Gaussian trust and reputation for fading MIMO wireless sensor networks

In this paper, we develop a framework for Gaussian trust and reputation computation in MIMO wireless sensor networks. An analytical framework is derived to characterize the direct interaction based reputation for fading MIMO wireless scenarios. Subsequently, this framework is expanded to incorporate second hand information based on indirect interaction which leads to a significant improvement in the efficiency of reputation computation. The proposed algorithm employs a Bayesian Expert Opinion model for reputation fusion and thus is robust against misinformation arising due to malicious sensor nodes. Further, we develop a novel trust computation scheme for the multivariate Gaussian reputation based MIMO wireless sensor network. We also extend the proposed reputation and trust computation model to the scenario of spatially correlated sensor measurements. Simulation results validate the performance of the proposed Gaussian trust computation framework and its resilience to the presence of malicious sensor nodes in the wireless sensor network.