Intelligent Hybrid Coalition for Distributed Sensor Network Clustering

As DSN nodes are intelligent, multi-agent paradigms are more im- portant to adapt to sensing circumstance tactically and semantically. So we pro- pose IHCF(Intelligent Hybrid Coalition Formation) model for constructing dynamic clusters of sencing nodes and synthesizing their sensing signals/informaiton under two regimes(task sharing, result sharng). In this model, a cooperation layer including negotiation and heuristic clustering processes is designed. Latent features and meanings in the cluster are emerged with the help of IHCF scheme. To do this, projection tecnhiques usning dimension reduction is applied. It shows more efficient and adaptable perfor- mance enough to compete with existing 2-tier classifier systems. Currently IHCF is designed for sensing areas of object recognition(e.g. face)