores MLBR-Coupon database for coupons, and if they are found, the pertained information along with a map is presented to the client. WHERE provides several services to content providers to help them develop applications that are helpful to users and are easily accessible. Application directory is a service mostly requested by the user. Other features of WHERE include mapping and routing. WHERE supports several APIs that provide multi-point mapping for applications. This is also the case for routing via car or on foot.

There were several languages involved in producing the application, but the main logic and the processing was the open-source server-side scripting language PHP (“PHP”, 2008). It operates well with the Apache Web server (“The Apache Software Foundation”, 2008) and therefore is optimized for web language parsing, but can also be run from the command line, and therefore could be utilized in the future for off-line operations such as database maintenance. PHP is a very versatile language for writing web applications and allowed us to create other web pages in XHML/JavaScript as well as the Jin language, employed by WHERE. Jin, however, is not standard, although it is similar to HTML (not XHTML). The major difference is that it is much more demanding with respect to proper syntax. If the syntax is not correct, it will simply crash the page (or the phone, if testing on a device), and does not degrade gracefully similar to standard browser-based HTML pages do. It has some rudimentary JavaScript-like functionality, but it is basically used for displaying pages. Despite all of its shortcomings, WHERE (utilizing Jin) is open source and is used to implement a quick prototype. CakePHP (“CakePHP”, 2008) is a PHP framework which utilizes the MVC (Models, Views, Controllers) paradigm. The framework was used to develop the administration page in order for registered users to add, view, modify and delete products, services, stores, and company information. CakePHP’s method of handling the interaction between the Models, Views and Controllers is not unique but enabled much functionality with little effort. Issues like error handling, data verification, and data cleaning were all performed by the CakePHP framework. Google Maps (Google Maps API Concepts”, 2008) is a particularly integral part of the clientside web interface. Due to the need for a hybrid mapping interface, it was important that the map takes up a large portion of the display to display all of the correct spatial information along with the relevant information to the search keyword.

211

Mobile Location-Based Recommender

The Google Maps API which is an open source API and facilitates the flexible design was used to display the map, the locations of the user, the relevant store locations, the contextual information for each store location (coupon information), and the directions from the user’s location to the store locations. The Yahoo User Interface (YUI) (“The Yahoo! User Interface Library (YUI)”, 2008) libraries are open source web libraries which take the burden of cross-browser functionality planning and debugging from the shoulder of the developer. YUI obfuscates browser quirks so that the developer can focus on what they want something to do in the browser, instead of how it will look on different browsers. We used YUI for page layout, AJAX functions, animation, and display of layered features. The layout utilized YUI Grids, which split up the page in predetermined “grids” using CSS. Using AJAX, we implemented a bookmarking and printing system for the user to remember and print their coupons (bookmarking deprecated in favor of printing). Using the animation features of YUI, we implemented a small animation which directed the user’s eye to the text-entry bar to enter their address. PostGIS (“What is PostGIS?”, 2008) is the database extension that is installed into the PostgreSQL which supports geo-computations. In order to calculate which stores are near a user, PostgreSQL, with PostGIS’ support computes a bounding box around the user. Then each location found within the bounding box is calculated for two reasons. Firstly, to ensure that the location actually is inside the radius specified by the user (some could be in the corners of the box), and secondly, is to sort and display the locations by distance. These calculations are done by the database, since the database is compiled and is much faster at geospatial calculations than the server-side scripting language which by nature, is not compiled. There were several key queries from the database that demonstrate the level of complication

212

derived from using such a complex database, and complex calculation of geospatial data. The first and foremost query, which is used on both the mobile device application as well as the web application, is the search query. The query has several variables which change per query and change the output of the query. They are search Longitude and Latitude, search Radius Distance, and Search Keyword. As mentioned earlier, the query first searches for entities with latitude and longitude within the bounding box of the current position’s latitude and longitude, plus and minus north/south, east/west given the search radius distance. This bounding box is created by using the “expand” function. Then it does the more intense calculation of ensuring the resulting locations are located inside the actual radius, not just the bounding box. This precise calculation is done by the “distance_sphere” function. It would be possible to do more precise calculations using the “distance_spheroid” function, but for the precision that we are looking for (0.1 miles), it is not necessary. The Search Keyword is self explanatory, but how it is used in the database needs explanation. For each product and coupon row, there is a column which keeps an extended index with the abbreviated version of all the words (e.g., “bottles” would be converted to “bottle”) in the description and title of the row. This is updated on creation, and on edit, all performed using triggers in the database. The Search Keyword is converted to a query statement by creating abbreviated versions of the words in the search string, and then comparing it to the index. This is done using the “ts_query” function which converts the keywords to a “stunted” form. The comparison is then between the new “ts_querried” form of the search and all of the products and coupons from all of the locations returned in the first part of the search, it then returns the locations containing each of the results. If the element found is a product, only the store location information will be returned, but if the element was a coupon, both the store location and the coupon information will be returned. The

