论文部分内容阅读
针对目前广域网流量多重分形模型不能准确拟合真实网络流量,文中基于瀑布过程建立了一个网络流量多重分形模型,该模型用Hurst指数分布来调节乘数因子,用Haar小波方法产生的序列来拟合随机变量。并通过真实网络流量来检验该模型,检验结果:引入了调节因子和小波序列的模型能够比其他分布更准确地拟合广域网流量。
In view of the fact that the current multi-fractal model of WAN traffic can not accurately fit the real network traffic, a multi-fractal model of network traffic is established based on the waterfall process. The model uses the Hurst exponential distribution to adjust the multiplier factor and the Haar wavelet method Random Variables. The model is verified by real network traffic and the test results: the model with the adjustment factor and the wavelet sequence can fit the WAN traffic more accurately than the other distributions.