利用重要性贡献矩阵确定通信网中最重要节点  

Finding most vital node by node importance contribution matrix in communication netwoks

作　　者：赵毅寰[1] 王祖林[1] 郑晶[1] 郭旭静[1] 

机构地区：[1]北京航空航天大学电子信息工程学院,北京100191

出　　处：《北京航空航天大学学报》2009年第9期1076-1079,共4页Journal of Beijing University of Aeronautics and Astronautics

基　　金："十一五"国防预研基金资助项目(41101040603)

摘　　要：针对已有的节点重要性评价方法,基于节点的移除可能导致网络拓扑结构变化的不足,提出了一种利用节点间关联特性的通信网络节点重要性评价方法.该方法定义的节点重要性贡献矩阵(NICM,Node Importance Contribution Matrix)考虑了网络中不同节点间的联接关系对节点重要性的影响,每个节点对其相邻节点重要程度的贡献与该节点的度有关,节点的初始重要性设为该节点的介数.利用该算法对典型网络的节点重要性进行了分析,并且与移除节点的方法进行了对比.实验结果表明,所提出的算法无须考虑网络拓扑结构的变化,结果精确并能正确评价节点对网络资源的控制能力.Node importance evaluation based on removal of nodes and their incident links may change the topology of the network, even splits the network into several disconnected parts. The neighbor nodes＇ node importance contribution matrix （NICM） was defined and an improved node importance evaluation method was proposed. A node＇s importance value was initialized as its betweenness. Different node contributed importance value dissimilarly to a certain node, according to its degree and their position relationship. The algorithm was compared with the node-removal based method in the analysis of advanced research project agency （ARPA） net topology. The results show that this algorithm accurately represents the node importance, with less calculation workload and no change of the net topology. Nodes＇ ability to control the network recourse is also involved in this method.

关 键 词：通信网络 最重要节点 度 介数 拓扑 

分 类 号：TN915.02[电子电信—通信与信息系统]