Node Trust Prediction Framework in Mobile Ad Hoc Networks

Trust quantification methods depends on the deployment region, network applications, level of security required. Considering the environment of mobile ad hoc networks, in this study we propose a new trust model. This new model is divided into two parts: trust assessment and trust prediction. The process of node trust assessment is based on node's historical behaviors, in which the trust decision factors include the subjective reputation and indirect reputation. Then utilizing the obtained trust data sequence, we propose an improved dynamic grey-Markov chain prediction measure to predict node's trust for future decision making. In which we eliminates grey bias via using unbiased GM(1,1) model. Moreover, in order to improve anti-jamming performance, we introduce the fuzzy classification method to classify the fuzzy states into some reasonable groups. Experimental results have been conducted to evaluate the effectiveness of the proposed model. It is proved in theory and practice that the prediction results not only are more reliable but also can help the decision maker with grasping the change tendency of trust and making proper decisions.