| 一种IP网络拥塞链路丢包率范围推断算法 |
| |
| 引用本文： | 陈宇,周巍,段哲民,钱叶魁,赵鑫.一种IP网络拥塞链路丢包率范围推断算法[J].软件学报,2017,28(5):1296-1314. |
| |
| 作者姓名： | 陈宇 周巍 段哲民 钱叶魁 赵鑫 |
| |
| 作者单位： | 西北工业大学 电子信息学院, 陕西 西安 710072;郑州航空工业管理学院, 河南 郑州 450015,西北工业大学 电子信息学院, 陕西 西安 710072,西北工业大学 电子信息学院, 陕西 西安 710072,解放军防空兵学院, 河南 郑州 450052,解放军防空兵学院, 河南 郑州 450052;通信网信息传输与分发技术重点实验室, 河北 石家庄 050000 |
| |
| 基金项目： | 国家重点基础研究发展计划（973计划）（2013CB329104）；国家自然科学基金（61103225）；通信网信息传输与分发技术重点实验室基金 |
| |
| 摘 要： | 针对大规模IP网络拥塞链路丢包率范围推断算法中存在的不足，提出一种贪婪启发式拥塞链路丢包率范围推断算法.借助多时隙路径探测，避开单时隙探测对时钟同步的强依赖；通过学习各链路拥塞先验概率，借助贝叶斯最大后验定位拥塞链路；提出聚类拥塞链路相关、性能相近路径集合的策略，通过对聚类路径集合中性能相似系数求解，循环推断拥塞链路丢包率范围.实验验证了本文提出算法的准确性及鲁棒性.
|
| 关 键 词： | IP 网络 拥塞链路推断 丢包率范围 贝叶斯最大后验概率 贪婪启发算法 |
| 收稿时间： | 2016/6/16 0:00:00 |
| 修稿时间： | 2016/7/28 0:00:00 |
Congested Link Loss Rate Range Inference Algorithm in IP Network |
| |
| CHEN Yu,ZHOU Wei,DUAN Zhe-Min,QIAN Ye-Kui and ZHAO Xin.Congested Link Loss Rate Range Inference Algorithm in IP Network[J].Journal of Software,2017,28(5):1296-1314. |
| |
| Authors: | CHEN Yu ZHOU Wei DUAN Zhe-Min QIAN Ye-Kui and ZHAO Xin |
| |
| Affiliation: | Institute of Electronic Information, Northwestern Polytechnical University, Xi''an 710072, China;Zhengzhou University of Aeronautics, Zhengzhou 450015, China,Institute of Electronic Information, Northwestern Polytechnical University, Xi''an 710072, China,Institute of Electronic Information, Northwestern Polytechnical University, Xi''an 710072, China,Air Defence Forces Academy of PLA, Zhengzhou 450052, China and Air Defence Forces Academy of PLA, Zhengzhou 450052, China;Science and Technology on Information Transmission and Dissemination in Communication Networks Laboratory, Shijiazhuang 050000, China |
| |
| Abstract: | Aiming at the shortcomings of congested link loss rate range inference algorithm in large scale IP network, a kind of congested link loss rate range inference algorithm based on the greedy heuristic method was proposed.The strong dependence on the clock synchronization of single slot E2E path measurements was avoided through using multiple slots E2E path measurements.Each congested link can be located through adopting the link congestion Bayesian Maximum A-Posterior (BMAP) after learning congestion prior probabilities of the links.The set consisting of paths with related congested links and similar performance was constructed.Through solving the performance similar coefficient dynamically, loss rate range of each congested link can be circularly inferred.The accuracy and robustness of the algorithm proposed in this paper was verified by experiments. |
| |
| Keywords: | IP network congested link inference loss rate range Bayesian Maximum A-Posteriori (BMAP) greed heuristic algorithm |
|
| 点击此处可从《软件学报》浏览原始摘要信息 |
|
点击此处可从《软件学报》下载全文 |
|
