I-centric communications - IEEE Xplore

WIRELESS WORLD RESEARCH FORUM

I-centric Communications: Personalization, Ambient Awareness, and Adaptability for Future Mobile Services Stefan Arbanowski, Pieter Ballon, Klaus David, Olaf Droegehorn, Henk Eertink, Wolfgang Kellerer, Herma van Kranenburg, Kimmo Raatikainen, and Radu Popescu-Zeletin

ABSTRACT The acceptance of next-generation mobile communication systems depends to a large extent on the services and applications that can be offered to customers. Tailoring the services to actual user needs is here considered to be crucial for the success of future wireless technology. The individual user, “I,” has to be put in the center of service provisioning. In this article we report the work developed by the Working Group 2 of the Wireless World Research Forum on a service infrastructure framework for the future wireless world. Major service capabilities such as personalization, ambient awareness, and adaptability are described along with a reference model focusing on I-centric communication. Since September 2001, the Wireless Work Research Forum (WWRF) Working Group 2 (WG2) has been active in the area of service architectures for future wireless systems. The vision of I-centric communication has been developed putting the individual user (“I”) in the center of service provisioning rather than offering inflexible services that are unaware of actual customer needs or situations. I-centric systems relieve the user of technology-related decisions, such as network selection, supporting his or her personal goals. Thus, all tasks a communication system has to perform must consider this I-centric paradigm. The objective of this research is to develop a communication system service infrastructure model that will include each individual in his or her environment with his or her preferences, and adapt services to different situations and resources in real time. Figure 1 shows the WWRF WG2 reference model for I-centric communications, addressing the individual user interacting in his or her personal communication space. It follows a topdown approach starting with the user in his or her individual communication space, defined by related contexts and objects. From a communication system point of view, I-centric services in such a communication space support three fea-

IEEE Communications Magazine • September 2004

tures: ambient awareness, personalization, and adaptability; these are discussed in the following. The service platform for I-centric communications is responsible for shaping the communication system, based on individual communication spaces, contexts, preferences, and ambient information. Preferences will be provided by personalization, whereas ambient information has to be provided by ambient awareness. The IP-based communication subsystem is responsible for providing interactivity between different objects in the communication spaces. It provides features such as call control, session control, and mobility management. IP communication is seen as the common denominator to harmonize heterogeneous network infrastructures. The wired or wireless networks layer implements all aspects of the physical connection(s) between different endpoints. Due to the hierarchical structure of the reference model, a connection in the IP-based communication subsystem might use different physical connections in the underlying networks. Devices and communication end systems provide the physical end system infrastructure that supports all other layers. The reference model does not address how the functionalities on the different layers are distributed between network entities and terminals. In this way, it fits both centralized service architectures (e.g., derived from the intelligent network) and purely decentralized architectures (e.g., peer-to-peer). The main features of I-centric communications (ambient awareness, personalization, and adaptability) affect all layers. Therefore, supporting functions have to be provided as a vertical solution. The reference model introduces the concept of a generic service element that implements common functionalities on all layers. A generic service element can be seen as a toolbox from which complex services can be assembled and executed dynamically. The vertical approach introduces I-centricity on all layers. Generic service elements such as service discovery or environment monitoring provide cross-system knowledge of the context. Service creation or

0163-6804/04/$20.00 © 2004 IEEE

63

Capturing value Context

from information goods, widely

Ambient awareness

perceived as being

seem to increase

Business model

Adaptation Conflict resolution

the user him/herself

Individual communication space

Service deployment

tailored to specific needs or initiated by

Environment monitoring

wireless services. However, services

Service creation

business model for

Personalization Service discovery

challenge for any

Service control

Service bundling

public goods, will remain an important

Objects

Application support layer

Generic service elements for all layers

Service platform

Service execution layer Service support layer Network control and management layer

IP-based communication subsystem

IP transport layer

their value to the customer.

