LogitTrust: A Logit Regression-based Trust Model for Mobile Ad Hoc Networks

Mobile ad hoc networks (MANETs) have been utilized to execute many applications in diverse environments. Trust is an effective mechanism to cope with misbehaving nodes. However, implementing trust in MANETs is confronted by several obstacles, i.e., no centralized authority, dynamic environments, and limited observations which hinder trust accuracy. In this work, we propose a novel logit regressionbased trust model called LogitTrust to model dynamic trust for service-oriented MANETs wherein a node can be a service requester (SR) or a service provider (SP). The novelty of our design lies in the use of logit regression to dynamically estimate trust of SPs based on their distinct behavior patterns in response to environment changes. We demonstrate that LogitTrust outperforms traditional approaches based on Bayesian Inference with belief discounting in terms of trust accuracy and resiliency against attacks, when given the same amount of limited observations, while maintaining a low false positive rate.