自相似业务流的合成性能研究

Research on Performance of Aggregating Self-similar Traffic Stream

在分组交换网中对业务流进行多时间尺度的测量时,业务流会产生相似的特性,利用分形点过程模型来分析自相似流是目前较为流行的方法。基于 OPNET 平台,对合成后的自相似流的参数,如分组平均到达速率λ、Hurst 参数 H 和分形开始时间分别进行了讨论,得出了合成后的业务流的统计特性。

Traffic stream displays the similar statistical properties across a wide range of time scales in Packet-Switched Networks. It's a popular way to analyze the self-similar stream with FRP model. It is discussed parameters of the aggregating self-similar traffic streams, including average arrival rate λ, Hurst parameter H and fractal onset time scale(FOTS). The statistics characteristic of traffic stream is also presented.