Bio-inspired context gathering in loosely coupled computing environments
Context-awareness is a key requirement of human-centric computing systems. Applications may ease user interaction or even anticipate the behavior of the user when utilizing information about the current context. The pervasive provision of context data represents a major challenge in this scope. For that reason we introduce an approach for gathering, disseminating, and interpreting context data in dynamic, highly distributed systems, which are mostly disconnected from central networking infrastructures. On the one hand we describe architectural consideration addressing functional elements and their organization in the computing environment. On the other hand we incorporate a model for request-driven context gathering and a biologically-inspired approach for weighting, storing and forwarding context data. The conceptual considerations are complemented with a description of our first efforts in realizing our approach on top of a peer-to-peer framework.