A Mutual Evaluation Based Trust Management Method for Wireless Sensor Networks

We propose a lightweight trust system for the clustered wireless sensor networks based on the mutual evaluation between the cluster heads and sensor nodes. We evaluate the trust level of a cluster head in two aspects, namely, trust level as a service provider and a supervisor. We consider multidimensional trust attributes to compute the global trust value of a node. By means of the Petri net (PN), we illustrate the performance of an entity in our trust model. Theoretical analyses as well as simulations are done and the results showed that our model with linear computational complexity had less memory and communication overhead when compared with the current state-of-art trust management scheme. In addition, our model can detect malicious and selfish entities (especially the malicious cluster head) and resist various attacks efficiently.