On-off attack detection in trust model using intra-daily variability for the IoT

The growth of the internet of things (IoT) increases the need to develop the trust computational model for heterogeneous networks with various IoT devices. Trust models are considered as an effective tool to mitigate insider attacks induced by IoT devices. However, trust models are exposed to on-off attacks, in which devices randomly exhibit good and bad behaviors to avoid being categorized as low-trust devices. The objective of this work is to recognize the malicious devices executing on-off attacks in IoT applications. This paper introduces an on-off attack detection strategy for the trust computational model based on the non-parametric index named intra-daily variability (IV). IV indicates trust fragmentation which depends on the frequency and the transitions between periods of low and high trust values of a device. The higher value of IV indicates the occurrence of fragmented trust values and the lower value of IV indicates the occurrence of non-fragmented trust values. Experimental results show that the proposed model outperforms the baseline methods by increasing the on-off attack detection rate