Wireless sensor network are composed of many wireless sensing devices namely sensor nodes, they are small in size, limited in resources and randomly deployed in harsh environment. Therefore, it is not uncommon for sensor networks to have malfunction behaviour, power shortage or network failure. To address this issue, we proposed a new scheme based on Causal Model Method (CMM) which applies fault sources analyzer for component-level fault diagnosis in wireless sensor networks. Our new method consists of three phases to define the node failure sources as “collect, classify, and correct”. Once the fault source has been classified, CMM mechanism will enable reconfigure process to compensate for the erroneous sensor nodes impact. We have conducted a simulation study to our work using the Georgia Tech Network Simulator (GTNetS). Our simulation results show that CMM can improve wireless senor networks performance and reliability.
