An Intelligent Broker Architecture for Context-Aware Systems

Context-aware computing is an emerging paradigm to free everyday users from manually configuring and instructing computer systems. As the general trend of computing is progressing towards an open and dynamic infrastructure, building context-aware systems can be difficult and costly. In order to build successful context-aware systems, we must develop an architecture to reduce the difficulty and cost of building these systems. This PhD. dissertation proposal describes a research plan to develop a broker-centric agent architecture that is aimed to relieve the burden of capability-limited agents of acquiring and reasoning about contexts, and to protect the privacy of users in a contextaware environment. The implementation of the Context Broker Archiecture will explore Web Ontology Language for modeling contexts and privacy policies, Jess for building a hybrid reasoning mechanism and JADE/FIPA for realizing broker behaviors and agent communications.