基于改进DoubleTree算法的网络拓扑发现方法  

A Topology Discovery Method Based on the Improved DoubleTree Algorithm

作　　者：乔宏[1] 张大方[1] 曾彬[2] 李明伟[1] 韩健[1] 

机构地区：[1]湖南大学软件学院,湖南长沙410082 [2]湖南大学计算机与通信学院,湖南长沙410082

出　　处：《计算机工程与科学》2010年第4期18-21,38,共5页Computer Engineering & Science

基　　金：国家973计划资助项目(2007CB310702);国家自然科学基金重大研究计划资助项目(90718008);国家自然科学基金资助项目(60673155;60703097)

摘　　要：Double Tree算法是一种重要的IP级网络拓扑发现方法,极大减少了探测冗余,但该方法会产生大量的网络通信量,增加了网络负担。本文在Double Tree算法的基础上,通过对全局停止集存放结构的改进,实现了Double Tree++算法。实验证明,该方法在达到95%的网络覆盖度的同时,减少了50%探测冗余,并且降低了探测源点间的网络通信量。The DoubleTree algorithm is one of the important IP level topology discovery methods. It can reduce probe redundancy dramatically; however, it also brings about a great deal of network traffic, promoting the network burden. In this paper, a DoubleTree＋ ＋ algorithm based on DoubleTree, is designed and implemented via improving the storage structure of the global stop set. The experiment shows that DoubleTree＋ ＋ can reduce 50% probe redundancy and decrease the network traffic among probe sources, while covering 95 % network nodes and links.

关 键 词：网络拓扑 探测冗余 DoubleTree通信量 

分 类 号：TP393.03[自动化与计算机技术—计算机应用技术]