Abstract:

Wireless sensor networks (WSNs) consist of a large number of sensor nodes. The clustering architecture can deal with self-organization of large-scale networks, so clustering is a standard approach to achieving efficiency and scalability. In the applications of topology discovering, geography routing, tracking and so forth, overlapping clusters are useful, and recognizing boundary nodes is important. Different from the former boundary search algorithm based on single node, a distributed self-organization overlapping clustering algorithm in a random network is proposed to generate overlapping clusters, by means of which the clusters borderline is fused to form boundary of WSNs. Moreover, the results of simulations indicate that the algorithm can construct balanced clusters, search the network boundary effectively and enhance the network survival period obviously.