Pervasive 2004

NEW: Pervasive 2005

Home
Message from the GC
Conference Committee
Call for Papers
How to present
Forum
Gallery
Submission Categories
Doctoral Colloquium
Hot Spots
Tutorials
Papers
Tech-Notes
Videos
Workshops
Submission
Advance Program
Print Program (PDF)
Overview
Keynotes
Doctoral Colloquium
Workshop
Tutorials
Papers & Technotes
Hot Spot Papers
Video Papers
Demonstrations
Conference Site
Registration &
Accomodation

Sponsors
About Linz
About Vienna
Contact Us
Reviewers
PC Members
Artwork














Subscribe to the
PERVASIVE mailing list:


 

Location and Travel Info

The doctoral colloquium will be held at the
University of Linz
Uni Center / Mensengebäude Altenberger Strasse 69
A-4040 Linz.
(see also an area map showing the location of the Uni Center, Hotel Sommerhaus, and the tram stop)

Linz has an International Airport but can also easily reached by car and train from Vienna.

Schedule of Talks

  Sunday April 18, Linz
  Room G
10:45-11:00 Welcome and Opening
A. Ferscha, G. Kotsis
  Session A1
Room D

Advisors (tentative list):
Gabriele Kotsis, Max Mühlhäuser, Wieland Schwinger

Session B1
Room H

Advisors (tentative list):
Alois Ferscha, Ismail Ibrahim Khalil, Tom Pfeifer, Albrecht Schmidt

11:00-12:30
12:30-14:00 Lunch Break
  Session A2
Room D

Advisors (tentative list):
Michael Beigl, Alois Ferscha, Gabriele Kotsis, Wieland Schwinger

Session B2
Room H

Advisors (tentative list):
Max Mühlhäuser, Ismail Ibrahim Khalil, Tom Pfeifer, Albrecht Schmidt

14:00-15:30
15:30-16:00 Coffee Break
  Session A3
Room D

Advisors (tentative list):
Michael Beigl, Gabriele Kotsis, Tom Pfeifer, Jari Porras, Wieland Schwinger

Session B3
Room H

Advisors (tentative list):
Ismail Ibrahim Khalil, Max Mühlhäuser, Albrecht Schmidt

16:00-17:30
18:30-19:30 Guided Tour at AEC, Museum of the Future
19:30-22:30 Dinner at Lentos, Museum of Art
  Monday April 19
  Session A4
Room D

Advisors (tentative list):
Michael Beigl, Gabriele Kotsis, Friedemann Mattern, Wieland Schwinger

Session B4
Room H

Advisors (tentative list):
Ismail Ibrahim Khalil, Tom Pfeifer, Albrecht Schmidt

09:00-10:30
10:30-11:00 Coffee Break
  Session A5
Room D

Advisors (tentative list):
Michael Beigl, Gabriele Kotsis, Friedemann Mattern

Session B5
Room H

Advisors (tentative list):
Alois Ferscha, Tom Pfeifer, Albrecht Schmidt, Wieland Schwinger

11:00-12:30
12:30-14:00 Lunch Break
  Session A6
Room D

Advisors (tentative list):
Michael Beigl, Friedemann Mattern, Ismail Ibrahim Khalil

Session B6
Room H

Advisors (tentative list):
Alois Ferscha, Gabriele Kotsis, Albrecht Schmidt, Wieland Schwinger

14:00-15:30
15:30-16:00 Coffee Break
  Session A7
Room D

Advisors (tentative list):
Ismail Ibrahim Khalil, Friedemann Mattern, Guangyou Xu

Session B7
Room H

Advisors (tentative list):
Michael Beigl, Gabriele Kotsis, Albrecht Schmidt, Wieland Schwinger

16:00-17:30

 

