北京邮电大学学报

  • EI核心期刊

北京邮电大学学报 ›› 2012, Vol. 35 ›› Issue (1): 10-14.doi: 10.13190/jbupt.201201.10.wenxx

• 论文 • 上一篇    下一篇

基于支持向量机和云模型的网络健康状态评估

温祥西,孟相如   

  1. 空军工程大学 电讯工程学院, 西安 710077
  • 收稿日期:2011-04-12 修回日期:2011-08-28 出版日期:2012-02-28 发布日期:2012-01-05
  • 通讯作者: 温祥西 E-mail:wxxajy@163.com
  • 作者简介:温祥西(1984-),男,博士生,E-mail:wxxajy@163.com 孟相如(1963-),男,教授,博士生导师
  • 基金资助:

    陕西省自然科学基金项目(SJ08F14,2009JQ8008);空军工程大学创新基金项目(201105)

Network Health Evaluation Based on SVM and Cloud Model

WEN Xiang-xi, MENG Xiang-ru
  

  1. Institute of Telecommunication Engineering, Air Force Engineering University, Xian 710077, China
  • Received:2011-04-12 Revised:2011-08-28 Online:2012-02-28 Published:2012-01-05
  • Contact: Xiang-Xi Wen E-mail:wxxajy@163.com

摘要:

分析了支持向量机分类方法的几何意义,结合健康评估问题的定义提出了一种基于支持向量机的网络健康评估方法.将健康评估转化为分类问题,以点到健康分类边界的距离为评估参数从整体上评价网络健康状况,降低了主观因素对网络评估的影响.采用云模型实现了健康度与评语域之间的不确定性映射,更符合评估的实际情况.基于opnet的仿真实验结果表明了该方法的可行性.

关键词: 网络健康状态评估, 支持向量机, 云模型, 网络仿真

Abstract:

The geometrical meaning of support vector machineSVMis analyzed. Integrated with the definition of health evaluation a method of network health evaluation based on SVM is introduced and translated into a classifying problem. The distance between the sample and health hyperplane is proposed to be the evaluation parameter. The health from the whole entity is evaluated and the subjective factor influences are reduced. The cloud model is used to realize the uncertained map between health degree and comment. The network simulation based on opnet shows feasibilities of this method.

Key words:

中图分类号: