Abstract: Aiming at the application characteristics of wireless Home Area Networks(HANs), this paper presents a K-means-based clustering algorithm for wireless HANs. The algorithm sets the optimal number of clusters in LEACH as the initial input parameters, and realizes the centralized and on-demand clustering on base station in home area networks, which makes use of Silhouette value to evaluate the optimal clustering. Experimental results show that this algorithm can get good clustering quality in wireless HANs simulation scene.

Key words: evaluation function, cluster, Home Area Networks(HANs), Wireless Sensor Network(WSN)

摘要： 针对无线家域网的应用特点，提出一种基于K-means的无线家域网分簇算法。以LEACH协议中的最优分簇个数作为K-means聚类的输入参数，在家域网基站上实现集中式按需分簇，并利用Silhouette值判定最优的分簇及簇头。实验结果表明，在无线家域网仿真场景中，该算法能获得较好的分簇聚类效果。

关键词: 评价函数, 簇, 家域网, 无线传感器网络

CLC Number: 

• TP301.6