Doctoral Colloquium Papers

  System Support For Ubiquitous Computing
   Erwin Aitenbichler, TU Darmstadt, Germany  
  A predominant issue in a ubiquitous computing world is that of future personal devices and how they communicate with their environment. This paper discusses smart digital identities that represent user's in the digital world and are able to perform operations such as authenticating the user to other parties. Furthermore, the environment with its many different devices, operating systems and network technologies, poses new challenges to communication middleware.
  Mobile Usability
   Thomas Grill, JKU Linz, Austria  
  Mobile telecooperative environments can be defined as the combination of mobility and the demands of a telecooperative environment. I will argue that the interaction design process for this kind of application areas must be different from the traditional desktop computing paradigm. Resultant a need for an adopted software engineering process for these environments is identified. As the standard engineering processes and methods could be applied in a broad way to the engineering of hardware and software within a mobile telecooperative environment, I conclude that this is applicable also to the area of usability engineering. Considering the restrictions given by such an environment the process has to be adopted. Based on a standard lifecycle approach I try to find a way of defining a process for a usability-engineering task for a mobile telecooperative environment. Concerning pervasive appliances existing interaction paradigmas can be varied, modified and enhanced. Also totally new paradigmas will have to be found and applied.
  Generating Adaptive User Interfaces Using Meta Description Languages For Device Independent Access
   Franz Gruber, SCCH, Austria  
  In this thesis an approach for generating device independent user interface descriptions and device dependant representation for mobile devices is presented. The user interface descriptions have to be adopted on device specifics therefore we call the user interface to be “adaptive”. To find out the factors where the description has to be adapted to, we use the Context information of the device. For mobile applications not only the mobile device itself has context information also when existing, a server application influences the client and the user interface. Therefore also the server context information can change the user interface adoption process to a specific device. Not the whole context information of client and server is required for generation of the user interface description and device adoption is needed, so the first key task is to identify and delimit the part of context information which is relevant for the user interface representation. A rough conceptual approach is presented in this paper, too. The second task is then to integrate the “user interface context model” to a user interface description language, not to forget the task of finding a suitable user interface description language for all devices. This work will then form the basis for intelligent user interfaces reacting on events of the execution environment.
  Designing Context-Aware Mobile Applications - User Perceptions For Usable UI’s
   Jonna Häkkilä, Finland  
  During recent years, the development of context-aware devices has gained increasing amount of attention. Mobile handheld devices form a potential group for context-aware application, since they are used in various different kinds of situations where prioritized applications and needs can change dynamically. Context-aware applications has to meet users’ demands in efficiency and usability and gain acceptance in privacy issues, which set challenges for designers. This study describes the experiences of research on usability and acceptability issues of context-aware mobile handheld devices.
  Personalization Attribute Management
   Kari Heikkinen, Finland  
  In this doctoral thesis a personalization management approach is presented. The thesis applies different personalization attributes on available pervasive applications and services within service domains. These attributes are the building blocks in the personalized provisioning of services and applications on the pervasive networks. The attributes are put on a layered model for management purposes; so that the layers ease the management of different attributes and the user can get measures for e.g. services in the domain. This approach for personalization management is to be applied in the enduser terminal, where the personal information is located. The personal attributes are exchanged with the service domain in the manner user has set. This attribute exchange is carried out between the personalization reference points of the end-user and the service domain. This thesis purposes to enhance the level of acceptability of technology and better user experience of services through the use of multidisciplinary research methodologies.
  Bringing Context Awareness Together
   Manasawee Kaenampornpan, University of Bath, UK  
