Towards Intelligent Behaviour for Autonomic Communications
Autonomic Communications is a trend inspired by other research initiatives which serve either the ultimate goal of autonomy or partial goals which contribute with their functional achievements to the overall purpose. Among those aiding trends is the MPEG-7 framework from ISO and IEC. MPEG-7 features the following activity types: navigation and access, spatial 2-D coordinates, color-structure descriptor, parametric motion, motion activity, face recognition, sound recognition and indexing description tools. The Pervasive Autonomic Communication Environment (PACE) which we present, in one of its scenarios,covers an entire whole metropolitan area. The individual autonomic elements are heterogeneous in nature and belong to various domains/owners. Cooperation modes within the PACE include: goal orientation with information sharing, resource exchange, and other techniques facilitating reaching the ultimate goal. Optimization measures for the autonomy of the system which we rely on include among others: learning to optimize the system behavior e.g. reduce number of hops, buffer adaptively to improve quality, perform smart caching, execute batches, and acquire information proactively.