Mobile Location-Based Recommender

reason for this is that the service is a deal finder, not a product finder. The reason businesses would want to input their products, is so that they can get more matches for their particular store, but there is really no reason for the product to be returned.

CONCLUSION AND FUTURE RESEARCH MLBRs are extremely important in the future of mobile computing. User can find their desired information based on their location. In this chapter, we presented an infrastructure and components of MLBRs and a prototype of MLBR-Coupon which can help users to find appropriate stores, with desired coupons, based on their location. Future research related to the prototype can focus on supporting users with both further options and better quality of service. To do this, one area of future research would be using user’s preferences and their feedback on their previous experiences for various purposes, such as providing user with the required information based on their purchasing habit and preferences. Also, extending the system with a social networking component will allow users to share their experiences with each other. Furthermore, the system could be used as a shopping schedule assistant, i.e., the user sets up their schedule, such as daily schedule, on the cell phone, and the system can provide them with information on the nearby stores on a specific day (and time) considering their schedule.

ACKNOWLEDGMENT The authors would like to thank Mathew Mendick for his assistance in developing the prototype in this project.

REFERENCES Apache Software Foundation. (n.d.). The Apache Software Foundation. Retrieved 8 Sept. 2008, from http://www.apache.org Baus, J., Cheverst, K., & Kray, C. (2005a). A Survey of Map-Based Mobile Guides. In Mimg, L., Zipf, A., & Reichenbacher, T. (Eds.), Map-based Mobile Services. Springer Berlin Heidelberg. doi:10.1007/3-540-26982-7_13 Baus, J., Cheverst, K., & Kray, C. A. (2005b). Survey of Map-Based Mobile Guides. In Mimg, L., Zipf, A., & Reichenbacher, T. (Eds.), Map-based Mobile Services. Springer Berlin Heidelberg. doi:10.1007/3-540-26982-7_13 Bellotti, V., Begole, B., Chi, E. H., Ducheneaut, N., Fang, J., Isaacs, E., et al. (2008). Activity-Based Serendipitous Recommendations with the Magitti Mobile Leisure Guide. Paper presented at the the twenty-sixth annual SIGCHI conference on Human Fctors in computing systems, Florence, Italy. CakePHP. (n.d.). CakePHP. Retrieved 8 Sept. 2008. Clarke, K. C. (2003). Getting Started with Geographic Information Systems (4 ed.)Upper Saddle River, NJ: Prentice Education. Cowen, D. (1988). GIS versus CAD versus DBMS: What Are the Differences? Photogrammetric Engineering and Remote Sensing, 54(11), 1551–1555. Curry, S. (2004). CAD and GIS: Critical Tools. Retrieved 10 Oct. 2009, from http://images.autodesk.com /apac_grtrchina_ main/files/4349824 _CriticalTools.pdf Google Maps, A. P. I. Concepts. (n.d.). Google Maps API Concepts. Retrieved 8 Sept. 2008, from http://code.google.com/ apis/maps/ documentation/

213

Mobile Location-Based Recommender

Hinze, A., & Buchanan, G. (2006). The Challenge of Creating Cooperating Mobile Services: Experiences and Lessons Learned. Paper presented at the the 29th Australasian Computer Science Conference, Hobart, Australia.

Sadun, E. (2009). The iPhone Developer’s Cookbook: Building Applications with the iPhone SDK: Pearson Education Inc. What is PostGIS? PostGIS Retrieved 8 Sept. 2008, from http://postgis. refractions.net/

Horozov, T., Narasimhan, N., & Vasudevan, V. (2006a). Patent Application Publication.

