网络随机接入ON-OFF重尾流的准入控制研究  

Research on Call Admission Control of Stochastic Accessing ON-OFF Heavy Tailed Flows

作　　者：姚正林[1] 刘金刚[1] 

机构地区：[1]中国科学院计算技术研究所-首都师范大学计算机科学联合研究院,北京100081

出　　处：《计算机工程》2005年第10期13-15,21,共4页Computer Engineering

摘　　要：近年来的许多研究表明,随着网络带宽的增大,业务量不断增多,网络中的数据流呈现出自相似性,具有很强的长相关特点,这就使得传统的基于短相关的Markov流量分析方法不再适用,该文对渐进自相似流进行了分析,在分析了系统输入固定数量叠加的ON-OFF重尾间隔流排队模型基础之上,提出了随机接入重尾数据流的准入控制算法,并进行了仿真分析。In recent years, lots of research have shown that flows in the network are the characteristic of self-similarity with bandwidth extending and services increasing. The self-similar traffic is long-range dependence. So, the traditional method of analyzing network performance based on Markov short-range dependent flows is not available. In this paper, the performance of asymptotic self-similar traffic has been analyzed. A call admission control (CAC) algorithm of stochastic number accessing heavy tail distribution interval ON-OFF flow has been proposed based on certain number of it. The model has been tested in simulating experiment, and has been proved available.

关 键 词：自相似 重尾分布 随机接入 准入控制 

分 类 号：TP393.03[自动化与计算机技术—计算机应用技术]