Enhanced Context Searching Based on Structured P2P

Most of context aware systems are based on client-server model which has some limitations, e.g. a bottle neck problem, single point of failure. The most disadvantage of client-server model is when all nodes need to update their context information frequently and quickly; it will cause massive information flooding in a system. By applying P2P model to context aware system, this problem can be reduced. However, it will create other problems, e.g. context searching will take more time when a number of hops are large, ontology distribution issue, problems of storing ontology and reasoning process. In this paper, we have proposed MF-P2P (Multi Finger P2P Table) which can reduce a context searching time in P2P model. Our design is based on Chord mechanism. By distributing processes to peers, we have shown that the system can perform a fast searching on high complexity of ontology and context.