  This paper reviews different context classification systems that have been used to define types of context. Although existing classification systems cover various types of context, in the development of context aware applications, only a few types of context have been used. This work aims to build a context classification model based on Activity Theory that brings together various research in context awareness. This provides a basis both for dialogue amongst context awareness researchers and for the implementation of a context awareness architecture.
  Self-Organizing And Scalable Service Discovery In Mobile Ad-Hoc Networks
   Michael Kreutzer, Universitaet Freiburg, Germany  
  Using service discovery as a key middleware component has the potential to hide the dynamic properties of mobile ad-hoc networks (MANETs) from the application layer. However, a service discovery approach should be optimized for the use in MANETs to manage both of the challenges: the constantly changing topology caused by the mobility of the nodes and the constantly changing size of the network caused by devices entering and leaving the network. This paper claims that a self-organizing and scalable service discovery approach called HYDRA provides a way to enable reliable access to basic services on the middleware layer. HYDRA makes use of the assumption that most of the required services are in the local surrounding of the service seeker. This assumption applies for the regarded scenario of mobile hospitals. HYDRA's distribution of the service discovery components is based on local views and local decisions and automatically adapts to the current size and topology of the MANET. The algorithm has been implemented in a test environment. In this experiment, the desired basic functionality of the algorithm has been reached and the paradigm of self-organization was fullled. The next step is the integration of HYDRA in the Service Location Protocol Version 2 (SLPv2) and a detailed analysis using a network simulator.
  Reachability Approach For Mobile Service Provision
   Reinhard Kronsteiner, JKU Linz, Austria  
  The use of wireless communication technology enhances the potential of mobile computing. Penetration with web-technology led to information-overload. Following the trend of ubiquitous computing and always-on connectivity for mobile services it will come to a non-manageable service-overload. Although we are permanently in coverage of many services, only a subset of them is relevant to us in our actual context. Limitation via restricted reachability to offered services is a way to keep the service landscape manageable for the user. This position paper covers a primary part of my PhD thesis. It introduces a notion to classify services out of their reachability and offers workarounds to service-filtering.
  A Logic-Based Framework For Qualitative Spatial Reasoning In Mobile GIS Environment With Applications To Navigation
   Mohammad Reza Malek, Technical University Vienna, Austria  
  Within the framework of this research the challenges raised by the limited resources on the mobile GIS environment will be addressed. In this research we propose a theoretical framework based on space and time partitioning in order to provide a paradigm that treats moving objects in mobile GIS environment. The idea is, to divide space-time into small parts and solve the problems recursively and the combination of the solutions solves the original problem. We motivate the use of influenceability relation as primary relation and show how a logic-based framework for representing and reasoning about qualitative spatial relationship over mobile agents can be defined. From application point of view, we concentrate here on finding a near optimal collision-free path because of its importance in robot motion planning, intelligent transportation system (ITS), and any mobile autonomous navigation systems.
  Understanding Sharing In Peer-To-Peer Networks
   Ronald Mannak, TU Delft, The Netherlands  
  Future mobile ad-hoc p2p (peer-to-peer) networks can provide reliable and robust personal and inter-personal connections between mobile devices in pervasive computing environments. Just as in current file-sharing p2p software, a mobile ad-hoc p2p network is dependent upon the willingness of users to share resources. Despite the large percentage of today’s file-sharing p2p users who do not share, there are those who do share. The aim of the current research is to understand factors that motivate users to share, with the goal of designing systems that can stimulate those who may otherwise not share. An increase in the number of users sharing resources will in turn lead to improved file sharing. Results of an informal exploratory study yielded four motivating factors for sharing resources, namely: Responsibility, Social visibility, Utility and Materialism. These factors are expected to play an even more important motivational role when users share files in mobile ad-hoc p2p networks, given the spontaneous nature of ad-hoc communities.
  An Architecture For Context Prediction
   Rene Mayrhofer, JKU Linz, Austria  
  Today’s information appliances appear very powerful, featuring on-device storage and processing power, communication technology and supporting many different applications. Context awareness is currently considered as one of the key issues for future device generations, with context prediction being the next step in research. The goal is not only to recognize the current context of an information appliance or its user, but also to predict the future context and thus enable the device to become proactive. In this paper, an approach to recognize and predict high level context information from low level sensor data is presented. Targeting a wide range of platforms, this approach has also been implemented in a software framework for on-line, un-supervised context prediction.
