A throughput descent and energy efficient mechanism for fault detection in WSNs

Abstract

In wireless sensor networks (WSNs), sensor nodes are deployed densely in large numbers to increase the quality of services (QoS). However, QoS is significantly affected by faults in such networks. The probability of faults in WSNs increases with increase of sensor nodes. In order to provide better QoS, it is essential that faulty nodes should be diagnosed and handled timely without affecting the underlying work of the network. In this paper, our main focus is on diagnosing transient and permanent fault by communicating an optimal number of messages. We propose a mechanism to diagnose such faults without degrading the throughput of WSNs. The proposed approach is reactive in nature, which during fault diagnosis generates traffic smoothly in the network due to which the network never degrades its performance. Scheme also allows faults to occur during diagnosis process. The result shows that our scheme diagnoses faults correctly and efficiently and is able to maintain the desired throughput during fault diagnosis.