Where Developer. (n.d.). Where Developer. Retrieved 8 Sept. 2008, from http://developer.where. com/ jin/dev welcome.jin

Horozov, T., Narasimhan, N., & Vasudevan, V. (2006b). Using Location for Personalized POI Recommendations in Mobile Environments. Paper presented at the the International Symposium on Applications on Internet Kurkovsky, S., & Harihar, K. (2005). Using Ubiquitous Computing in Interactive Mobile Marketing Meng, L., & Relchendbacher, T. (2008a). The State of the Art of Map-Based Mobile Services. In Meng, L., Zipf, A., & Winter, S. (Eds.), Map-Based Mobile Services. Springer Berlin Heidelberg. doi:10.1007/978-3-540-37110-6_1

Yahoo. (n.d.). Yahoo! User Interface Library (YUI). Retrieved 8 Sept. 2008, from http://developer.yahoo.com/yui/ Yang, F., & Wang, Z. M. (2009). A mobile locationbased information recommendation system based on GPS and WEB2.0 services. Transactions on Computers, 8(4), 725–734. Yeung, A. K. W., & Lo, C. P. (2007). Concepts and Techniques of Geographic Information Systems (2 ed.). Upper Saddle River, NJ: Pearson Prentice Hall.

Meng, L., & Relchendbacher, T. (2008b). The State of the Art of Map-Based Mobile Services. In Meng, L., Zipf, A., & Winter, S. (Eds.), Map-Based Mobile Services. Springer Berlin Heidelberg. doi:10.1007/978-3-540-37110-6_1

Yu, C. C., & Chang, H. P. (2009). Personalized Location-Based Recommendation Services for Tour Planning in Mobile Tourism Applications. In Proceeding 10th International Conference, EC-Web 2009, Linz, Austria, pp. 38-49.

Park, M. H., Hong, J. H., & Cho, S. B. (2007). Location-Based Recommendation System Using Bayesian User’s Preference Model in Mobile Devices. Paper presented at the 4th International Conference, UIC 2007. PHP. Retrieved 8 Sept. 2008 from http://www.php.net

Zenebe, A., Ozok, A. A., & Norcio, A. (2005, 22-27 July 2005). Personalized Recommender Systems in E-Commerce and M-Commerce: A Comparative Study. Paper presented at the HCII 2005, The 11th International Conference on Human-Computer Interaction, Nevada.

Postgre, S. Q. L. Retrieved 8 Sept. 2008, from http://www.postgresql.org/ Rashid, O., Coulton, P., & Edwards, R. (2008). Providing Location Based Information/Advertising for Existing Mobile Phone Users. Personal and Ubiquitous Computing, 12(1).

214

KEY TERMS AND DEFINITIONS Location-Based Services (LBSs): Prepare location-centric information and deliver information to user’s current location. Geo-positioning sensors, to determine user’s current location, and wireless communications, to deliver information, are the most important components of LBSs.

Mobile Location-Based Recommender

Recommender: Any system that provides recommendation based upon user’s preferences. Mobile Location-Based Recommender (MLBR): A recommender that provides locationcentric recommendation and delivers recommendation to user’s current location. Navigation: The process of moving from one location to another.

M-Commerce: Refers to the process of conducting commerce through wireless communications and mobile devices. Mobile Computing: Refers to the development on and utilization of mobile devices. Location-Based Social Networking (LBSN): A LBS for Social Networking (SN) in that members of the network share information based on their current location with one another.

215

216

Chapter 14

Success Cases for Mobile Devices in a Real University Scenario Montserrat Mateos Sánchez Universidad Pontificia de Salamaca, Spain Roberto Berjón Gallinas Universidad Pontificia de Salamanca, Spain Encarnación Beato Gutierrez Universidad Pontificia de Salamanca, Spain Miguel Angel Sánchez Vidales Universidad Pontificia de Salamanca, Spain Ana María Fermoso García Universidad Pontificia de Salamanca, Spain

ABSTRACT Mobile devices have become a new platform with many possibilities to develop studies and implement projects. The power and current capabilities of these devices besides its market penetration makes applications and services in the area of mobility particularly interesting. Mobile terminals have become small computers, they have an operating system, storage capacity so it is possible to develop applications that run on them. Today these applications are highly valued by users. Nowadays we want not only to talk or send messages by mobile terminal, but also to play games, to buy cinema tickets, to read email… We can bring these capabilities in our pocket. The University may not be aware of this fact. The students, due to their age, are the main users and purchasers. In this sense, this article will present three applications we have developed for mobile devices. Nowadays they are being used in Universidad Pontificia de Salamanca. All of them work on a university scenario and use different kind of services. DOI: 10.4018/978-1-60960-042-6.ch014 Copyright © 2011, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited.

