Trust Evaluation Model against Insider Attack in Wireless Sensor Networks

摘要

Insider attack is one of the key attacks in wireless sensor networks(WSNs). Despite its destructive force to network functions is huge, insider attackers are difficult to be detected only by the traditional techniques. Meanwhile, existing secure routing schemes are also vulnerable to insider attack and low energy-efficient due to their deterministic nature. Once a normal node is compromised or captured, the adversary can endanger all information transmitting through this node. In this paper, we propose a novel trust evaluation model to enhance the data security in WSNs. Under our design, each node computes the trust value of its 1-hop neighbors based on their multiple behavior evaluation with no requirement on a prior knowledge about normal/compromised sensor activities and build a trust management. Moreover, our scheme is application friendly, which can be employed to monitor many aspects of sensor networking behaviors and build the secure routing. In our secure routing, besides security, high energy-efficiency is also considered. Simulation results indicate that data shares can bypass most compromised nodes in the transmission path.