基于节点聚合度的生成簇算法  

Clustering algorithm based on convergence degree

作　　者：杜胜永[1] 柴乔林[1] 王华[1] 

机构地区：[1]山东大学计算机与科学技术学院,山东济南250061

出　　处：《计算机应用》2006年第4期948-950,共3页journal of Computer Applications

摘　　要：无线传感器网络可以用无向图表示,节点及其一跳邻节点构成一个子图。子图节点总数与边总数的比值作为该节点的聚合度,选择子图中聚合度最大的节点作为簇头,建立网络逻辑簇结构。与传统的最小ID算法和最大连通度算法相比,仿真结果表明该算法建立簇之间的重叠度最低。Wireless sensor network could be described as an undirected graph. The convergence degree of the node was determined by the ratio of the number of sub-graph＇s vertexes and number of the edges. The node with the greatest convergence degree was selected as the cluster head to build up the logical structure of the network. The experiment results show that the proposed method yields the lowest overlap of the clusters comparing with the LID algorithm and the MAXD algorithm.

关 键 词：无线传感器网络 节点聚合度 网络重叠度 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]