基于mallat算法的自相似网络流量随机建模

Stochastic Simulation of Self-similar Network Traffic Based on Mallat Algorithm

传统的网络流量自相似模型均从分形函数入手,计算复杂。文中应用小波变换对流量序列样本进行mallat多分辨分析,并重构各层数据,再随机拆分组合所得重构序列来进行仿真。所生成的仿真流量符合聚合方差估计,具有较好的自相似性。

Traditionally, selfsimilar network traffic is modeled by fractional model, the simulation and the calculation of which is complicated. Instead, this paper introduces traffic modeling by decomposing traffic into wavelet coefficients and scale coefficients with different scale by mallat algorithm and adding up the random parts of the different coefficients into the simulation traffic. The simulation traffic is tested and the selfsimilarity is confirmed.