自相似流量下自适应AQM算法性能研究

Research on Performance of Adaptive AQM Algorithm of Self-Similar Traffic

针对目前网络流量存在的自相似特性,提出了一种自相似流量下的主动队列管理算法——IARED算法。该算法首先根据网络流量的自相似和长相关特性,利用自相似流量的自相关函数来设置平均队列长度计算公式的权值,再根据当前平均队列长度与目标队列长度变化率和当前平均队列长度与上一时刻平均队列长度变化率两个参数来动态调整最大包丢弃概率。仿真实验结果表明,该算法能很好地适应自相似网络的流量变化,可以很好地控制队列长度,降低丢包率和保持较低的排队延迟。

Aimed at the self - similar characteristic of network traffic, an improved adaptive AQM algorithm for self - similar traffic, namely IARED, is proposed. Based on the self- similarity and long- range dependence of the network traffic, the algorithm uses the autocorrelation function to set the average queue length and dynamically adjust the maximum packet dropping/marking probability acoording to two parameters, namely the change ratio of the current average queue length versus target queue length and change ratio of eurrmt av- erage queue length versus last average queue length. The simulation results show that the improved algorithm can accomrrzxtate the change of the self- similar traffic and control the queue length very well, so as to decrease the packet loss rate and keep low queue delay.