Networks

Wired or wireless networks

Terminals

Devices and communication end systems

■ Figure 1. The reference model for I-centric communications. conflict resolution introduces programmability and reflection to the system. Complementing these technical issues, the business model for I-centric communication identifies the relationships and information flows between all active stakeholders and their roles within an I-centric system. The business model helps to identify reference points between all involved stakeholders and supports the assessment of the applicability of I-centric services to different business domains. The following sections introduce the results obtained in WWRF WG2. They explain how personalization, ambient awareness, and adaptability can be provided in future mobile communication systems. The concepts described represent a common understanding between academia and industry.

BUSINESS MODELS The vision for I-centric communications has two major implications for future business models. First, the influence of business considerations on the functional architecture increases. In telecommunications, the technological infrastructure has usually been set up before the actual business model. As both liberalization and convergence continue to shake up the structure of the telecommunications sector, organizational and financial aspects are increasingly becoming leading factors in new telecommunication systems development. This implies that rather than resulting from an R&D trajectory toward a coherent technological system, an I-centric communications system will be made up of various technologies configured around specific user requirements as identified in the market [1]. A lot of speculation has gone on about the

64

precise configuration of these technologies in the future wireless value network. In general, it can be argued that business models for mobile services have traditionally been characterized by an important dependency on the underlying network infrastructure, resulting in a rather closed model with a central “gatekeeping” role for the mobile network operator. This is slowly changing. Recent research reveals a number of decisive business factors that are starting to reshape the functional architecture: Due to increasingly uncoordinated and even divergent investment strategies of established operators as well as of new entrants, the emerging network environment will be characterized by strong heterogeneity. Interoperability will become a crucial factor determining the success of new wireless business models. Capturing value from information goods, widely perceived as public goods, will remain an important challenge for any business model for wireless services. However, services tailored to specific needs or initiated by the user him/herself seem to increase their value to the customer. The I-centric vision emphasizes this further by allowing users to take a central role in the production and distribution of new services. Business and market strategies of (coalitions of) firms in different information and communication technology (ICT)-related sectors will determine whether future wireless solutions will be network-centric or terminal-centric in terms of the distribution of intelligence. Until now, three different partly conflicting approaches can be observed: service-based (e.g., i-mode, Vodafone live), platform-based (e.g., MS Smartphone, Symbian Series 60), and protocol-based (e.g., SMS, MMS). Ad hoc networking may prove to be a disruptive technology in this evolution.


The vision for I-centric communications has introduced the

Reference point

concept of individual Info

Food

communication

News

spaces. An individual user is interacting with objects in his communication Administrative domain Stakeholder

Administrative domain

Administrative domain Stakeholder


tion between different individuals is done by sharing

■ Figure 2. Mapping objects to business relationships. Second, there will be increasing flexibility of roles and actors. This refers to the unbundling of functions and roles, and to their peer-to-peer characteristics. The I-centric vision implies the convergence of traditional telecommunication systems, Internet-based systems, and the emergence of new applications. The borders between traditional roles and administrative domains, such as network provider, content provider, service provider, and retailer, are blurring. An individual user may become a network provider (ad hoc networking), content provider (e.g., music sharing), service provider (peer-topeer), or even a retailer. Additionally, the roles may change in the same active context, implying a very flexible business model. This raises the question of how to integrate the individual user and the dynamic character due to changing business roles into a single generic business model. The vision for I-centric communications has introduced the concept of individual communication spaces. An individual user interacts with objects in his communication space. Communication between different individuals is done by sharing objects of their individual communication spaces. The particular physical resources used for a certain communication request are determined in the activation process of contexts and objects. These physical resources belong to administrative domains where certain roles are assigned. As users act in different contexts, the relationships to the administrative domains involved in a communication process have to be managed. This causes the dynamic assignment of roles and employment of different reference points from the viewpoint of an individual (Fig. 2). Reference points are interfaces describing the interactions taking place between the different roles. For example, the retailer reference point describes the relation between the consumer and the retailer. This fact is further amplified as the environments themselves are highly dynamic, and characterized by ad hoc and peer-to-peer communication, which may take place without any centralized control. The dynamic relationships between individual users and objects require new functions such as online subscription, online accounting, micropay-

space. Communica-

objects of their ment, and federation between ad hoc communication environments. Emerging challenges include the temporal unavailability of objects and services, and questions related to who pays whom for providing or using any physical resource. Such issues are addressed by specifying reference points between all instances involved. At those reference points, instances support the functions requested above. For example, for a micropayment function between two objects beneath exchanged information, the relationships to other objects, which might act as certifying instances, such as billing or accounting servers, have to be specified at respective reference points.

individual communication spaces.

