5th International workshop on middleware for pervasive and ad-hoc computing (MPAC 2007)

For more than a decade pervasive and ad-hoc computing have captured the imagination of the research community. The systems challenges they pose have been the focus of much research in the area. In recent years, as a result of these efforts a variety of middleware platforms and abstractions for pervasive and ad-hoc computing have been proposed aiming to address the identified challenges. Besides the pervasive and ad-hoc computing vision, this work is also motivated by a belief that underpinning middleware mechanisms are central in weaving together the multitude of sensing, computing, communication and information technologies driving developments in this area. However, despite significant progress in this area, a number of challenges remain. Proposed middleware platforms tend to focus on only some of the issues, e.g. location and context-awareness, reputation and trust managements systems, leaving others open. How these separate platforms are to be composed and interoperate to realise the pervasive computing vision is still a challenge. How to introduce autonomic capabilities to the individual platforms and their compositions remains unclear. At the same time, the proposed models and abstractions that these platforms incorporate are still lacking in theoretical rigour and foundation to allow formal analysis of their properties. Moreover, as the number of these models, abstractions and infrastructure components increases, the need emerges to develop appropriate frameworks for the comparative evaluation of alternative proposals both from the point of view of application developers and the operators of the infrastructure. Such evaluations require the development of benchmarks for pervasive and ad-hoc computing systems. Within this context the MPAC2007 workshop building on the success of the 2003, 2004, 2005 and 2006 workshops, sought to further develop the roadmap for research on the essential middleware abstractions and infrastructures for ad-hoc and pervasive computing. In addition to this, the workshop also continues to play an important role as a venue for discussing novel middleware abstractions and infrastructures. This year the workshop received a high number of quality submissions. Eleven papers out of the twenty eight originally submitted were chosen for these proceedings. These papers cover a broad range of issues including service orientation and discovery, support for self-* properties, event-based systems for pervasive environments and sensor networks, privacy management, and reasoning for context-awareness and situation recognition. More specifically, Dimitrov et al. propose a probability reasoning framework for context-awareness in smart homes, while Holzmann proposes a rule-base reasoning approach for recognition of complex spatiotemporal situations of artefacts. Taherian and Bacon describe a state-base publish/subscribe framework for multi-user sensor networks, while Selim et al. introduce a replication based type and attribute oriented event-based middleware over a structured peer-to-peer network. Chen and Lukkien describe an overlay architecture for secure service orchestration within service-oriented virtual communities, and Bonino da Silva Santos et al. propose a service-oriented middleware providing support for context-awareness. Edwards et al. identify the challenges that self-* properties pose in pervasive computing and how commonly used strategise like dynamic software update, service discovery, transparent replication and logical mobility are deployed to address them, while Cheung-Foo-Wo et al. describe a middleware based on an event-driven software component model with a schema-based self-adaptive architecture for distributed services composition. Trinta et al. evaluate a role playing games developed on top of a service-oriented middleware for cross-media games. Viterbo et al. describe a discovery service based on the notion geographic location scope. Finally, Wakeman et al. propose a scheme for managing privacy in trust management with pseudonymous group membership