基于泊松过程的 ATM 网 VBR 视频业务新模型

A New VBR Video Service Model Based on Poisson Process in ATM Networks

提出了一种新的ＶＢＲ视频业务模型：Ｇａｍｍａ到达的双泊松随机过程．该模型的分形维数小于１，具有长时相关性，能较好地描述具有突发性的实际数据业务，而且具有参数少、形式简单、便于分析等特点．由于其建立在传统的泊松过程上，给进一步对该模型进行排队分析带来了极大的方便．较详细地分析了这一新模型及其参数，得出了一些重要结论．

A new VBR video service model and a double stochastic Poisson process, in which the mean rate of occurrence of events is a Gamma process, are proposed. The fractal dimension is less than one, and it is of long range dependence. It is able to describe better the actual data service coming all of a sudden. It is characterized with fewer parameters, simple form, and convenient for analysis. As it is based on Poisson process, it is very easy to carry out a queuing analysis for the model. The model and its parameters are discussed, and some important conclusions are drawn.