PERSONALIZATION The phenomenon of information flooding applies not only to the entire Internet, but also to individual Internet sites, persons, and the wireless world as such. It has been proven that users, especially of mobile devices, are not willing to deal with complex functionalities and graphic user interfaces (GUIs). Personalization aims to increase the usefulness and acceptance of digital information and applications, since the user can manage his/her own individual information and communication space so that he/she can select, configure, and arrange presented information individually. Here an important service achievement is that the customer recognizes his/her space with increased attendance, facilitating fast orientation and usage. Besides explicit personalization, there is personalization where the set of offered information is adapted to the user implicitly by learned schemes [1]. The basic element used for personalization is the context of a user. A context may consist of many aspects such as the user’s needs, preferences, history, and behavior; location-related aspects such as physical coordinates and velocity; and ambient conditions, technical aspects like bandwidth of the network and capabilities of the terminal, business rules that apply, and so on. Context information can thus be defined as any information that can be used to characterize the situation of an entity. Personalization is considered the key factor for

65

Uniform

Highly differentiated

Quantitative, few

Rule-based

CASE

Qualitative, complex

Key products attributes

Customer needs, product space

Endorsement

Collaborative filtering

■ Figure 3. Technologies to achieve personalization. success of mobile devices and services. Information and services become increasingly tailored to individual user preferences and characteristics. Preferably, the services should automatically and in real time adapt themselves to changes in context. As the mobile user moves around, the services have to deal with a very dynamic user environment. Furthermore, adaptation to the context should be hidden from the user, providing him/her with optimal experiences and additional value. On the other hand, the environment of the user can and should be influenced by the presence and activities of the user, and adjust itself accordingly. The primary goal behind personalization is to make usage easier and the perception of the communication space richer, and enable personalized filtering of the global communication space to each individual communication space. Profile information has to be adapted dynamically using automated learning functionality. Wireline and mobile telecommunications systems are converging toward an all-IP communication system. In this environment, an extended personalization concept is needed that enables value networks (e.g., value chains) of content providers, network providers, and service providers to offer personalized services to mobile users in a way that suits their individual needs at a specific place and time. Personalized services will provide, next to data and content, emotions and experiences to the user. Depending on different needs and properties, several technologies are used to achieve personalization. Figure 3 depicts some typical technologies to acquire user profiles used for different kinds of customer needs (ranging from uniform to high differentiation) and the properties of desired products. Rule-based systems could be used to learn preferences by the observation of user behavior according to predefined rules. Computer-assisted self explication (CASE) systems support the user in specifying a detailed profile online. In endorsement systems user feedback is exploited to learn preferences. Col-

66

laborative filtering is used to compare and combine preferences with groups of similar profiles. These techniques aim to adapt existing systems by filters or content adaptation. Upcoming developments need to go beyond those techniques and also take several other technologies, like ambient awareness, into account. Several standardization activities are going on in several bodies like the Internet Engineering Task Force (IETF), the World Wide Web Consortium (W3C), and the Open Mobile Alliance (OMA) to handle different kinds of profiles and contexts and their adaptation to users’ needs. This reflects the upcoming need for personalization of information, services, and their presentation to users. In conclusion, new requirements for information and service personalization arise, addressing the handling of an extendible set of user preferences, the definition of multiple user profiles associated with different service usage situations, support for service accessibility to user presence, and profiling of different networks and services themselves, to name only a few.