Success Cases for Mobile Devices in a Real University Scenario

INTRODUCTION In recent years mobile devices (mobile phones, PDAs, ...) have suffered a large and fast evolution; this fact has caused companies and individuals demanding information and services over mobiles environments. These devices have evolved considerably since their birth; the great progress came in 2001, when appeared the first devices with LCD color display. In same year but in Japan it was released the third-generation phones (3G) that they were based on UMTS (Universal Mobile Telecommunications System); main innovation was the incorporation to the device of a second camera to video calls. Nowadays these 3G devices are the most used. Current mobile devices are very powerful. They are capable of transmitting, receiving and storing information, connecting to networks and running applications. Therefore, they are a very interesting platform for new research projects. University may not be aware of this fact. Students are the main users and purchasers. In this sense, this article will present three applications we have developed for mobile devices. They are being currently used in Universidad Pontificia de Salamanca. All of them work in the university scenario and they use different type of services. Structure that we will follow in our chapter is: first we are going to show with data that there is a continued demand for mobile devices. Market penetration and demand of applications and services for this type of devices are growing up. We’ll continue with a summary of the most important mobile services: voice services, messaging, multimedia, video, applications that run over mobile devices and data services. After that, we’ll emphasis on location based services (LBS) which is the main technologies available to carry out the location: cells and GPS (Global Positioning System). Then we will describe the main technologies to develop applications for mobile devices. Neverthe-

less mobile devices have some constraints which are necessary to be on mind when developing applications or services for mobiles. After that, we’ll explain some study cases that we have developed and implemented in a university scenario; next paragraph we are going to explain briefly these three applications. The first application is MovilPIU: it is a mobile application that provides to students the same services that those provided in UIP (University Information Points). Some of these services are free (request student record by example) and other services are paid (such as request certificates, purchase dining tickets). Also we’ll explain how we have resolved the payment of these services. MovilPIU is a data service and it is accessible through browsing. The second one that we describe is MoBiblio: we have developed this application to improve and to speed up the management of basic services of a library; exactly loan, return and renewal of books in the library of Universidad Pontificia of Salamanca. It is based on both message services: push and pull-push; besides it uses data services through browsing in order to provide other services such as access to the library catalogue. HouseMobile is the third application that we are going to study; help student to find an accommodation in Salamanca. It is based on location services; the application displays maps with both the position of the student (mobile device) and the position of accommodation searched; also, it provides to the user the guidance over a map from her current position towards the position of accommodation. Finally, we will finish describing what improvements we are going to incorporate to the above applications. Besides we will describe the actions that we are going to undertake at the present and in the future as regards mobile applications and services: Android and new development frameworks.

217

Success Cases for Mobile Devices in a Real University Scenario

Figure 1. Mobile phone penetration in the European Union (October 2007, lines per 100 inhabitants)

BACKGROUND

Figure 2. Mobile phone penetration in Spain

Mobile Device Expansion Last decade the use of mobile phones has suffered a large growth and its market penetration has been spectacular. For example, in Spain in 2007 was exceeded for the first time the 100 lines per 100 inhabitants (Comisión del Mercado de las Telecomunicaciones, 2007). Specifically in 2007 Spain reached 109 lines per 100 inhabitants (see Figure 1). A few years ago it was unthinkable to reach the barrier of 100 lines. In the Figure 2, it can be seen a Pyramid Research survey in 2007 (Pyramid research, 2007). This study shows the dramatic growth of mobile lines in Spain. This growth has not happened so fast ever with other technologies, and it notes that the market is at a very high saturation levels.

de Paz, & Corchado, 2009). All existing applications are based on one or more of the following basic services (Open Mobile Alliance, 2008) (Telefónica, 2007): •

Mobile Services The enormous market penetration of these devices has caused that the characteristics and possibilities of them are exploited to provide new services and new applications (Sánchez Vidales & Beato Gutierrez, 2008) (Bajo, et al., 2009) (Bajo, de Paz,

218

•

Voice service. It was the first service in mobile terminals. It is mainly used for communication between people and it is the main service. This service justifies the possession of a terminal. Nowadays, it represents 80% (Gartner, 2007) of operators business. Short Message Service. Popularly it is known as SMS. This service allows sending text between mobile terminals with a maximum of 160 characters. Despite being short text messages, the service is the most

