小波技术在网络流量特性分析中的应用

The Applications of Wavelet Technique in Network Traffic Analysis

网络流量的特性分析一直是通信网络性能分析的一个极其重要的问题。小波变换能将在时域上高度复杂相关的信号分解成小波域上近似彼此不相关的过程,从而使得在时域上分析非常困难的问题,在小波域上变得简单。基于小波技术的网络流量特性分析,将网络流量过程分解成不同尺度下的小波系数(细节)和尺度系数(背景),从而刻画出网络流量的自相似特性和多分形特性。

The characteristic analysis of network traffic is an important issue in communication network. Due to wavelet transform, highly correlated signals in time domain become nearly uncorrelated in wavelet domain. Through analyzing the scaling coefficient (detail) and wavelet coefficient (background), we can capture the characteristic of self-similar and multifractal of network traffic.