| 基于重要事件采样仿真技术的长相关网络业务性能评价 |
| |
| 引用本文： | 金志刚,舒炎泰,王继东.基于重要事件采样仿真技术的长相关网络业务性能评价[J].系统仿真学报,2002,14(1):122-124. |
| |
| 作者姓名： | 金志刚 舒炎泰 王继东 |
| |
| 作者单位： | 天津大学电子信息学院计算机系,天津300072 |
| |
| 基金项目： | 国家自然科学基金的资助(No.69872025)，天津市自然科学基金重点项目资助(No. 993800211) |
| |
| 摘 要： | 论述了重要事件采样仿真技术，分析了其仿真原理和实现方法，给出了典型情形下的仿真算法。对以分数自回归整合滑动平均过程(FARIMA（p，d，q)过程)为输入过程的单服务队列系统的仿真表明的系统溢出率与业务的长相关特性有密切关系。仿真结果与大偏差技术等分析结论较为一致，验证了仿真结果的可靠性。
|
| 关 键 词： | IS仿真 长相关业务 性能评价 |
| 文章编号： | 1004-731X(2002)01-0122-03 |
| 修稿时间： | 2000年11月23 |
Importance Sampling in Performance Evaluation of Long Range Dependence Network Traffic |
| |
| JIN Zhi-gang,SHU Yan-tai,WANG Ji-dong.Importance Sampling in Performance Evaluation of Long Range Dependence Network Traffic[J].Journal of System Simulation,2002,14(1):122-124. |
| |
| Authors: | JIN Zhi-gang SHU Yan-tai WANG Ji-dong |
| |
| Abstract: | This paper proposes the importance sampling technology for performance evaluation of computer networks. The importance sampling theory and the typical algorithm in network simulation are discussed. Given a queue fed by a FARIMA process, the simulation results show that the long range dependence of traffic gives rise to the overflow of queue. The importance sampling significantly reduces the time of simulation. The simulation results are accordance with the analysis result by large deviation technology and other analysis result. |
| |
| Keywords: | importance sampling long range dependence traffic performance evaluation |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