Success Cases for Mobile Devices in a Real University Scenario

•

•

popular in history of mobile telephony. Gartner (Gartner, 2007) says that in 2006 936,000 million SMS were sent in the world and it is expected to reach 2.3 trillion messages in 2010. Multimedia Message Service (MMS). Much of current mobile phones can store and send ringtones, logos, photographs, graphics or music. These have become small multimedia consoles. Multimedia message reception is not as popular as sending SMS. On the one hand because not all terminals have this service and on the other hand because of the price. In 2006 there were sent 8300 million multimedia messages (less than 1% of total) and in 2010 they will reach 47,400 (less than 5%) (Gartner, 2007). Location service. This service is based on the nature of GSM (Global System for Mobile Communications) network (Dru & Saada, 2001) (Schiller & Voisard, 2004). The coverage of a network operator is established on the basis of cells, the size of the cells is dependent on the density of antennas. In urban areas where the number of antennas is large, the size of the cells is smaller; in this case the location is established with an accuracy of about 100 to 500 meters. Obviously this is less accurate when we are on roads or populations with a lower density of antennas.

Location service is available on all phones, but its precision is not very good and therefore the uses of this service are very limited. However, at present time there are mobile devices already incorporating GPS (Global Positioning System). GPS is a satellite navigation system that can determine the position of an object with accuracy up to centimeters, making it much more interesting. •

Video services. 3G/HSDPA (High-Speed Downlink Packet Access) terminals allow

•

•

video calls and audio in real time. In addition, this service lets you send, receive and play video. Using this service, for instance, you can access to video-monitoring circuit through the mobile device or watch television. In any case, video service is not very popular, but in future, experts are betting on it. Mobile software applications. Mobile terminals have become small computers. They have a powerful processor, an operating system and storage capacity that make possible to run mobile applications on it. To develop this software we can use languages like J2ME (Java Micro Edition) (Sun. JavaMe) (de Jode, 2004), C++ or C# (Wigley, Moth, & Fod, 2007). These features expand significantly the number and type of projects can be done on this new platform. Data service. The last generation of mobiles has the possibility of Internet connection. So, this connection can be the base for other services. For example, e-mail and browsing Web are now two of the main applications of data service and it is expected to grow in the future.

Location-Based Services (LBS) Location service is a service increasingly used to provide applications and services because users request services fit to their location context. LBS (Schiller & Voisard, 2004) use the actual position of the terminal to provide the service. With LBS the communication and interaction is possible in two ways; the user provides his actual context to the services provider, that is, he provides his position in an automated way and his preferences and necessities; then the provider of services provides to the user the requested information fitted to her necessities and to her actual positioning. There are several definitions of the locationbased services concept; for some authors (Schiller

219

Success Cases for Mobile Devices in a Real University Scenario

& Voisard, 2004), LBS is a service of information that is available through mobile devices and it uses the positioning of the device to provide the service. With LBS it is possible to develop services and applications available through the mobile devices and these applications could give answers to questions such as where am i?, what are there nearby? and how can i go? Some possible applications of the LBS (Dru & Saada, 2001) are for example emergency services to locate a user in situations where he cannot give his position, tourist services to indicate user the nearby interesting places or tracking services to use it in packet transport, etc. The first LBS was message-based services, SMS and MMS (Fangxiong & Zhiyong, 2004); these services were very limited because of technological restrictions. The terminals that supported these services were simple, with poor user interfaces that did not offer many possibilities for presentation of information. Nowadays LBS are developed to be supported by powerful terminals and others technologies (Lei & Hui, 2006) such as WAP browsing (Open Mobile Alliance, 2008), JavaMe (Sun. JavaMe) or i-mode (DoCoMo, 2005). The LBS has an add-value when these interact with maps services to provide geographic information; by example, to provide guidance information between the location of mobile device and others geographic points, or to provide services that are about specific distance walking from the position of the terminal. The LBS emerge (Shiode, Li, Batty, Longley, & Maguire, 2004) from 3 technologies: mobile technology, Internet and geographic information systems with spatial databases. Considering who is the entity that begins the service, the LBS could be classified into two groups: pull-services and push-services. With pull-services, the user carries out the request for the service. With push-services the information or service is not directly requested by the user.

220

All location-based service has four components (Schiller & Voisard, 2004): •

•

•

•

Mobile device. Devices from which the user requests services and where the results are returned. Mobile communications network. Mobile phone network through which the users send from the mobile device both the user data and the service request. Also through it, the answer to information requested by the user is received. Positioning component. With this component is possible determining the device/ user position. The position could be obtained in several ways. Service-data provider. This component is the responsible of processing the request service and it also provides different services, such as calculating the position, finding a route, finding specific information of interest nearby for a user, etc. Provider may content the requested information by the user, but this information usually will be requested by the data provider to others contents providers (yellow pages, directories, etc).

