Wireless Sensor Network Intrusion Detection Model Studies

Wireless sensor network (WSN) is a special kind of wireless Ad hoc network. It has been widely applied in many ways. Because of the unique characteristics (e.g. lack of physical protection of nodes and limited energy, etc), wireless sensor network is easier to attack. So far, Wireless sensor network security solutions are basically from the research point of view of prevention. But they can not be based on prevention against malicious internal attacks. In practice, the damage of internal attack may be greater than external attack. Therefore, intrusion detection system presents a second line of defense network protection, and is essential for providing a highly secure information system. Modeling can effectively identify potential intruders, and provide further protection. This paper focus on intrusion detection model in wireless sensor network.Because of the unique characteristics of WSN, traditional network intrusion detection system cannot be deployed in WSN. At present, there have been some studies on intrusion detection in wireless sensor network. Without considering the impact of the data source, these intrusion detection Models restrict their capabilities to specific attacks only. The attacks of the five layers of wireless sensor networks are different. It is difficult to improve intrusion detection accuracy by detecting a certain kind of attack. Therefore, an effective intrusion detection model should be able to detect different attacks.On the basis of the intensive analysis of the traditional data fusion limitations and security in wireless sensor network, a Mobile Agent Multilayerd Intrusion Detection Model(MAMIDM) is proposed in this paper. The data source is classified into intracluster data source and intercluster data source. By combining the multi-layer conception with the mobile agent technology, the function of selecting the different processing according to the different data source can be realized. The global detection algorithm located in the global detection engine is proposed to integrate the local detection engine and the mobile agent detection engine. Results show that mobile agents can reduce the energy consumption of wireless sensor network and extend the networks lifetime. Multi-layer intrusion detection can provide a higher detection rate and low false alarm rate.