论文部分内容阅读
全网链路流量监测对流量工程和网络攻击监测都有重要意义但是一直很难实现。该文提出了一种基于随机矩阵理论(RMT)的监测方法,利用流量协方差矩阵和随机矩阵理论预测结果进行比较,从两者差异中提取流量的时间相关信息。在网络中布置少量观测点,利用协方差矩阵的最大特征值能够准确的获取高速和低速链路的流量信息,并且通过最小特征值监测到观测点是否工作。该方法节省存储资源,计算时间短,获取信息多,是一种高效的监测方法。
The whole network link traffic monitoring traffic engineering and network attack monitoring are important but has been difficult to achieve. In this paper, a monitoring method based on stochastic matrix theory (RMT) is proposed, which uses traffic covariance matrix and stochastic matrix theory to compare the prediction results and extract the time-related information of traffic flow from the difference between them. A small number of observation points are arranged in the network, and the maximum eigenvalues of the covariance matrix can accurately obtain the traffic information of the high-speed and low-speed links. The minimum eigenvalue is used to monitor whether the observation point is working or not. The method saves storage resources, calculates time is short, access to more information, is an efficient monitoring method.