首页 | 本学科首页   官方微博 | 高级检索  
     

准自相似业务流的SPN模型及其评价
引用本文:仝爱军,林闯,徐明伟.准自相似业务流的SPN模型及其评价[J].计算机工程与应用,2002,38(15):55-59.
作者姓名:仝爱军  林闯  徐明伟
作者单位:1. 中国联通有限公司河北分公司,石家庄,050011
2. 清华大学计算机科学与技术系,北京,100084
基金项目:国家自然科学基金(编号:90104002,60173012),国家973重点基础研究发展规划项目(编号:G1999032707),国家863高技术研究发展计划(编号:2001AA112080)
摘    要:该文研究了近似自相似业务的马尔可夫模型,提出了该模型的相位型表达式并据此将模型从离散时间域转换到连续时间域,笔者对一个处理自相似业务到达的简单队列系统进行了研究,比较了系统在实际跟踪的仿真情况下和模型数值分析情况下的性能,模型数值分析用无穷状态随机Petri网实现,研究表明:马尔可夫模型虽能较好地近似自相似性,但仍需进行改进。

关 键 词:无穷状态随机Petri网  自相似  马尔可夫模型  性能评估
文章编号:1002-8331-(2002)15-0055-05
修稿时间:2002年3月1日

Modeling and Evaluation of Pseudo Self-Similar Traffic with Stochastic Petri Nets
Tong Aijun,Lin Chuang,Xu Mingwei.Modeling and Evaluation of Pseudo Self-Similar Traffic with Stochastic Petri Nets[J].Computer Engineering and Applications,2002,38(15):55-59.
Authors:Tong Aijun  Lin Chuang  Xu Mingwei
Affiliation:Tong Aijun 2 Lin Chuang 1 Xu Mingwei 11
Abstract:This paper addresses the suitability of a recently suggested approach for approximating self-similar traffic with a Markovian model.The phase-type nature of the proposed approach is identified and used to transform it from the discrete-time to the continuous-time domain.This paper then investigates the performance of a simple queuing system subject to self-similar arrival traffic,thereby comparing the result of trace-driven simulation with a measured self-simi-lar trace to those derived from a numerical analysis of suggested model.The numerical investigations are performed us-ing a special class of stochastic Petri nets which is particularly suitable for analyzing queuing-model like situations.The result indicates that the suggested Markovian traffic model needs still to be improved,even though the properties of self-similar are well approximated.
Keywords:Infinite-State Stochastic Petri nets(iSPN)  self-similar traffic  Markovian model  performance evaluation
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号