基于Alpha稳定自相似过程的网络业务流建模

On the Modeling of Network Traffic Based Alpha-stable Self-similar Processes

针对实际网络传输中存在的自相似现象,介绍了自相似的含义,指出了采用al-pha稳定过程作为建模理论的理由,给出了alpha稳定过程及线性分形稳定噪声(LFSN)过程的定义和相关性质,提出了一种新的网络业务流模型,并对新模型进行了仿真分析比较,证实新模型在描述突发性方面优于其他的自相似网络模型,因此新模型能更好地描述实际网络传输的主要特征。

With self-similar phenomenon emerging in real network traffic, this paper introduces the concept about self-similarity. And then the paper points out the reason why uses alpha-stable processes as the basic of modeling, at the same time indicates the definitions and properties of the alpha-stable process and the Linear Fractional Stable Noise (LFSN) process. Based on above states the paper advances a new network traffic model. By the simulation experiments, it is shown that the new model is better than any other self-similar models. So the new network model not only captures both the burstiness and the long range dependence of real network, but also has great flexibility.