Intrusion Detection for Wireless Sensor Network Based on Traffic Prediction Model
Han Zhijie, Wang Ruchuang · Physics Procedia · 2012
In this paper, the authors first propose an efficient traffic prediction algorithm for sensor nodes which exploits the Markov model. Based on this algorithm, a distributed anomaly detection scheme, TPID(Traffic Prediction based Intrusion Detection), is designed to detect the attacks which make more influence on packet traffic, such as selective forwarding attacks, DOS attacks. In TPID, each node acts independently when predicting the traffic and detecting an anomaly. Neither special hardware nor nodes cooperation is needed. The scheme is evaluated and compared with other method in experiments. Results show that the proposed scheme obtain high detection ratio with less computation and communication cost.