首页 | 本学科首页   官方微博 | 高级检索  
     

基于延迟参数的可用带宽主动估测算法
引用本文:卞静,王泽强,张光昭.基于延迟参数的可用带宽主动估测算法[J].电子与信息学报,2006,28(10):1926-1929.
作者姓名:卞静  王泽强  张光昭
作者单位:中山大学科学计算与计算机应用系,广州,510275;中山大学电子系,广州,510275
基金项目:国家自然科学基金;广东省自然科学基金
摘    要:可用带宽是重要的网络资源,但它的实际测量存在许多困难。该文基于延迟趋势模型,提出一种端到端的可用带宽估测算法SSP(SRdeg and SRdet Parameters),设计了反映网络传输性能的参数SRdeg 和SRdet ,通过端系统对参数进行计算,主动调整发送速率,从而快速估测到网络的可用带宽,达到有效利用网络资源的目的。在NS2平台进行了参数和SSP算法的性能测试实验,并与当前的带宽估测工具Pathload进行了比较。实验结果表明,SSP算法估测带宽的精度和效率都高于Pathload。

关 键 词:可用带宽估测  延迟趋势模型  Pathload  参数SRdeg和SRdet
文章编号:1009-5896(2006)10-1926-04
收稿时间:2005-11-16
修稿时间:2006-05-08

An Active Estimation Algorithm for Available Bandwidth Based on the Delay Parameters
Bian Jing,Wang Ze-qiang,Zhang Guang-zhao.An Active Estimation Algorithm for Available Bandwidth Based on the Delay Parameters[J].Journal of Electronics & Information Technology,2006,28(10):1926-1929.
Authors:Bian Jing  Wang Ze-qiang  Zhang Guang-zhao
Affiliation:Department of Scientific Computation and Computer Application, Zhongshan University, Guangzhou 510275, China; Department of Electronics, Zhongshan University, Guangzhou 510275, China
Abstract:Available bandwidth is one of the important network resources, but it difficult to measure in practice. In this paper, SSP( SR degand SR detparameters), an active end-to-end available bandwidth estimation algorithm based on delay trend model, is proposed. Two parameters SR degand SR detare designed which can indicate the network transmission performance changes. By calculating these parameters, the end system can adjust the transmission rate so that the available bandwidth can be estimated quickly as well as the network resources can be effectively used. The parameters are tested by NS2 simulation platform. The performance of SSP algorithm is compared with Pathload, a currently used tool on available bandwidth estimation. The simulation experiments show that the proposed algorithm SSP is more efficient and accurate than that of Pathload.
Keywords:Pathload
本文献已被 CNKI 维普 万方数据 等数据库收录!
点击此处可从《电子与信息学报》浏览原始摘要信息
点击此处可从《电子与信息学报》下载全文
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