There are different methods or technologies to compute the positioning of a user/mobile device; we should choose one technology or other depending on the requirements of the application to develop: •

Network-based positioning (Kaaranen, Ahtiainen, Naghian, & Niemi, 2001). This technology uses the mobile communications network. The positioning is carried out by the base stations; the devices send signals to the base station and this is detected by them. It is possible to calculate the positioning because the coordinates of base stations are known and it is possible to measure (or at last to approximate) the

Success Cases for Mobile Devices in a Real University Scenario

•

•

distance between the terminal and the base station. To compute the distance, three or more received signals should be measured. There are different methods to measure the positioning: AOA (Angle of Arrival), TOA (Time Of Arrival), E-OTD (Enhanced Observed Time Difference). With this technology the accuracy of positioning is variable (between 50 mts and kms); at urban zones the accuracy is higher, but in rural zones it could be lower. Terminal-based positioning (Agency, 2002). It is also named as positioning by satellite. This technology uses an infrastructure of satellites of earth orbit and a receiver terminal. Terminals to compute the position are based on the received information through the signals of radio of 3 or more satellites. This technology has a higher accuracy (between 5 mts and 40 mts) than those based on network mobile positioning, but its use is restricted to outdoor. It is independent of mobile phone network. GPS is the most extended but there are others such as Glonass or Galileo. Local positioning (Krishnamurthy & Pahlavan, 2004). This is a positioning method oriented to constrained areas; it is based on transmission of signals to short distances. It is basically used for locationbased services in indoors such as shopcenters, museums, super-stores, etc. Some of these methods are based on WLAN (Wireless Local Area Network), Bluetooth or RFID (Radio Frequency Identification).

Technologies of Development Despite the fact that the mobiles devices and terminals have advanced considerably, nowadays they have some restrictions yet; we have to consider these restrictions when we are going to develop them. The mobile devices (pda, phones,...) have usually low-power computing, small memory

size, small screens with low resolutions, limited capacities of input, run-time depending of the battery, low and variable bandwidth, among others. So, it is necessary new technologies to develop and to exploit applications and services for mobile phones. Some of this technologies are WAP (Wireless Application Protocol), J2ME or JavaMe (Java 2 Micro Edition) or the development platform of Windows Mobile among others. WAP (Open Mobile Alliance, 2008)(WAP 2.0. Technical White Paper, 2002) is an Open International Standard for applications that uses wireless communications. It is the specification of a set of protocols to normalize the way in which wireless devices can access and interact with services and information easily and quickly. WAP is oriented to mobile devices with displays restricted, small keyboards and low bandwidths. This protocol allows to operate to applications and services over a large number of networks (CDMA -Code Division Multiple Access-, WCDMA -Wideband CDMA-, UMTS -Universal Mobile Telecommunications System-, GSM -Global System for Mobile Communications-, …). It is basically oriented to present contents in the terminal. The version 2.0 presents important improvements with regard to version 1.0. One of these improvements is a better design of content over portals using XHTML and WCSS. WAP is supported by most of mobile devices having a micro-browser. Another possibility is to develop the services using J2ME or JavaME (Sun. JavaMe) (de Jode, 2004) (Lei & Yin, 2004) (Yuan, 2004), that is, with the “mini” version of Java. It is a Java platform specially oriented to devices with limited capacities over the personal computer version. It is oriented to the develop applications running on mobile phones, although it is possible to develop client/server applications which interact through network services. On the other hand, in order to develop applications running on mobile devices with terminals based on Windows Mobile (Wigley, Moth, & Fod, 2007), it must be possible use the platform

221

Success Cases for Mobile Devices in a Real University Scenario

of Microsoft using Microsoft Visual Studio and Windows Mobile SDK; with these technologies, it is possible to produce software in native code (Visual C++) and in managed code (Visual C#, Visual Basic .Net). The Windows Mobile Platform offers features such as data connectivity seamless and it has enhanced security.

•

•

ACADEMIC APPLICATIONS FOR MOBILES IN A UNIVERSITARY SCENARIO Next, we are going to expose three projects1 based on different mobile services and that they have been development to provide several services to the university community.

