A lightweight anomaly mining algorithm in the Internet of Things

Abstract

The security of Internet of Things (IoT) has already become a thorny problem because of opening deployment and limited resources. Thus, as the essential part of intrusion detection anomaly mining gets more and more attention. However, complexity of algorithm is the vital issue due to the specialty of IoT. Meanwhile, traditional methods with Euclidean distance may cause misjudgment at some extent. So this paper proposes a lightweight anomaly mining algorithm which employ Jaccard coefficient firstly as the judging criterion instead of Euclidean distance. The experiment verifies the availability of proposed algorithm.