A Trust Management Framework for Detecting Malicious and Selfish Behaviour in Ad-Hoc Wireless Networks Using Fuzzy Sets and Grey Theory

摘要

As wireless network applications evolve, they require interaction between many entities (protocols, middleware etc), and an increasing requirement for the design of secure applications among the entities is trust management. Consequently, many new attacks against networks are aimed at the trust management. In this paper, we develop a new trust management framework (TMF) for wireless networks. The proposed framework applies Grey theory combined with Fuzzy Sets to calculate a node's trust value based on observations from neighbour nodes'. The new TMF employs multiple rather than a single parameter to decide the resulting trust value. Simulations conducted in an 802.11 based wireless network show that the new framework can not only identify abnormal trust behaviour, but can also effectively find which aspect of the metrics used to establish the trust value for a node are abnormal, and hence identify the strategy the attacker is using against the TMF.