基于距离和分布的无线传感器网络分簇算法  

Clustering algorithm for wireless sensor networks based on distance and distribution

作　　者：廖鹰[1,2] 齐欢[1] 王晓红[1] 李伟群[1] 

机构地区：[1]华中科技大学控制科学与工程系,湖北武汉430074 [2]信息工程大学电子信息工程系,河南郑州450001

出　　处：《华中科技大学学报（自然科学版）》2012年第6期29-33,共5页Journal of Huazhong University of Science and Technology(Natural Science Edition)

基　　金：国家重点基础研究发展计划资助项目(2007CB307100);国家自然科学基金资助项目(60774036);湖北省自然科学基金重点资助项目(2008CDA063);中央高校基本科研业务费专项资金资助项目(C2009Z025Y)

摘　　要：考虑随机分布节点的剩余能量以及节点相对基站的位置,针对基站位置的非均匀无线传感器网络,提出了一种基于节点位置和分布密度的多跳自组织分簇算法.该算法在分簇准备阶段,根据节点分布密度和相对基站的距离确定分簇的半径,均衡分簇能耗;在簇头选举阶段,利用节点的剩余能量和节点连接密度信息,选择最优的节点成为簇头;在分簇建立阶段,限制分簇跳数,有效降低簇内通信量.通过一系列的仿真实验,验证了算法在节点均匀和非均匀分布情况下均能取得较好的性能,建立更为均衡的分簇结构,显著提高网络生存周期.When the residual energy and position of nodes in random distribution considered, a self-or- ganization muti-hop algorithm was proposed for clustering basing on position of base station and con- nection density to generate clusters in wireless sensor networks with inhomogeneous distribution. In cluster preparation process, the algorithm determined the radius of clusters according to node density and relative distance of the base station to balance energy consumption of clusters. The optimal node became cluster-head on the basis of the residual energy and node density in cluster-head election process, and restricted hops of cluster decreased communication traffic effectively in cluster set-up process. The performance of the novel algorithm was illustrated with a series of simulated tests, which indicate that the new algorithm can establish more balanceable clustering structure effectively in uniform and random distribution of nodes, and enhance the network life cycle obviously.

关 键 词：无线传感器网络 非均匀分布 分簇 生存周期 连接密度 

分 类 号：TN915.4[电子电信—通信与信息系统]