Multidimensional Trust-Based Anomaly Detection System in Internet of Things

Trust Management (TM) has been playing an important role in dealing with security and privacy issues in the Internet of Things (IoT). Following this trend, we propose a trust-based anomaly detection system which provides a closed loop of trustworthiness computing, decision-making and trust reevaluation. The proposed trust model considers multidimensional trust elements including reputation, Quality of Service (QoS) and social relationship, the result of which is employed to instruct the device to take appropriate security policies against its peers. Moreover, the detected anomaly event will trigger the reevaluation of the peers trustworthiness. To evaluate our system, we consider a shopping mall scenario with a great many of IoT devices, and the simulation results show our system achieves very low false alarm rate under proper trust level threshold.