基于小波包分析的网络流量随机模拟  

Stochastic simulation of network traffic based on wavelet packet analysis

作　　者：刘渊[1] 马汝辉[1] 

机构地区：[1]江南大学信息工程学院,江苏无锡214122

出　　处：《计算机工程与设计》2008年第10期2439-2440,2443,共3页Computer Engineering and Design

基　　金：国家预研基金项目(A1420061266)

摘　　要：提出基于小波包分析的网络流量随机模拟算法,通过应用小波包变换将实际测量的网络流量分解为2个不同分量(n为分解的层数),再将这些分解长度相等的每个分量进行等分,进行这些等分分量的随机组合形成2个分量,再重构,最后对模拟合成的网络流量进行自相似特性的分析。实验结果表明,用该方法模拟的网络流量具有较好的自相似性。A simulation algorithm of network traffic based on wavelet packet analysis is presented, the network traffic collected from real network is decomposed into different 2^n variable based on wavelet packet （n is the decomposed layer）. The recomposed 2^n variable is randomly obtained from lots of the same length variable, then are decomposed from the original 2^n variable. The recomposed 2^n variable is analyzed by self similarity. Statistical analysis shows that the network traffic generated by this way possesses high degree of self similaxity.

关 键 词：小波包 小波变换 网络流量 自相似参数 自相似性 

分 类 号：TP393.01[自动化与计算机技术—计算机应用技术]