A Gaussian-Based Trust Model for Nodes of Sensor Networks

At present, the security mechanism of the sensor network is almost based on node authentication and data encryption, which can address security issues for Internet connection and information transformation. However, it lacks effective information security monitoring mechanism for node betrayal and malicious attacks occurring inside sensor networks. By analyzing factors including node performance, behavior and environment impact, etc., this thesis proposed a Gaussian-Based Trust Model (GBTM) for nodes of sensor networks against the insufficient adaptability of the current trust model. By GBTM, network monitoring cell can be utilized to build multiple Gaussian functions corresponding to monitored nodes to raise the conditional adaptability for computing the credibility of monitored nodes. And the online behavior of nodes can be evaluated by setting trust value in long or short term so that automatic credibility state and the credibility monitoring across the Internet can be realized. From the simulation analysis, it can be judged that GBTM has higher accuracy and dynamic adaptability compared with other known trust models.