一种基于自回归模型的网络异常检测方法

Network anomaly detection approach based on autoregression model

随着网络技术的不断发展,计算机病毒、网络攻击等问题也日益严峻。维护网络的安全和稳定,是一个亟须解决的问题。针对该问题,介绍了一种基于自回归模型的网络异常检测方法,该方法将局部的网络流量看作统计学上近似的平稳。OPNET上的仿真实验表明,该方法能有效检测出网络异常,误报率低。

With the development of network technology, it is more and more serious to deal with the computer virus and network attacks. It is critical to maintain the security and stableness so as to ensure the efficiency and proper operation of the network. To solve this problem, This paper introduced a method of network abnormity detection based on autoregression model. It assumes the local network traffic is statistically steady. The simulation on OPNET demonstrates this method can effectively detect network anomalies and its misstatement rate is low.