  Context Management For Heterogeneous Administrative Domains
   Filipe Meneses, University of Minho, Portugal  
  When accessing user context, context-aware applications often interact directly with sensors or have to deal with specific space representations. This work addresses context representation and management for mobile users. It proposes a generic solution based on a Context Container where the user context can be represented by an unlimited number of dimensions. The proposed solution is based on a Context Manager that integrates the raw data acquired by sensors and enriches user context with new calculated and estimated dimensions.
  Adaptive User Support In Agent Based Dynamic Environments
   Jan Misker, TU Delft, The Netherlands  
  In this paper the concept of an ad hoc agent environment as a way for users to interact with an ambient intelligent environment is introduced. Utilizing agents, users can interact on a level that best suits their needs and capabilities, leaving tedious chores to the agents. The user interacts with this environment as a whole, instead of interacting with individual applications on individual devices. Devices and services in the environment have to be more or less independent, which fits well with the notion that agents are autonomous. Intelligence enables the agents to learn about their user and adept the environment in such a way that the users cognitive resources are not overly addressed. A research application that demonstrates some of the issues involved when employing agents has been developed. The application shows that there is a tension between the amount of control a user can have over an agent and the autonomy of the agent.
  Towards An Architecture And New Modeling Appraoch For Context-Based Security In Pervasive Environments
   Ghita Kouadri Mostefaoui, University of Fribourg, Switzerland  
  An infrastructure for context-based security managements in pervasive environments is presented. The core components of the system are described with a focus on the followed approach to modeling context-based security policies.
  Position-Based Composition of Video Scene From Distributed Sensors
   Phuong-Hoang Nguyen, France Telecom R&D Dih/Ocf, France  
  We address the management of distributed video sensors and their automated configuration for "follow me" interpersonal visual communication. A model for description of video sensor device metadata is specified using a variant of a generic XML-based sensor modelling language. These metadata may integrate an X3D-based description of the sensors' relative positioning information. This is used to compose a 3D video scene. Navigation in this 3D video scene is based upon the positioning information of the target persons, making it possible to automatically follow them through the sensor network.
  Reconfigurable Software Architecture For Communication Layers In Pervasive Computing Systems
   Mahdi Niamanesh, Sharif University of Technology, Iran  
  Due to existence of different communication standards and upcoming ones, it is necessary to have suitable devices to support standards. Using current technology, every standard almost has a special supporting device and is somewhat implemented in hardware. As a result, a pervasive computing network needs huge amount of hardware devices, which upgrading of such a network can be very expensive or even infeasible. To confront the problem, reprogrammable and reconfigurable standards on a (general-purpose) processor can be proposed using the concepts of software radio and active networks. The main question, which is addressed in my thesis, is how to implement the communication layer of pervasive computing systems. The communication layer, which is placed on top of the nearly general-purpose physical career, includes all radio processing, data linking and networking layers. Ideally, the physical layer in such systems does not include any protocol- (standard-) specific processing and therefore it is a common platform to present common services for higher layers. A suitable framework, which includes architectural reconfiguration, control and management, is necessary to handle reconfiguration operation and conflict resolutions.
  Automatically Generating User Interfaces For Appliances
   Jeffrey Nichols, Carnegie Meelon University, USA  
  I am exploring the use of automatic user interface generation for improving the experiences that users have with computerized appliances in their pervasive computing environment. I am building the personal universal controller (PUC) system, which will automatically generate graphical and speech user interfaces for an appliance on intermediary “user interface devices” such as personal digital assistants (PDAs) or mobile phones. Each interface is generated from an abstract specification of the appliance, written in a language I am designing, that describes the complete functionality of the appliance. The PUC system will also be able to generate a single combined user interface for multiple connected appliances, such as a home theater. Additionally, I am exploring how previous interface design decisions can be incorporated into future interface designs to achieve consistency for the user. I intend to evaluate the completed system by conducting user studies that compare the au! tomatically generated appliance interfaces to the functionally identical manufacturers’ user interfaces.
