Learning the Dynamics for Anomaly Detection in Wireless Sensor Networks
Yi Gao, Chun Chen, Jiajun Bu, Wei Dong, Lei Rao And Xianghua Xu
Anomaly detection is an essential functionality ofWireless Sensor Networks (WSNs). However, the dynamic nature of WSNs makes the anomaly detection challenging. The environmental dynamics or internal topological dynamics pose difficulties to differentiate abnormal behaviors from normal ones. We observe that indirect correlations among multiple attributes of a sensor node can be utilized to capture and model the dynamic behaviors. Prior studies overlooked indirect correlations while in this study we exploit it for detecting anomaly efficiently and accurately. We propose ICAD, an indirect correlation based anomaly detection approach. By applying the Markov chain, the dynamic behaviors are modeled and subsequently used to detect anomalies. Compared with prior approaches, ICAD can detect different types of anomalies simultaneously and the performance is not sensitive to network dynamics. ICAD is implemented based on TinyOS/TelosB platform. Evaluation results show that ICAD achieves high detection accuracy with low overhead.
Keywords: Wireless sensor networks, anomaly detection, indirect correlation