Joint Behavior Detection in Wireless Sensing Networks

Node Security is very important because there are many kinds of node behavior coexisting in an unknown network. Most of node trust evaluation existing is based on energy element and communication element, which is not sufficient to describe behavior security of a node. This paper evaluates node trust value by the node behavior effect on network security. It presents behavior cognition parameters and joint cognition model for 9 types of common node behaviors to set a correct trust value. The results show that in a static and dynamic network environment it reaches a high grading accuracy which can exclude malicious node from network activity effectively.