Fuzzy based Adaptive Threshold Determining Method in Sensor Networks

Many sensor network applications are dependent on the secure operation of networks, and serious problems may result if the network is disrupted or injured. An adversary using these restrictions can compromise sensors and use them to inject bogus reports into the network. Bogus reports can lead to both false alarms and the depletion of limited resources. A statistical en-route filtering scheme (SEF) was proposed to detect such bogus reports during the forwarding process. In SEF, it is important to determine the number of the message authentication code (MAC) to attach to the event report when a real event occurs. The number of MACs which attached to the report trades off detection power and overhead. We propose a threshold determining method to save energy using a fuzzy logic. Fuzzy rule-based systems are used to divide the sensor field into several areas considering the network situation. Each area uses a different number of MACs so that the size of the report can be reduced by the distance between the base station and the region where the event occurs in the hop count. Simulation demonstrates that the proposed method is resilience and energy efficient. Key words: Sensor networks, false data filtering, fuzzy logic, statistical en-route filtering 1