A trust-based scheme employing evidence reasoning for IoT architectures

IoT architectures heavily depend on wireless resource-constrained nodes that collect and transmit a plethora of information. The broadband wireless medium enables malicious nodes to perform a number of attacks. Cryptographic means along with trust-based schemes are often deployed to combat these attacks. We propose a centralised trust-based scheme where all nodes monitor their neighbors, and report their findings to a fusion center. Each report is followed by a reliability measure, computed using the Signal-to-Interference-plus-Noise-Ratio. Fusion is performed employing a belief distribution with reliability method. We consider malicious nodes that periodically drop packets based on some probability. The evaluation results show that the proposed scheme can accurately detect the malicious behavior.