Bio-inspired monitoring of pervasive environments

Abstract

Successful deployment of the pervasive computing paradigm is based on the exploitation of the multitude of participating devices and associated data. It becomes therefore evident that there is a necessity to provide optimal resource discovery mechanisms, the effectiveness of which will constitute the foundation for the efficient operation of pervasive computing environments. To mitigate the drawbacks brought by the inherent nature of pervasive environments, i.e. dynamicity, heterogeneity and scalability of the network infrastructure, we propose to employ P2P overlay networks that lead to more manageable topologies and optimize resource monitoring by scaling down the degree of complexity. In particular, we take advantage of bio-inspired self-organization mechanisms to construct reliable P2P overlays and thus provide more robust and adaptive monitoring solutions. High-level policy-based management operations driven by monitored context information enable a further level of runtime adaptation and optimization, subject to the overall application requirements. We report here on our ongoing work in this research area.