Trust management system for the internet of things (IOT)

Abstract

The Internet of Things comprises of diverse network of homogeneous and heterogeneous objects (Things) that are exchanging data continuously and can be accessed through networks ubiquitously. The IoT nodes consist of limited resources i.e. energy, processing, communication and computation capabilities. Therefore, such limitations create challenging issues in providing security and raise the possibility of attacks and risks. A trust management model is considered as a method to defend IoT system against malicious attacks and provide reliable data exchange. In this thesis, we focus on identifying malicious behaviour of a node. Our goal is to detect and mitigate potential On-Off attacks in a multi-service IoT environment. The proposed model employs distributed trust management scheme. So, each node is autonomous in decision making and communicating directly with the peer nodes in the network. This assists in gathering direct information, to evaluate trust among neighbour nodes. Trust is evaluated dynamically based on node(s) behaviour. A malicious node may behave well (Off state) for a certain time to attain positive trust value and later deliberately exploits and misbehave (ON state) to affect peer nodes, in case of On-Off attack. The proposed algorithm efficiently detects the misbehaving node and notifies the peer nodes about identified malicious node to mitigate and avoid potential On-Off attack(s) in the IoT network. Keywords: IoT, Trust, Trust Management System, Security, Dynamic, Distributed, Attacks, Weighted-Sum.