Invited Talk

Keynote Talk

PERFORMANCE ISSUES IN
MOBILE COMMERCE APPLICATIONS

Gabriele Kotsis

Institut für Telekooperation
Johannes Kepler University Linz
Altenberger Strasse 69, 4040 Linz, Austria
gabriele.kotsis@jku.ac.at

Abstract
Recent years have seen a sustained growth of interest in mobile computing and communications. Indicators are the rapidly increasing penetration of the cellular phone market in Europe, or the mobile computing market growing nearly twice as fast as the desktop market. In addition, technologic advancements have significantly enhanced the usability of mobile communication and computer devices. From the first CT1 cordless telephones to todays Iridium mobile phones and laptops/PDAs with wireless Internet connection, mobile tools and utilities have made the life of many people at work and at home much easier and more comfortable. We can conclude that mobility and wireless connectivity are expected to play a dominant role in the future in all branches of economy. This is also motivated by the large number of potential users (a US study reports of one in six workers spending at least 20 percent of their time away from their primary workplace, similar trends are observed in Europe). The addition of mobility to data communications systems has not only the potential to put the vision of "being always on" into practice, but has also enabled new generation of services, e.g., location-based services. In this talk, we will discuss the fundamental performance issues related to the development and deployment of mobile commerce applications and services.
Mobile applications are based on a computational paradigm, which is quite different from the traditional model, in which programs are executed on a stationary single computer. In mobile computing, processes may migrate (with users) according to the tasks they perform, providing the user with his or her particular work environment wherever he or she is. To accomplish this goal of ubiquitous access, key requirements are platform independence but also automatic adaptation of applications to (1) the processing capabilities that the current execution platform is able to offer and (2) the connectivity that is currently provided by the network. Mobile services and applications differ with respect to the quality of service delivered (in terms of reliability and performance) and the degree of mobility they support, ranging from stationary, to walking, to even faster movements in cars, trains, or airplanes. A particular challenge is imposed by (interactive) multimedia applications, which are characterized by high QoS demands. New methods and techniques for characterizing the workload and for QoS modeling are needed to adequately capture the characteristics of mobile commerce applications and services.
A fundamental necessity for mobile information delivery is to understand the behavior and needs of the users, i.e. of the people. Recent research issues include efficient mechanisms for the prediction of user behavior (e.g. location of users in cellular systems) in order to allow for proactive management of the underlying networks. Besides this quantitative evaluation user behavior can also be studied from a quantitative point of view (how well is the user able to do her or his job, what is the level of user satisfaction, etc.) to provide information to other services, which can adapt accordingly. This kind of adaptation may for example include changes in the user interface, but also chances in the type of information transmitted to the user.
From a telecommunications infrastructure point of view, the key enabling technology for mobility are wireless networks and mobile computing/communication devices, including smart phones, PDAs, or (Ultra)portables. Wireless technologies are deployed in global and wide area networks, (GSM, GPRS and future UMTS, wireless broadband networks, GEO and LEO satellite systems), in local area networks (WLAN, mobile IP), but also in even smaller regional units such as a campus or a room (Bluetooth). Research on wireless networking technologies is mainly be driven by the quality of service requirements of distributed (multimedia) applications with respect to the availability of bandwidth as well as performance, reliability, and security of access.
Being provocative, one might state, that the situation that application developers are facing nowadays in mobile computing is similar to the early days of mainframe computing. Comparatively "dumb" clients with restricted graphical capabilities are connected to remote servers over limited bandwidth. Although significant improvements have been achieved increasing the capabilities of networks and devices, there will always be a plethora of networks and devices and the challenge is to provide a seamlessly integrated access as well as adaptability to devices in application development making utmost use of the available resources.
We conclude, that not only in development but also from a performance point of view, the integration of multimodality and multimedia, the changes in user behavior, the specific characteristics of the underlying infrastructure (networks and devices) as well as the adaptability of applications and services must be considered in any analysis and evaluation. In the talk, we will present a case study of a transponder-based mobile service and discuss the problems and difficulties in performance prediction and tuning.