MovilPIU This first project is an academic application. Students can access a range of services offered by the University (student record requesting, obtaining certificates, purchase of dining tickets,...) using their smart card in the UIP. Our goal is to facilitate students to access to these services. MovilPIU allows access to these services from a computer, a mobile phone or a PDA. For the project will be necessary to achieve:

Figure 3. MovilPIU Architecture

222

Identify the student: Nowadays the University community has a user account to access to computers. Through this account the user is identified and authenticated. The same account will be used to identify users in MovilPIU using Active Directory. Study and implementation of services: MovilPIU must be able to present information to the user according to the device used to having access to it. We will use the following formats: WML, XHTML and HTML (Wugofski, Lee, Soo Mee, Watson, & Mee Foo, 2000).

Moreover, we need to work with the same information (databases) that we are now used by the UIP, so that the UIP will continue to function without making any changes. •

Payment: Some services in the UIP need a pre-payment. For example, if they wish to apply for a mark certificate, students will have to pay the amount of the fee. This payment will continue to be. We use the rules defined by Spanish banks for this payment and we use a Web application to manage this function.

Success Cases for Mobile Devices in a Real University Scenario

Figure 4. Mobile and Web user identification

Figure 5. MovilPIU application

Architecture MovilPIU The architecture used in MovilPIU is a three-tier design: Model, View and Controller (MVC). Communication between components, controller and model is through SOAP messages. Really the model is a Web service (see Figure 3). The answer to the user is obtained by transforming the XML information retrieved from the model using XSLT stylesheet. Different CSS are applied depending on user-agent. The supported formats are HTML, XHTML and WML.

Results MovilPIU As we can see in Figure4, the user must access via mobile or Web. After that, depending on the

user profile, it will be possible to execute different academic functions (Figure 5). MovilPIU is divided into three parts: “Academic Information” (Figure 6), “Certificates Request” (Figure 7) and “Dining tickets Request” (Figure 8). The application works by using the navigation, via mobile and via web, through links. Parts in which the application is divided are described below.

Academic Information The application allows students to request their academic record. The student can choose between the different graduates that he has studied in the University.

223

Success Cases for Mobile Devices in a Real University Scenario

Figure 6. Academic Information

Figure 7. Certificates request

Figure 8. Dining tickets request

224

Success Cases for Mobile Devices in a Real University Scenario

Certificates Request In this option MovilPIU allows students to request certificates for their qualifications in the University.

Dining tickets Request This option allows to buy dinning tickets. Not only students but also teachers and staff can access to this option and to buy dining tickets. The purchase is recorded and is sent to the bank at the end of the month. The price is showed near to the kind of ticket.

MoBiblio Nowadays most of libraries are computerized; its management and access to its catalogue are both carried out digitally; otherwise, most of them use

the Web to provide some of their basic services, such as general information request, catalogue access, queries about the state of the loans and reservations and renewals of loans and reserves, among others. Recent studies have concluded that only 40 per cent of people have available an Internet connection, and therefore, these services could be available for these people from where they are connected to Internet. If we consider that 100% (Pyramid research, 2007) of users usually use mobile phones, if the basic services could be available on mobile devices, then these services could be available to any user at any time in anywhere (Figure 9). We propose the architecture of a platform to provide the basic-services of a library using mobile devices, which are at the same time so advanced and so restricted. Besides, these services are developed and implemented for a particular library2 using the architecture proposed. With this

Figure 9. Schema with the state of a library before and after Mobiblio

225

Success Cases for Mobile Devices in a Real University Scenario

Figure 10. Push message schema to provide alert services

platform basic-services of a library could improve in dramatically in quality and available. The platform proposed has a subsystem of messaging to provide through the mobile phone some basic-services that nowadays libraries provide through traditional post or e-mail. Specifically alerts or warnings services related to loan service, for example the expiry date of a loan, or availability of a reserved book/resource (Figure 10). From such alerts/warnings, the user could carry out several actions; the user could return the loaned resource/book or in case of an alert about the reservation go to the library to loan the book that it is available. To develop this subsystem we have used a push message service, in which the library is the actor that initiates the communication. In these cases the user should have been signed up for the services previously in order that the library could have both the mobile phone number and the user authorization. In the case of expiration date of a resource, the user has the option of renewing the resource;

226

this is possible sending another SMS, so in this case the communication is established bilaterally. Another bilateral service that it has been implemented is the renewal of a loan and the query of state of loans and reserves of a user (Figure 11). To develop these services it has been used a pullpush schema; the user is the actor that initiates the communication, and the library answers back with another message showing if the renewal has been possible or not. In case of query of information, the queried information will return. The platform has other subsystem to provide browsing services through the development of a WAP portal (Figure 12). This subsystem has been implemented because, there are several services more suitable for the user using browsing than messaging; some of them are querying the catalogue, updating the user information, or requesting general information about the library.

