Multi-Feature Transmission Behavior Analysis-Based Intrusion Detection in IoT Networks

Abstract

Intrusion attack in Internet of Things (IoT) has been well studied. There exist number of approaches which detect the intrusion attack like payload and signature based approaches. However, they suffer to achieve higher performance in detecting intrusion attack in IoT networks. To handle this issue, an efficient Multi Feature Transmission Behavior based Intrusion Detection model (MFTBIDM) is presented in this paper. The method considers payload, frequency, signature, behavior of IoT nodes in the network to perform intrusion detection. To perform this, the Edge server monitors the frequency of data transmission, the signature of data packets towards any service, payload of data being sent, and behavior of IoT nodes. Accordingly, the method computes different trust score like Payload Trust Score (PTS), Frequency Trust Score (FTS), Signature Trust Score (STS), and Behavior Trust Score (BTS). Using all these trust scores, the method computes the value of Legitimate Transmission Score (LTS) to perform intrusion detection. The proposed method improves the performance of intrusion detection.