首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于网络演算的有损分形整形器及性能模型
引用本文:张连明.基于网络演算的有损分形整形器及性能模型[J].华南理工大学学报(自然科学版),2009,37(5).
作者姓名:张连明
作者单位:湖南师范大学;华南理工大学
摘    要:随着Internet的不断演化和发展,它已成为一个开放的复杂巨系统,为改善Internet复杂网络流量自相似等突发特征给网络性能带来的影响,提出了一种适合自相似等突发流量的有损分形整形器的一般数学模型,证明了分形到达约束受限于原始到达曲线,推导并得出了有损分形整形器的输出特性与输入流量的自相似特征参数以及分形整形曲线之间的关系,分析了有损分形整形器的队列长度与延迟等性能,讨论了有损分形整形器的引入对网络端到端延迟、数据丢失总数以及平均丢失率等性能的影响,分别得到了上述各性能指标上界与流量自相似参数的关系,给出了在允许最大平均丢失率的情况下的最小缓存容量.这对于基于分形整形器的拥塞控制策略和流量控制方案的评价,以及分形整形器自身参数的设计是具有参考作用.

关 键 词:自相似流量  有损分形整形器  确定性网络演算  性能模型  上界  
收稿时间:2008-7-10
修稿时间:2008-9-28

Research on Performance and Model of Loss Fractal Regulators in Network Calculus
Abstract:Internet has become an open giant complex system with the evolution and development of Internet. In order to improve the influence of bursty characteristic of self-similar traffic on Internet complex network on network performance, a general mathematical model of the loss fractal regulator for controlling self-similar traffic is proposed, and it was proved that the fractal arrival constraints are restricted to the original arrival curves. The relationships between the output properties of the loss fractal regulator and the self-similar parameter of the input traffic and the shaper curve and the shapers are derived. The performance of the loss fractal regulator is analyzed, such as, queue length and delay. The influences of the introduction of the loss fractal regulator on the end-to-end delay and the packets loss total and the average loss rate are discussed. The correlation of the bounds on the above performance metrics and the parameter of self-similar traffic were given. At last, the buffer minimum in the loss fractal regulator was obtained under the maximum value of the average loss rate. These works and results of the analyses have the practical significance for the evaluation of congestion control scheme and traffic management scheme of self-similar traffic on Internet complex network and the configuration of the parameter of the fractal regulator parameter.
Keywords:self-similar traffic  loss fractal regulator  deterministic network calculus  performance model  upper bound
点击此处可从《华南理工大学学报(自然科学版)》浏览原始摘要信息
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