【作者】 杨志豪； 胡光岷；

【Author】 Yang Zhihao~1 Hu Guangmin~2 School of communication and information engineering,UESTC

【机构】 成都电子科技大学宽带与光纤重点实验室；

【摘要】 针对现有网络拓扑识别方法在中间节点协作问题上的两难选择,本文提出一种局部协作条件下的有限混合模型估计方法,并用于网络拓扑识别。该方法利用与traceroute协作度中间节点的ICMP报文信息,直接确定有限混合模型中的未知变量,可以有效减少某些聚类团之间的计算,获得准确的网络拓扑结构。仿真实验表明本文的算法可以在部分中间节点不协作的情况下,获得完整的网络拓扑结构,同时相对于经典的网络层析成像方法可以减少探测包的发送数量,提高算法的精度和稳定性。更多还原

【Abstract】 In this paper we proposed a new network estimation algorithm using Finite Mixture Model with Partial Association due to the tradeoffbetween the association of the internal routes and the complex of the Tomography methods.The algorithm which we proposed uses the ICMP information to calculate the unknown variables in the model and reduce some computation between clusters effectively to form the topology.The simulation results show that our algorithm can get the whole topology when a part of the internal nodes don’t associated,and can reduce the number of probe packets,improve the accuracy and stability of the algorithm to the traditional tomography methods.更多还原