Boundary Discovery Based on Overlapping Clustering in Non-Uniform Wireless Sensor Networks

Wireless sensor networks (WSNs) detect and monitor the outside physical state by the sensor nodes organizing automatically. Clustering is a standard approach for achieving efficient and scalable performance in WSNs. In the applications of analysis topology discovering, geography routing, tracking and so forth, we need to resolve how to search for the boundary nodes and form the network boundary. Different from former boundary discovery algorithm basing on single node, this article proposes a self-organization overlapping algorithm for clustering (N-SOCA) to generate overlapping clusters, and then carries out clusters boundary fusion to realize WSNs boundary. The simulation result indicates that the algorithm can obviously enhance the network life cycle and generate the network boundary effectively. Furthermore, N-SOCA and the boundary discovery algorithm proposed in this paper are scalable regardless of the network size.