期刊文献+

改进的基于最大似然的快速拓扑估计方法 被引量:1

Improved fast topology estimation method based on maximum likelihood
下载PDF
导出
摘要 基于最大似然的网络拓扑估计方法能够获得全局最优的估计结果,优于一般局部最优化和节点对融合方法,但在网络规模较大时存在计算复杂度较高的缺点。首先证明了网络拓扑估计似然函数是单峰的且峰值为最大值;然后利用单峰特征,改进了现有基于最大似然的拓扑估计方法,降低了计算复杂度。最后,Matlab和NS-2仿真结果证明,在不降低拓扑估计准确率的情况下,改进算法将计算复杂度减少了30%~40%。 Maximum likelihood based topology estimation method can obtain globally optimal estimation resuhs, and its performance is better than that of other methods such as general local optimization and node fusion method. However, when network scale is large, the topology estimation is computationally intensive. To solve this problem, firstly, the authors proved that the network topology estimation likelihood function is a single-peaked function and the peak value is maximum value. Using the single-peaked property of likelihood function, the current maximum likelihood based topology estimation method had been improved. Finally, the results of Matlab and NS-2 simulations show that the improved method cuts computational complexity by 30% -40% without reducing topological estimation accuracy.
出处 《计算机应用》 CSCD 北大核心 2011年第1期212-214,238,共4页 journal of Computer Applications
基金 国家自然科学基金资助项目(60872033) 新世纪优秀人才支持计划资助项目(NECT-07-0148)
关键词 拓扑估计 最大似然 网络层析成像 三明治包 马尔可夫链蒙特卡洛算法 topology estimation maximum likelihood network tomography sandwich packet Markov Chain Monte Carlo (MCMC) algorithm
  • 相关文献

参考文献9

  • 1SIAMWALLA R, SHARM R, KESHAV S. Discovering Internet to-pology [C/OL]// IEEE INFOCOM'99. Washington, DC: IEEE, 1999 [ 2010 -03 -02]. http://citeseerx, ist. psu. edu/viewdoe/ download?doi = 10.1.1.1. 3717&rep = repl &type = pdf.
  • 2CASTRO R, COATES M, LIANG G, et al. Network tomography: recent developments [J]. Statistical Science, 2004, 19(3) : 499 - 517.
  • 3COATES M, HERO A O, NOWAK R, et al. Internet tomography [J]. IEEE Signal Processing Magazine, 2002, 19(3): 47 -65.
  • 4RATNASAMY S, MCCANNE S. Inlerence of multicast routing trees and bottleneck bandwidths using end-to-end measurements [ C]// INFOCOM'99: 18th Annual Joint Conference of the IEEE Computer and Communications Societies. Washington, DC: IEEE, 1999. 353 - 360.
  • 5DUFFIELD N G, HOROWITZ J, PRESTI F L, et al. Muhicast to-pology inference from end-to-end measurements[ EB/OL]. [ 2010 -01 -06]. http://www, ce. uniroma2, it/~ lopresti/Articoli/topologyITC. pdf.
  • 6CASTRO R, COATE M, NOWAK R. Likelihood based hierarchical clustering [ J]. IEEE Transactions on Signal Processing, 2004, 52 (8) : 2308 -2321.
  • 7COATE M, CASTRO R, NOWAK R, et al. Maximum likelihood network topology identification from edge-basod unicast measurements[ C]// Proceedings of the 2002 ACM SIGMETRICS Interna- tional Conference on Measurement and Modeling of Computer Systems. New York: ACM, 2002:11-20.
  • 8SItIN M-F, ItERO A O, III. Hierarchical inference of unicast network topologies based on end-to-end measurements [J]. IEEE Transactions on Signal Processing, 2007, 55(5) : 1708 - 1718.
  • 9马知恩,王锦森.工科数学分析:下册[M].北京:高等教育出版社.2002:55-57.

同被引文献14

引证文献1

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部