AMBIENT AWARENESS Ambient awareness uses a transection of context, focusing on the user’s information that is ambient in the sense of situation and environment. Ambient in I-centric systems refers to the situational context in which an individual user or actor is. Ambient awareness deals with sensing and exchanging the ambient information of a user in the human communication space. Three main parts are identified: spatial information (geographical data like location, orientation, speed, and acceleration), physiological information (depicting life conditions, e.g., blood pressure, heart rate, respiration rate, muscle activity, and tone of voice), and environmental information including temperature, air quality, and light or noise level [1]. The goal of ambient awareness is to acquire and utilize information about the situation of an actor to enable services in an active context (i.e., personalize and adapt services to a certain situation at a certain moment in time). In ambient awareness a key aspect is situation sensing in which a sensing device detects environmental states of different kinds and passes them on (to the context interpreter functionality). An ambient awareness support architecture must collect ambient information about all relevant entities in the user’s environment from diverse sources, process that data, and distribute the processed information to several different types of applications running on distinct mobile or fixed devices. The architecture should scale to a high number of information sources, applications, mobile devices, and networks. The exchange of contextual parameters must be supported between different network or administrative domains and heterogeneous devices. Hiding as many details as possible of context acquisition and exchange from applications and services matches the user’s desire (and fits well within the I-centric view). Thus, intelligent resource discovery (and sharing) is also necessary for the service architecture. Figure 4 sketches the “spider role” of the ambient awareness functionality between (heterogeneous) sensors in a real-world environ-


Sensory equipped Wi-Fi environments

Ad hoc networks

Restaurant services

Graphical Statistics: Port 4 File View Measurements

File\

Grap File\

GPS equipped body area networks

Grap File\

Grap

4\

4\ 4\

4\

Video application schedule Presence application Instant messaging

Sensory equipped home environment

Ubiquitous WAN

Ambient awareness

Anywhere, anytime

Graphical Statistics: Port 4 Graphical Statistics: Port Graphical Statistics: Port 4 Graphical Graphical Statistics: Port 4 Graphical Statistics: Port Graphical Statistics: Port 4 GraphicalGraphical St 4 Graphical Statistics: Port Graphical Statistics: Port 4 Graphical Graphical Statistics: Port 4 Graphical Statistics: Port Graphical Statistics: Port 4 Graphical

Sensory equipped corporate environment Location services

■ Figure 4. The spider role of ambient awareness.

ment, and tailored applications and services. In the sequel the main functionalities of an ambient awareness architecture are highlighted. In order to acquire ambient information, sensors or human–machine interfaces gather information about actors (humans) and their environmental states. Actors can also provide this information themselves. A special feature of ambient is position and location information. This part of the user context is mostly exploited in today’s mobile services. Advances in sensor technology (smaller and integrated on chip) and extension of sensor types (e.g., body sensors that are nonintrusive in user perception) are needed to increase the capabilities for adaptation of services to, and cooperation with, the environments of actors. Aggregation of the sensed values in relation to a known reference is needed for correct interpretation. Ambient information can be significantly enriched by integrating sensed information (e.g., coordinates) with situational models (e.g., topology information) in order to retrieve higher-level concepts (e.g., the person is in a particular room). Different parts of the actor’s situation and the context of the information may have different weights in determining the ambient of the actor. In addition, the order of processing may affect the resulting ambient awareness. Expressing the relative weights and ordering of processing can be done in a formal rules language or as ontology, making contextual processes, like database selection, transcoding, and annotation, part of the process of evaluation. Other key technologies needed for intelligent processing are reinforcement learning, incremental clustering, and rule-based deduction. Many ambient-aware applications make use of


