DTMS-IoT: A Dirichlet-based trust management system mitigating on-off attacks and dishonest recommendations for the Internet of Things

Abstract

The Internet of Things (IoT) is based on the idea that things surrounding the human living space can be connected to the Internet, allowing smarter living environments and more comfortable life. However, adoption of the IoT might not be approved unless security issues are solved. In this paper, we propose a new Dirichlet based trust management system for the IoT called DTMS-IoT. This system detects nodes malicious behaviour which permits to mitigate both on-off attacks and dishonest recommendations. It also uses service levels and things capacities to reinforce security. As in many trust management systems introduced in the literature, the computation of trust in our solution is based on direct observations and recommendations. The novelty of our approach can be summed up in two aspects. The first aspect concerns recommendations that we do not consider unless direct observations are not credible. However the second aspect tries to prevent On-Off attacks using the precaution factor and mitigates dishonest recommendations based on k-means selection algorithm and guarantor things. The effectiveness of the proposed system is proved by simulation against on-off attacks and good/bad mouthing attacks.