FID: Fuzzy Based Intrusion Detection for Distributed Smart Devices

Internet of Things (IoT) applications are driving the world to be smart by providing limitless services with optimized time and resources. Security and safety are paramount to ensure the reliability of IoT applications. For this purpose, Intrusion Detection Systems (IDS) must be employed in order to discover the intrusions and to contain and possibly correct their impact on the network. Reported IDS are not feasible in the context of the IoT. On the one side, rule based IDS require excessive storage cost to save and update attack signatures and they cannot detect the unknown attacks. On the other side, behavioral based IDS are more feasible in the context of IoT but they fail to detect some known attacks. In this paper, a Fuzzy based IDS is proposed where the undesired behavior is expressed using fuzzy logic. Also, after determining the fuzzy values, the decision is filtered by the Base Station. The parameters' values that influence on the proposed approach are determined following extensive simulations. Analytical study and simulation results show that the proposal outperforms existing solutions.