[1]刘卫江,龚俭,丁伟,等.基于最小二乘法的流长度分布估计方法[J].东南大学学报(自然科学版),2006,36(3):467-471.[doi:10.3969/j.issn.1001-0505.2006.03.028]
 Liu Weijiang,Gong Jian,Ding Wei,et al.Method for estimation of flow length distributions based on least square method[J].Journal of Southeast University (Natural Science Edition),2006,36(3):467-471.[doi:10.3969/j.issn.1001-0505.2006.03.028]
点击复制

基于最小二乘法的流长度分布估计方法()
分享到:

《东南大学学报(自然科学版)》[ISSN:1001-0505/CN:32-1178/N]

卷:
36
期数:
2006年第3期
页码:
467-471
栏目:
计算机科学与工程
出版日期:
2006-05-20

文章信息/Info

Title:
Method for estimation of flow length distributions based on least square method
作者:
刘卫江123 龚俭23 丁伟23 程光23
1 东南大学计算机科学与技术学科博士后流动站, 南京 210096; 2 东南大学计算机科学与工程学院, 南京 210096; 3 江苏省计算机网络技术重点实验室, 南京 210096
Author(s):
Liu Weijiang123 Gong Jian23 Ding Wei23 Cheng Guang23
1 Post Doctoral Station for Computer Science and Technology, Southeast University, Nanjing 210096, China
2 School of Computer Science and Engineering, Southeast University, Nanjing 210096, China
3 Key Laboratory of
关键词:
抽样报文 IP流 概率 最小二乘法
Keywords:
packet sampling IP flows probability least square method
分类号:
TP393
DOI:
10.3969/j.issn.1001-0505.2006.03.028
摘要:
为了得到未抽样流的分布特征, 提出一种新的由抽样报文流数据来估计原始未抽样流长度分布的方法.首先分析了产生一个定长抽样流的原始流的概率分布模型,并根据这个概率分布特征给出了长流一个非常简单的估计.然后构造了关于短流的方程组,利用流的重尾分布特性和最小二乘法对该方程组进行求解,得到了短流的估计.理论分析表明该估计方法有效地控制了时间复杂程度,实验测试结果也表明该算法对于分布的估计是精确的,估计精度与EM算法相当.
Abstract:
A novel method for estimation of original flow length distributions from sampled flow statistics is proposed to obtain the distribution feature of unsampled flows. First, the probability distribution model of original flow for a sampled flow of fixed length is analyzed, and simple estimation for large flows is described according to the analysis result. Then, estimation for short flows is obtained by constructing equations involving short flows and solving them using the heavy-tailed feature of flow and the least square method. The theoretical analysis shows that the computational complexity of this method is well under control, and the experimental results demonstrate that the distributions inferred from the proposed method are as accurate as those from the expectation maximum(EM)algorithm.

参考文献/References:

[1] Roberts J W.Traffic theory and the Internet [J]. IEEE Communications Magazine,2001,39(1):94-99.
[2] Claffy K C,Polyzos G C,Braun H W.Application of sampling methodologies to network traffic characterization [C] //Proc of ACM SIGCOMM ’93.New York:ACM Press,1993:194-203.
[3] IETF.Packet sampling(psamp)[EB/OL].(2005-02-02)[2005-06-30].http://www.ietf.org/html.charters/psamp-charter.html.
[4] Duffield N G,Grossglauser M.Trajectory sampling for direct traffic observation [J].IEEE/ACM Trans on Networking,2001,9(3):280-292.
[5] Duffield N G,Grossglauser M.Trajectory sampling with unreliable reporting [C] // IEEE INFOCOM 2004.Hong Kong,2004:1570-1581.
[6] Cisco.Sampled Cisco [EB/OL].(2002-12)[2005-06-30].http://www.cisco.com/en/US/products/sw/iosswrel/ps1829/products_feature_guide 09186a 0080081201.html.
[7] Brownlee Nevil.NeTraMet Version 4.4 [EB/OL].(2002-12)[2005-06-30].http://www2.auckland.ac.nz/net/Accounting/ntm.Release.note.html.
[8] Estan C,Varghese G.New directions in traffic measurement and accounting:focusing on the elephants,ignoring the mice [J]. ACM Transactions on Computer Systems,2003,21(3):270-313.
[9] Hohn N,Veitch D.Inverting sampled traffic [C] //Internet Measurement Conference 2003.New York:ACM Press,2003:222-233.
[10] Duffield N G,Lund C,Thorup M.Properties and prediction of flow statistics from sampled packet streams [C] // Proc of ACM SIGCOMM Internet Measurement Workshop 2002.New York:ACM Press,2002:159-171.
[11] Duffield N G,Lund C,Thorup M.Estimating flow distributions from sampled flow statistics [J]. IEEE/ACM Transactions on Networking,2005,13(5):325-336.
[12] Claffy K C,Braun H W,Polyzos G C.A parameterizable methodology for internet traffic flow profiling [J].IEEE Journal on Selected Areas in Communications,1995,13(8):1481-1494.
[13] 陆璇.应用统计[M].北京:清华大学出版社,1999:80-89.

备注/Memo

备注/Memo:
基金项目: 国家重点基础研究发展计划(973计划)资助项目(2003CB314803)、教育部科学技术研究重点资助项目(105084)、 江苏省网络与信息安全重点实验室资助项目(BM2003201)、江苏省博士后科研资助计划资助项目.
作者简介: 刘卫江(1969—),男,博士,教授, wjliu@njnet.edu.cn.
更新日期/Last Update: 2006-05-20