sensory information obtained using diverse sensors, making cross-modality sensory modeling for ambient awareness a challenging research issue. The presence of multimodality and availability of relevant ambient information provide the need to introduce intelligent human–machine interface management, which enables intelligent I/O behavior adaptation. This means that the input/output behavior of a user end system (e.g., the modality) can change dynamically based on the actual context. For example, the user interface (UI) to an email system is typically text-based, whereas it would become speech-based if the user is located in a car. Research on appropriate models for structuring ambient-aware applications is also needed. The massive number of actors and objects in the wireless world can generate enormous amounts of information. This information cannot be processed efficiently without enormous computing power. Luckily, the ubiquitous availability of network connectivity and computing resources allows for building global software infrastructures for distributed computing and storage that can be used as a basis for intelligent processing of contextual information, and providing this information to applications and services. This global infrastructure consists of hardware and software entities that provide wellknown functionality and services. Examples include abstractions of devices such as mobile phones, PDAs, home appliances, and various software components. An interoperable middleware technology that enables uniform access to this software infrastructure in order to support easy and rapid service creation is necessary [2]. Standards to describe ambient information

67

App

App

App

Open APIs

Modeling services

Ontology service

Mobile distributed information base

Semantic matching engine

Perception service

Event notifications

Environment monitoring

UI UI support support

Generic service elements

Distributed application framework Execution environment

Mobile middleware

Internet protocol suite Operating system

■ Figure 5. The provision of open APIs. are a prerequisite of making the information available to applications. Only a few well defined mechanisms, frameworks, and standards exist that specify how contextual information can be made available to applications and how data is exchanged between protocol layers. For the exchange of location-based information between domains no open Internet standards exist; also, the problem of selectively transmitting, replicating, and/or analyzing data is not completely understood. Software support for context-aware applications is investigated by several approaches, such as by architectural (partial) solutions [4, 7], frameworks [8, 9], toolkits [3, 6], and service infrastructures [5, 10]. Standardization in this area is mainly performed within OMA, W3C, and Parlay, but focuses on specific aspects like location information (Parlay and OMA), privacy and terminal capabilities (W3C), or SyncML (for synchronizing information between devices). Ultimately, standards to exchange the situational circumstances of an individual and generic support for this exchange as well as standards to secure protection of sensitive data are needed.

ADAPTABILITY Within our work there is a wide consensus that beneath personalization and ambient awareness, adaptability will be one fundamental characteristic of services and applications for the wireless world of the future. By adaptation we mean the ability of services and applications to change their behavior when circumstances in the execution environment change. Based on the functional requirements for adaptive applications we have identified eight generic service elements (Fig. 5). They include environment monitoring, event notification, distributed application framework, perception service, modeling services, mobile distributed information base, ontology service, and semantic matching engine. We have not identified service discovery and auto-configuration as generic service elements. Instead, they are included in the distributed application framework due to their crucial importance [1].

68

The basic principle of adaptability — the behavior of an application changes when the circumstances change — requires that the system detects changes and notifies about them. Therefore, the generic service elements must include environment monitoring and event notification. Adaptive applications need to change their behavior. In some cases this is achieved by changing an algorithm internal to the application. However, in many cases an alternative approach is much more plausible through replacing some components of the application or relocating them into other network elements or service nodes. Therefore, the generic service elements must include a distributed application framework that enables seamless and flexible replacement and relocation of application components, but also allows flexible combination of service elements. Adaptation into the runtime changes in the system structure requires configuration of the service infrastructure. In addition, the user-centric approach requires that such an adaptive configuration be automatic. For example, when a node is plugged into the system, the services of it need to be available and discoverable by the other system users. Therefore, the execution framework must support auto-configuration and service discovery. Adaptive applications are based on various models: models of user preferences and behavior, models for quality of service (QoS) of connectivity, models of environment, and so on. These models are implications of extensive use of artificial intelligence. Due to different needs of applications, behaviors, and environments, the models needing to be supported are manifold: Bayesian, probabilistic, regressive, predicative, propositional, logical, and so on. Therefore, the generic service elements have to provide basic means of modeling, without restricting the actual form of the model, including a perception service that provides a distributed information base to store and retrieve pieces of knowledge, a model builder that takes care of parameter estimation according to the given estimation criterion, a model combiner that builds up a combined model from given submodels, and a model evaluator that returns the output of the model. Data management, including perceptions, user preferences, device capabilities, and application requirements as well as a user’s own data, will be of fundamental importance for mobile users. Therefore, there is a need for provision of a mobile distributed information base. This distributed database system, replicated worldwide, must incorporate several features such as consistency, efficiency in its access, reliability, and high availability. Additional essential features include intelligent synchronization, role-based views, and transactional operations with user-defined semantics of correctness. The models used to describe the state of the world of interest behavior of users or characteristics of the environment (e.g.,in a distributed system) need a common set of concepts so that apples and oranges are not intermixed. Therefore, the generic service elements must include an ontology service that provides a well defined basic set of concepts and a way to introduce new concepts with well defined semantics. In computing, three different worlds meet:


