基于流量预测和相关系数的异常检测

安全问题是无线传感器网络应用面临的重要挑战之一。提出了一种基于混沌时间序列预测和相关系数相结合的异常入侵检测方法,该方案首先对正常情况下无线传感器网络节点的流量应用混沌时间序列方法进行预测,然后根据传感器节点的流量预测序列和实际流量序列的相关系数变化来进行异常检测。实验结果表明,该方案在入侵检测率达到相当高的程度,与当前典型的WSN入侵检测方案相比较具有更优越的性能。

Traffic Prediction-Based and Correlation Coefficient for Anomaly Detection

Security is one of the key problems in wireless sensor networks (WSN) applications.An anomaly detection approach is proposed based on chaotic time series prediction and correlation coefficient (CTSP),firstly the method of chaotic time series prediction is applied to predict the traffic of WSN nodes in normal,then we use correlation coefficient of traffic prediction series and real traffic series of WSN nodes to identify anomalous nodes.The experimental results demonstrate that the scheme achieves higher acc...