A dynamic trust model in internet of things.

Indicating trust or distrust of a node is a key issue in the trust management of IoT. However, there are some challenges for the management with anonymous nodes, inaccurate communication, etc. Due to huge number of nodes in IoT, one of the possible solutions is predicting trust and distrust values. In this paper, we propose a dynamic trust model based on direct and indirect trust computation, and the most important part of a model, trust prediction. The prediction method mainly depends on the combination of exponential smoothing and a Markov chain. We employ exponential smoothing to predict trustiness and a Markov chain to fix the deviation of the prediction. To test our scheme, we create a simulation to evaluate its performance and effectiveness. The simulation results are encouraging. [ABSTRACT FROM AUTHOR]