一种具有自相似性的分组业务流模型

A Novel Packet Traffic Model with Second-order Self-similarity

以两种状态 Markov链调制的 Poisson过程 (MMPP(2 ) )为原型 ,建立了一种具有自相似性的分组业务流模型 ,并从模型中引出时间尺度的概念。当调节模型参数时 ,该模型既表现长程依赖性 L RD(L ong- Range Dependence) ,又表现出短程依赖性 SRD(Short- Range Dependence)。仿真结果表明 ,这种参数可调模型能更好符合通信网业务的实际情况。

A novel packet traffic model based on the two-state Markovian chain modulating Poisson process (MMPP(2)) is introduced and evaluated. The fluid rate of the model takes on LRD (Long-Range Dependence) characteristic as well as SRD (Short-Range Dependence) characteristic. Time scale is derived out from this model, which stands for the length of the packet in practice. Hurst parameters in the model can also be adjusted, with which second-order self-similarity is examined. The model is well suited for the VBR traffic, and is primarily useful in the present and future networks.