自相似网络业务流建模方法与比较

随着网络服务的发展．多媒体数据逐渐成为网络业务流的主要内容。基于多媒体数据的网络业务流对服务质量(QoS)提出了更高的要求。通常对服务质量的要求必然转化为对网络资源的需求，然而这个需求的量化必须基于一个能够准确描述网络变化的网络业务流数学模型。同时，网络业务流的不同特征以及研究目的的不同使得自相似网络业务流建模成为一个挑战性的问题。近来，一系列的方法被用于自相似网络业务流建模，本文的作者将这些方法分为两类共四种方法。本文全面分析比较了这些方法，并提出一种分类机制用于根据不同的研究需要确定相应的建模方法。

Self-Similar Network Traffic Modeling Approaches and Comparability

With the diversification of network service, multimedia data have been gradually becoming the main content in network traffics. The network services based on multimedia data put forward higher requirement for QoS (Quality of Service). In general, the requirement for QoS must be transformed into the need of network resources, whereas the quantification of this need is based on a mathematical model of network traffic which can exactly describe changes of network traffic- Furthermore, the different features of network traffic and purposes of research have been made self-similar network traffic modeling a challenging research question. Recently, there have been a number of approaches applied to self-similar network traffic modeling. The authors of this article classify these approaches into two categories and four approaches. A thorough analysis and comparison among these approaches is discussed in this paper- Furthermore, we propose a novel scheme to determine an appropriate modeling approach to satisfy the requirement of a specific research.