摘要
为了对网络进行有效的测量和管理,对流量模型的研究是非常重要的。论文研究了相关的流量模型,提出了一种基于回归方程方法的流量模型,证明了该方法的有效性,获得了最终仿真结果。该模型的核心有三点:(1)通过主动测量的方法获得路径流量和流量分配概率;(2)通过流量和流量分配概率计算结点对流量;(3)推导了回归方程。
In order to manage the whole network,a very important method is traffic model.In this article,the authors survey the recent advances in the research of traffic model and introduce some of the typical algorithms.At last,a novel traffic model is proposed based on the active probe method.First,we can get time interval by inject probe packet into the network.Second,we can acquire regress formula based on M/M/1 queuing model.Last,we get traffic model by using the Initial Gap Increasing(IGI)algorithm to acquire traffic.Our experiments show that measurements that allow us to estimate the competing traffic to within about 30%.
出处
《计算机工程与应用》
CSCD
北大核心
2005年第9期9-12,共4页
Computer Engineering and Applications
基金
重庆市自然科学基金资助(CSTC2004BB2195)
关键词
流量模型
流量矩阵
主动测量
IPMP
回归方程
traffic model,traffic matrix,active measurement,IP measurement protocol,regress formula