  Enabling Contextual Services In Wireless Context Area Networks
   Mario Pichler, SCCH, Austria  
  This paper represents a work -in-progress description of the WCAN (Wireless Context Area Networks) project at Software Competence Center Hagenberg (SCCH). It describes the idea of subdividing the real world into areas that represent a specific context, like railway stations, airports, sport stadiums, concert halls etc., and to provide mobile people therein with services that are context specific. The suitability of the Open Services Gateway initiative (OSGi) Service Platform and Multi-User Publishing Environment (MUPE) Application Platform for the realization of contextual services is investigated. The motivation for this dissertation project, its current status, and planned further work are illustrated.
  Supporting Localized Activities In Ubiquitous Computing Environments
   Helder Pinto, University of Minho, Portugal  
  A localized activity, i.e., an activity that is realized within a single location, requires the combination of artefacts either available in the environment or carried by the user. This work explores how ubiquitous computing environments can provide a user-centered support to localized activities by integrating the ubiquitous computing infrastructure with the user personal information space and resources. This implies several integration problems, namely the integration with personal information spaces and with user resources (e.g., information services and physical devices). This work identifies the main integration issues and explores technological and architectural solutions for the problem, aiming to contribute with a comprehensive framework supporting integration between ubiquitous computing and user environments.
  Determination of Time And Location In Large-Scale Dynamic Networks of Tiny Sensors
   Kay Roemer, ETH Zürich, Switzerland  
  So-called sensor nodes combine sensors, processors, wireless communication capabilities, and autonomous power supply in a tiny device. Large-scale networks of these untethered devices can be deployed unobtrusively in the physical environment in order to monitor a wide variety of realworld phenomena with unprecedented quality and scale. A fundamental service in sensor networks is the determination of time and location of events in the real world. This task is complicated by various challenging characteristics of sensor networks, such as their large scale, high network dynamics, restricted resources, and restricted energy. We develop new approaches for determination of time and location under these constraints, and devise design principles based on our experience. We illustrate the practical feasibility of our approaches by a concrete application.
  Virtual Distances Used for Optimization of Applications in the Pervasive Computing Domain
   Kari S. F. Schougaard, University of Aarhus, Denmark  
  This paper presents the notion of virtual distances - communication proximity - to describe the quality of a connection between two devices. We use virtual distances as the basis of optimiza- tions performed by a virtual machine where a part of an application can be moved to another device if this gives a performance improvement. As best-e rt connections are used the quality of service cannot be negotiated, but the virtual distances are monitored by the virtual machine.
  Pervasive Information Retrieval
   Zia Syed, Robert Gordon University, UK  
  With the emergence of pervasive computing environments, mobile information retrieval is becoming an exciting new domain that challenges information scientists. In the context of mobile user, we argue that most of the task-related information that a mobile user needs, possess strong correlation with the spatio-temporal aspects of the information and its sources. However, often such information is present within the very surrounding of the user, but currently inaccessible. We suggest that peer-to-peer information sharing mechanism can help us tap this unheeded information. Effective retrieval of contextually-related information is critical in a mobile scenario, and we suggest exploitation of contextual attributes of both the mobile users and the information itself so as to enhance the performance of the overall information seeking process.
  Event Dissemination Service For Pervasive Computing
   Sasu Tarkoma, Helsinki Institute For Information Technology, Finland  
