Approach for modelling trust in cluster-based wireless ad hoc networks

In this study, the authors propose a cluster-based trust management model that efficiently detects the malicious nodes and restricts them to be on a route in wireless ad hoc networks. In contrast to previous works, the trust of a node is calculated using various trust attributes having substantial effect on reliable routing in the networks. In the proposed model, each node periodically predicts the value of each trust attribute about other nodes using autoregression. Subsequently, the direct trust is estimated using the weighted combination of trust attributes and it is fine tuned using proportional-integral model. All these recommendation trusts, from common neighbours, are collected and combined by the clusterhead to quantify the trust, and hence the routing is reliable and secure in the proposed model. Simulation results show that the proposed trust model provides better throughput and packet delivery ratio in presence of malicious nodes compared to other existing schemes.