user preferences, device capabilities, and application requirements. Adaptive applications need to match the three worlds so that user preferences, device capabilities, and application requirements are met. This requires common representation of preferences, capabilities, and requirements as well as matching rules. In the future matching cannot only be string comparison but needs to be based on semantics. Therefore, the generic service elements must include a semantic matching engine that selects the most appropriate alternative from those available for a given criterion.

CONCLUSION The article presents the first results obtained in our work in the area of service architecture for the wireless world. Our belief is that the I-centric paradigm will be introduced stepwise in the future and will provide rich opportunities for new products, business models, and attractive services for customers. We would like to thank all participants of WWRF WG2 for their contributions and the lively discussions we have had over the last years.

REFERENCES [1] WWRF Book of Visions, to appear. [2] H. van Kranenburg et al., “Ubiquitous Attentiveness: Enabling Context-Aware Mobile Applications and Services,” 1st Euro. Symp. Ambient Intelligence, Veldhoven, The Netherlands, Nov. 2003. [3] A. K. Dey, D. Salber, and G. D. Abowd, “A Conceptual Framework and a Toolkit for Supporting the Rapid Prototyping of Context-aware Applications,” Human–Comp. Interaction, vol. 16, 2001, pp. 97–166. [4] T. Itao et al., “Context-Aware Construction of Ubiquitous Services,” IEICE Trans. Commun., vol.E84-B, no. 12, Dec. 2001, pp. 3181–88. [5] J. I. Hong and J. A. Landay, “An Infrastructure Approach to Context-aware Computing,” Human–Comp. Interaction, 2001, vol. 16, 2001. [6] P. Castro and R. Muntz, “Managing Context for Smart Spaces,” IEEE Pers. Commun., vol. 7, no. 5, Oct. 2000, pp. 44–46. [7] B. Kummerfeld et al., “Merino: Towards an Intelligent Environment Architecture for Multigranularity Context Description,” Wksp. User Modeling for Ubiquitous Comp.g, Mar. 2003. [8] T. Kindberg et al., “People, Places, Things: Web Presence for The Real World,” Tech. rep. HPL-2000-16, HPLabs, http://www.cooltown.hp.com/ [9] P. Mendse, C. Prehofer, and Q. Wei, “Context Management with Programmable Mobile Networks,” IEEE Comp. Commun. Wksp., Oct. 2003. [10] K. A. Hawick and H. A. James, “Middleware for Context Sensitive Mobile Applications,” P. M. C. Johnson, and C. Steketee, Eds., Proc. Australian Info. Sec. Wksp. and Wksp. Wearable, Invisible, Context-Aware, Ambient, Pervasive and Ubiquitous Comp., vol. 21, Conf. Res. and Practice in Info. Tech., pp. 133–42, Adelaide, South Australia, 2003.

BIOGRAPHIES STEFAN ARBANOWSKI ([email protected]) is head of the Competence Center for Open Communication Systems at Fraunhofer FOKUS. He holds an M.Sc. and a Ph.D. in computer science from the Technical University of Berlin. There he has given lectures on brroadband networks, mobile and personal communication, and agent and middleware technologies. He has participated in several national and international projects, such as IN/TMN Integration, the TINA auxiliary project Personal Communication Support in TINA, UMTS Network Aspects, and VHE Trail Specifications. Since 2001 he has been active in the Wireless World Research Forum (WWRF). He is chairman elect of WWRF WG2 for 2004. PIETER BALLON ([email protected]) is senior consultant and head of the Networked Economy and Cultural Industries


