Anomaly Detection in Wireless Sensor Networks Based on Improved GM Model

Abstract

Aiming at the problems of poor detection effect, high rate of missed detection and high rate of false detection in traditional methods, an anomaly detection method for wireless sensor networks based on improved GM model is proposed. The multilateral measurement method is used to locate the nodes in the wireless sensor network, and the state tracking of the running track of the located nodes is carried out. According to the tracking results, the self similarity between nodes is measured by Hurst index. Based on the measurement results, the improved GM model is used to predict the abnormal nodes. The abnormal values of the wireless sensor network are calculated by the distance between adjacent points, and the state of the current node is judged, this completes the anomaly detection of wireless sensor networks. The experimental results show that the proposed method is effective in anomaly detection of wireless sensor networks, and the rate of missed detection and false detection is low.

Keywords

• hurst index
• network anomaly
• node positioning
• self similarity
• wireless sensor