  Event-based computing is a generic enabler for the next generation mobile services and applications that need to meet user requirements irrespective of time and location. The event paradigm and publish/subscribe systems allow clients to asynchronously receive information that matches their interests. The requirements of pervasive computing present new challenges pertaining to event delivery that need to be solved. The research question addressed in this paper is how to meet the requirements of the pervasive environment. We outline an architecture for mobile computing that addresses two key requirements: terminal mobility and user mobility. The system consists of access servers, event channels and a mechanism for locating event channels. The architecture uses filter merging for supporting high accuracy in event delivery, reducing communication cost, and improving event processing on terminals and servers. Experimental results based on the merging system are also examined.
  Interoperative Systems For Replenishment
   Sébastien Truchat, University of Erlangen-Nuremberg, Germany  
  Most of the mobile services currently offered are built on the typical internet architecture where the mobile device depends on a reliable communication to the server on which the application runs. Through the example of the replenishment system HORN, another type of architecture based on short range wireless communication and local applications will be introduced. An analysis of the problems and needs related to the development of mobile services in this context will lead to a strategy to enhance and reuse existing infrastructures for new services and other platforms.
  Sitmod: A Tool For Modeling And Communicating Situations
   Martijn Vastenburg, TU Delft, The Netherlands  
  Home environments will change in the near future due to use of awareness technology. Future homes will be able to sense situations and react appropriately. User profiles can be used to personalize the aware home environment. This study researches the use of situated user profiles, in which the elements of the profile are linked to situations. First, a formal model for home situations is constructed based on the results of qualitative user studies. In a later stage of the study, the role of the profiler as a mediator will be investigated.
  Dichotomy In The Plasticity Process: Architectural Framework Proposal
   Montserrat Sendín, Escola Universitària Politècnica, Spain  
  Everybody knows that mobile computing provides us with tremendous versatility, making it possible to extend interaction to a wide range of increasingly divergent contexts in which multiple variable parameters come together. However, the more this versatility increases the more challenging the goals are becoming. On the one hand, the complexity in the design of User Interfaces (henceforth UIs) increases, and on the other hand the need for dynamic adaptation becomes more noticeable. In this thesis I analyse the problem of plasticity of UIs and I discern between the capacity required to automatically generate a UI for a concrete contextual situation, starting from an abstract specification –generic UI -, and the capacity of incremental adaptation that the specific UI obtained should adopt in real time as the user goes across new contexts of use . Therefore I distinguish and define two types of plasticity with their respective goals that I name explicit plasticity and implicit plasticity respectively. Furthermore, I propose a client/server-based architectural framework to accommodate this continuous and diverse variability accompanying the system throughout its period of utilisation, which takes into account the two identified goals that I am looking to solve separately but in co-ordination.
  Mobile Agents With Intrusion Detection During Transfer For Pervasive Computing
   Degan Zhang, Tsinghua University, Beijing, P.R. China  
  Based on Pervasive Computing paradigm, we have developed a set of key technologies for Smart Space, such as Smart Meeting Room, with multiple mobile agents for intrusion detection during seamless transfer in this paper, and discussed the advantage of mobile agent, the type and working mechanism of mobile agent, the scenario of smart space in the era of pervasive computing with moblie agents. Based on multiple mobile agents, a number of problems can be solved in the process of sealess mobility, especially, intrusion detection. The developed technologies have been success-fully integrated and demonstrated in our prototype system. The efficiency of our researches has been tested by the demo.
  An Architecture For Building Customizable Context-Aware Applications By End-Users
   Tao Zhang, TU Munich, Germany  
  The application logic of many current context-aware systems, that is, which action to perform in response to sensor events, is hard coded and not modifiable by the user. Hard-coded application logic makes the system difficult to personalize, while the invisibility and the unmodifiability of the application logic make it difficult for the user to get a common understanding about the system behavior. Users may stop using the system since they have the feeling of losing the control over it. In this paper, we bring together two research fields, context-aware computing and rule-based systems, and propose an architecture which enables end-users to customize and control context-aware applications. We report on a usability case study conducted in courses of teaching activities. We conclude this paper with a status report and an outlook into the future work.
Pervasive 2004      April 18-23      Linz / Vienna, Austria      Back to Top