ABSTRACT

Wireless Sensor Network (WSNs) have become a new information collection and monitoring solution for a variety of application. In WSN, sensor nodes have strong hardware and software restriction in terms of processing power, memory capability, power supply and communication throughput. Due to these restrictions, fault may occur in sensor. This paper presents a distance based fault detection (DBFD) method for wireless sensor network using the average of confidence level and sensed data of sensor node. Simulation results show that sensor nodes with permanent faults and without fault which was judged as faulty are identified with high accuracy for a wide range of fault rate, and keep false alarm rate for different levels of sensor fault model and also correct nodes are identified by accuracy.