Research Team at the TNO Institute for Strategy, Technology and Policy in the Netherlands. He specializes in new business models for mobile and broadband services and platforms. He holds a Master’s degree in modern history and a postgraduate degree in information and library sciences. KLAUS DAVID ([email protected]) is full university professor and head of the chair of communication technology at the University of Kassel (UoK). He has 12 years of industrial experience in major companies like HP, Bell Northern Research, IMEC, T-Mobile (as head of group and UMTS project leader), and IHP (as head of department), with five years of international experience in the United Kingdom, Belgium, the United States, and Japan. He was co-founder of the ACTS On The Move project, one of the first projects about mobile middleware concepts, and initiator of the IST WINEGLASS project. O LAF D ROEGEHORN ([email protected]) is a senior researcher at the chair of communication technology at UoK. He holds an M.Sc. in computer science from the University of Dortmund, Germany. He is a Ph.D. student at the University of Duisburg with a scholarship grant from the DFG. He was a project leader at IHP on mobile middleware for 3G networks. He is the main contact at UoK for the EC 6th Framework Program, WWRF, and WWI. He heads the Mobile Middleware competence center at UoK. His research focus is on the mobile wireless Internet, including mobile middleware, personalization, and the development of mobile services.

In computing, three different worlds meet: user preferences, device capabilities, and application requirements. Adaptive applications need to match the three worlds so that the user preferences, device capabilities, and application requirements are met.

HENK EERTINK ([email protected]) received his M.Sc. in 1987 and his Ph.D. in 1994, both in computer science. He is a senior member of scientific staff at Telematica Instituut Central Organization, Enschede, the Netherlands. At Telematica Instituut he is responsible for the coordination of the research program Infrastructures for Context Awareness. He has more than 10 years post-doctoral experience in the area of distributed systems, software engineering, content distribution, and networking technology. His current research interests are in middleware for mobile applications, mobility-aware content distribution, and context awareness support at the middleware level. W OLFGANG K ELLERER [M] ([email protected]) heads the service platform and middleware technology group at NTT DoCoMo's European Research laboratories in Munich, Germany. Before he joined DoCoMo he worked at the Institute of Communication Networks at Munich University of Technology. In 2001 he was a visiting researcher at the Information System Laboratory of Stanford University, California. He holds Ph.D. and Master’s degrees in electrical engineering and information technology from Munich University of Technology. He is a member of the ACM. He is vice chairman elect of WWRF WG2 for 2004. H ERMA V AN K RANENBERG ([email protected]) received her M.Sc. in 1988 and her Ph.D. in 1992, both in electrical engineering. She is a member of scientific staff at Telematica Instituut with research and management experience in several projects in the field of content engineering, mobile services, and middleware for heterogeneous IPbased mobile environments. She is active in WWRF in the area of ambient awareness and personalization. Her present research interests are in the area of context awareness, mobility management, and tailoring of next-generation mobile services in a multi-access environment. KIMMO RAATIKAINEN [M'81] ([email protected]) has M.Sc. (1983) and Ph.D. (1990) degrees in computer science from the University of Helsinki. Since 1986 he has been employed in the Helsinki University Computer Science Department, as a full professor since 1998. Since January 2000 he has been a part-time principal scientist at the Nokia Research Center and from January 2002 also at the Helsinki Institute for Information Technology. His current research interests include wireless communication, middleware for mobile distributed systems, and operating systems. RADU POPESCU-ZELETIN [SM] ([email protected]) is a professor at the Technical University of Berlin and director of the Fraunhofer Institute for Open Communication Systems (FOKUS). He was chairman elect of WWRF WG2 from 2002 to 2003. He graduated from the Polytechnical Institute Bucharest, Romania, got his Ph.D. from the University of Bremen, Germany, and his habilitation from the Technical University of Berlin. He is Doctor honoris causa of the Polytechnical Institute Bucharest and Professor honoris causa of the Catholic University of Campinas, Brazil.

69