Proactive trust classification for detection of replication attacks in 6LoWPAN-based IoT

Abstract

The 6LoWPAN standard has been widely applied in different Internet of Things (IoT) application domains. However, since the nodes in the IoT are mostly resource constrained, 6LoWPAN is vulnerable to a variety of security attacks. Among others, replication attack is one of the severe security threads to IoT networks. This paper therefore proposes a trust-based detection strategy against replication attacks in IoT, where a number of replica nodes are intentionally inserted into the network to test the reliability and response of witness nodes. We further assess the feasibility of the proposed detection strategy and compare with two other strategies such as brute-force and first visited strategy via a thorough simulation. The evaluation takes into account the detection probability for compromised attacks, the execution time of transactions and rate of communication failure. The simulation results show that while maintaining detection runtime on average 60 s for up to 1000 nodes, the proposed trust-based strategy can significantly increase the detection probability to 90% on average against replication attacks and in turn significantly reduce the communication failure.