Grid-based Fault-Tolerant Event Detection in Wireless Sensor Networks

Abstract

In this paper, we present a grid-based fault-tolerant event detection scheme for wireless sensor networks. The sensor field is divided into square-shaped cells, called virtual grids here, where sensor nodes in each grid form a cluster. Each grid is further divided into subgrids of equal size. Each cluster head receives sensor readings of its member nodes and counts the number of sensor nodes reporting a "1" (an unusual reading) in each of its subgrids. An event is detected by applying a threshold test and finding square regions with highest detection rate. Cluster heads communicate with their neighboring cluster heads to exchange the numbers of nodes reporting a 1, requiring only a small amount of communication overhead to make a final decision on an event. High event detection performance is achieved even for a relatively high fault probability by employing a moving average filter to tolerate most transient faults. Computer simulation shows that the proposed scheme can achieve extremely low false alarm rate, while maintaining high event detection accuracy.