Abstract:This paper targets the identification of faulty sensor nodes, which is one of the most important sources of faults in wireless sensor networks. To explore an effectual technique for node self detection, a new algorithm, Node Self Detection by History data and Neighbors (NDHN) is proposed in this paper. This algorithm collects the characteristics of the nodes to compute the biases through historical measurements and neighbors' data to make a judgment. The algorithm is hierarchical and thus can be extended for large networks. Simulation results indicate that the identify accuracy is high.

A Self Detection Technique in Fault Management in WSN