A trusted framework for secure routing in wireless ad hoc networks

This paper proposes a novel quantitative trust model that supports both first-hand (direct) and second-hand (recommendation) trust opinion to calculate the final trust of each node in the network. In order to compute direct trust of a node, the proposed trust model utilizes the theory of ARMA/GARCH to predict the trust parameters and a probabilistic model to combine those parameters. The proposed trust model collects recommendation trusts from common neighbors of the node under review and combines these trusts to get the resultant trust using a weighted combination model. Based on proposed trust model, a routing protocol has been driven to provide trustworthy routes. Simulation results show that the proposed protocol gives significantly higher packet delivery fraction as compared to CBRP even in presence of malicious nodes in the network. In summary, our proposed model is lightweight in terms of computation and powerful in terms of flexibility and accuracy in managing trust in WANET.