Towards a flexible middleware for context-aware pervasive and wearable systems
Marco Muro, Michele Amoretti, Francesco Zanichelli, G. Conte · 2010
Both ambient intelligence and wearable computing do need novel hardware and software, yet another enabling factor is middleware, which should be highly capable, flexible and efficient, allowing for the reuse of existing pervasive applications when developing new ones. In the considered application domain, middleware should also support quick runtime re-configuration of systems, interoperability among different platforms, efficient communications, and context awareness. In the on-going “everything is networked” scenario scalability appears as a very important issue, for which the peer-to-peer (P2P) paradigm emerges as an appealing solution for connecting software components in an overlay network, allowing for efficient and balanced data distribution mechanisms. In this paper we go through the state of the art of middleware for pervasive and wearable computing, and we outline our approach that is geared to autonomic systems and based on the P2P paradigm as well as on dynamic refactoring of software entities.