Biography

Univ. Prof. Mag. Dr. Gabriele Kotsis – Prof. Kotsis received her master degree in 1991 (honoured with the Award of the Austrian Computer Society), her PhD in 1995 (honoured with the Heinz-Zemanek Preis) and
the venia docendi in 2000 (computer science, from the University of Vienna). She was working as a researcher and teacher at the University of Vienna (1991-2001), at the Vienna University for Economics and Business Administration (2001) and at the Copenhagen Business School (2002). Since December 2002 she is holding a full professor position at the Telecooperation Department at the Johannes Kepler University Linz. Her research interests include performance management of computer systems and networks, workgroup computing, mobile and Internet computing, telemedia and telecooperation. She has experience in national and international research project in those areas, including for example the EU-funded international BISANTE project on network traffic modelling and simulation, where she was technical leader, or the EMMUS project on Multimedia Usability where she was project coordinator. Gabriele is author of numerous publications in international conferences and journals and is co-editor of several books. She is member of IEEE and ACM and acting president of the AustrianComputer Society. She is actively participating in the organization of international conferences.

Invited Talk 1

Processing Streams of Spatial Data
Stephane Bressan, Tok Wee Hyong

Computer Science Department
The School of Computing (SoC)
National University of Singapore
Singapore

Abstract
We propose and study sequential non-blocking algorithms for the processing of spatial joins on continuous data streams with unpredictable arrival rates. Given two incoming streams of spatial data represented by their bounding boxes, the algorithms immediately and continuously compute and output the pairs of data from each stream whose bounding boxes intersect.
The different algorithms we propose take advantage of different possible characteristics of the data such as clustering of the arrivals to build indexes or summaries to accelerate the production of results. We comparatively analyze the performance of the proposed algorithms using several synthetic and real-life datasets.

Biography

Dr. Stéphane Bressan is a senior lecturer in the Computer Science department of the School of omputing (SoC) of the National University of Singapore and Adjunct Associate Professor at the Malaysian University of Science and Technology (MUST). He joined the National University of Singapore in 1998. He is the coordinator of the Electronic Commerce Laboratory at SoC. He joined MUST in 2004. He graduated in 1987 with a degree in Computer Science, Electronics and Process Automation from the Ecole Universitaire D'Ingénieurs de Lille (France) and received his Ph.D. in Computer Science in 1992 from the
Laboratoire D'informatique Fondamentale of the University of Lille. In 1990, he joined the European Computer-industry Research Centre (ECRC) of Bull, ICL, and Siemens in Munich (Germany). In 1994, he was appointed site leader of the Database Platform project and principal investigator and work-package manager for the European IDEA ESPRIT project on Intelligent Databases. From 1996 to 1998, he was a research associate at the Sloan School of Management of the Massachusetts Institute of Technology (MIT) working on the Intelligent Integration of Information.

Stéphane Bressan's work has been published and presented at various occasions including the 1995 G7 summit on the Information Society, the 1995, 1997 and 2000 ACM-SIGMOD conference on the management of data, the 1996 and 2000 conference on Extending DataBase Technology, or the 2002
ICDE international conference on data engineering. He is the author or co-author of more than 100 conference, workshop, and journal articles, and books and book chapters. He is a member of the program and editorial committees of several conferences and journals including the 2002 ACM SIGMOD conference on the management of data, the 2003 ICDT international conference on database theory and the 2004 VLDB international conference on very large databases. Stéphane Bressan's domain of research is the integration and the management of disparate information, i.e. the integration and management
of multi-modal and multimedia information from distributed, heterogeneous, and autonomous sources. His research includes core database research, modeling, extraction, integration and processing of disparate
information, peer to peer systems, and information indexing, retrieval, and extraction, as well as applications in domains such as Geographical Information Systems and Computational Linguistics.

Invited Talk 2
Pending confirmation