高阶间隔估计算法在网络流量监测中的应用研究

Research on High Order Interval Estimation Algorithm in Network Traffic Monitoring

当前的网络流量监测方法比如报文抽样和时间间隔抽样,存在资源消耗大和精度低2个方面的问题,本文提出了一种高带宽及高传输速率情况下的流量监测方法,即以均匀分组采样方法为理论的高阶间隔估计算法﹒首先根据网络流量的动态变化,实时调整采样次数;再利用低阶采样数据及优化算法进行高阶估计;最后采用相对熵进行采样性能分析与估计﹒仿真结果表明,随着高阶间隔的逐渐增大,流量监测的精度相比报文采样和时间间隔采样得到逐步提高,而资源消耗却显著减少﹒

With the gradual increase of the network bandwidth and transmission speed the current network traffic monitoring methods such as packet sampling and time interval samplings,there are two problems in resource consumption and low precision.In this paper a method of high bandwidth and high transmission rate is proposed,that is a high order interval estimation algorithm based on uniform sampling method.According to the dynamic change of the network traffic,the sampling time is adjusted in real time.High order estimation using low order sampled data and optimization algorithm.Finally,the relative entropy is used to analyze and estimate the sampling performance.The simulation results show that with the increasemen of the high order interval,the accuracy of traffic monitoring is gradually improved compared with the sampling and interval sampling,while the resource consumption is significantly reduced.