基于 K-means＋＋的无线传感网分簇算法研究

摘要

For the problems of uneven clustering,cluster head election unreasonable and single form of data transmission in traditional hierarchical algorithm,this paper proposed an improved wireless sensor networks clustering alogrithm LEACH-KPP based on K-means++.Firstly,it used K-means++algorithm divided the cluster evenly in the cluster stage,then used im-proved cluster head election function selected cluster head in cluster head election stage,at last used hybrid transmission me-thod transmitted data by single-hop and multi-hop dynamically according to the distance between cluster and cluster head, cluster heads in the stage of fusion data transmission.OMNet++simulation result and derivation of time complexity shows that LEACH-KPP extends the lifetime of network and is better than traditional hierarchical algorithm in the residual energy and alive nodes number.%针对传统分层路由算法存在的分簇不均匀、簇头选举不合理以及数据传输形式单一等问题，提出基于K-means＋＋的无线传感网改进分簇算法LEACH-KPP。首先在成簇阶段采用K-means＋＋算法实现均匀分簇，随后在簇头选举阶段使用改进簇头选举函数选取簇头，最后在融合数据传输阶段根据簇头与基站、簇头与簇头之间距离动态选择单跳与多跳的混合传输方式传输数据。OMNet＋＋仿真结果与时间复杂度推导表明，LEACH-KPP延长了网络的生存周期，在节点剩余能量与后期存活数目上都优于传统分层路由算法。