Success Cases for Mobile Devices in a Real University Scenario

Figure 11. Pull-Push message schema to request actions or information

Figure 12. Schema to provide services by means browsing (to query catalog, to request general information, ...)

227

Success Cases for Mobile Devices in a Real University Scenario

Architecture The architecture proposed to develop and to implement this platform (Figure 13) is based on MVC (Model-View-Control) pattern of design; this pattern separates the part that implements the business logic from which generates the presentation of contents and from the component that takes on all the client requests. With this architecture

each component is independent of others, so to carry out a migration of this platform to another library, only it would be necessary to change the data access. The model handles the access to the database using JDBC (Java Database Connectivity), so it makes the searches, insertions and updates of the data. Controller has been implemented by means of a class that extends the HttpServlet class. This

Figure 13. Proposed architecture to develop the platform to provide basic services of a library.

228

Success Cases for Mobile Devices in a Real University Scenario

Figure 14. Display with a WAP page with general information about the library (a) and with the main menu (b)

component receives request from users and after this and depending on the type of these requests, it carries out a connection through of the factory of models and it carries out the calling to the model. The view will generate the contents that are displayed on the mobile phones in a dynamic way using JSP (Java Server Pages) technology, All messaging services have been implemented using the same architecture, being the model component, the same that it has been used for developing browsing services. With push services, in which the library is the actor that it initiates the communication, the request to the controller is carried out by the master class periodically (for example, one time for each 12 hours); the master class carries out the request about which are the loans out of dated or which are the available reservations. In this case, the view will not generate jsp pages, but it will generate messages that they are transmitted to mDirectSender (Tempos21. MDirect, 2007) and this one will send them to the phone operator and this one will send them to the user. In case of pull-push services, in which the user is the actor that initiates the communication, it will be the servlet mContentService (Tempos21. MContent, 2007) which receives the SMS from the user through phone operator; after that, the back message is built by the view with the requested

data to the model; finally the message will be send to the phone operator by means of mDirectSender (Tempos21. MDirect, 2007).

Description of Services Being in mind the nature of the available services, we could distinguish two types: services of data available through browsing (Figure 12, Figure 14) and services available through messaging (Figure 10, Figure 11). The most important browsing services are: •

Access to the catalog to query the resources/books (Figure 15); it is possible carry out queries by different fields (author, title, isbn, …) in the same way that for online or web catalogues but now using the mobile. If the resource is loaned or is not available, in this moment, through the same page, it is possible to carry out the reservation of the book. Query of personal reservations (Figure 16); with this option is possible to query reserved resources or books. The user has to authenticate with her name/surname and her number of user. Query of loaned books/resources (Figure 17a); with this service, the user could query

229

Success Cases for Mobile Devices in a Real University Scenario

Figure 15. Wap-page to query a book on the catalogue and page with a result

•

the books loaned; and also to renew the loan. As in the previous case, in order to query this information the user has to authenticate. Querying and updating the personal information (e-mail, address, phone number, …)(Figure 17b).

The services available by means messaging are: •

Services of alerts of loans out of date and alerts of available reservations; they are push services and the user receives in their mobile phone an alert SMS warning about

the event occurred. In these SMS, it is provided the type of service and the title and author of the book. Next messages are examples of this type of messages, mobiblio:Loan out-of-date:Title:Operating Sysem. Author: Stallings. mobiblio:Available Reservation: Structure Data in C. Author: Joyanes. •

Services of querying of loans, reserves of books, and renewal of loans; these services are based on pull-push technology and with them the user through a SMS makes a request of information or asks for an ac-

Figure 16. Wap-Page to authenticate a user and Wap-Page with a result of a query about the reservations of a user

230

Success Cases for Mobile Devices in a Real University Scenario

Figure 17. Diplay with the results of a query about loaned books (a) and diplay to update the personal information of a user (b)

tion to the system. The system provides to the user the information requested through a SMS (loaned resources or available reserves) or the result of carrying out the action (if it was possible or not the renewal). The SMS to make a request to short number could be like some of the next,

mobiblio loan mobiblio reserv mobiblio renewal