随机延迟污染对IP网络坐标系统的影响及其抑制方法研究  

Influence of stochastic delay contaminations on IP network coordinate system and contamination mitigation methods

作　　者：张敏[1,2] 周亮[2] 黄琼[3] 阳小龙[1] 隆克平[1] 

机构地区：[1]北京科技大学计算机与通信工程学院,北京100083 [2]电子科技大学通信与信息工程学院,四川成都611731 [3]重庆邮电大学通信与信息工程学院,重庆400065

出　　处：《通信学报》2011年第12期29-35,共7页Journal on Communications

基　　金：国家重点基础研究发展计划("973"计划)基金资助项目(2007CB310706;2012CB315905);国家自然科学基金资助项目(60873263;60932005;61172048;61100184);四川省青年基金资助项目(09ZQ026-032);教育部新世纪优秀人才计划(NCET-09-0268)~~

摘　　要：针对实际网络中存在各种随机延迟污染严重影响IP网络坐标系统的时延预测性能和稳定性问题,从侧重于维持实测时延统计特征无损失和增强抑制时延随机波动2个角度,分别提出t-Filter和Δ-Filter 2种随机延迟污染抑制方法。通过PlanetLab数据集的仿真结果表明:与MP-Filter相比,它们能更有效地提高IP网络坐标系统的时延预测准确性、系统稳定性和收敛性。There are various stochastic delay contaminations in the IP networks, which have a great impact on the performance of IP network coordinates system in terms of the delay prediction accuracy and the coordinate stability. Here, from the perspective of keeping the statistical characteristic of delay without a loss and suppressing the abnormal delay jitter, two contamination mitigation methods were proposed respectively, i.e., t-Filter and A-Filter. Through the PlanetLab datasets, compared with MP-Filter, the simulation results show that both t-Filter and A-Filter can enhance the delay prediction accuracy, the system stability and the system convergence in the more effective methods.

关 键 词：IP网络坐标 网络距离 T检验 时延梯度 

分 类 号：TP393.01[自动化与计算机技术—计算机应用技术]