A Distributed Bayesian Algorithm for data fault detection in wireless sensor networks

This paper proposes a new Distributed Bayesian Algorithm (DBA) for data fault detection in order to address problems of data fault in wireless sensor networks. Although many methods for data fault detection have been proposed, the diagnosis accuracy is still very low especially in the situation of the densely deployed WSN with a large number of faulty nodes. In DBA, the Bayesian network is introduced to calculate the fault probability of sensor nodes, and the fault probability will be adjusted by exploiting the border nodes to improve the accuracy of fault probability and avoid the negative impact of the large number of faulty nodes. A comprehensive simulation has been made to compare the performance of DBA and Distributed Fault Detection (DFD). The simulation results show DBA can improve the fault detection accuracy significantly even in the situation with a large number of faulty nodes.