一种基于反馈的K-means分簇算法研究

发布时间：2018-06-18 18:01

本文选题：无线传感器网络 + K-means算法　；参考：《信号处理》2017年08期

【摘要】：针对典型的LEACH分簇式路由协议分簇不均匀,簇头节点分布随机导致网络能量消耗大的情况,本文提出一种基于死亡节点数目反馈的K-means分簇算法。首先通过K-means算法划分簇的个数,选择簇的中心节点为该簇的簇头,并通过位置集中性得到集中性较大的若干个节点为主簇头群,其中最大的为主簇头,自此完成初始化。此后用一个受死亡节点数调控的自适应打分函数更新每一轮的簇头和主簇头。主簇头只用于融合并传输数据并不负责感知环境信息。仿真实验结果表明:本算法相较LEACH以及传统的基于K-means的分簇算法,在整个网络的生存时间上分别提高了35%和25%。同时证明:反馈机制的加入和主簇头的选取都有利于网络寿命的提升。
[Abstract]:In this paper, a K-means clustering algorithm based on the number of dead nodes is proposed in this paper. Firstly, the number of clusters is divided by the K-means algorithm, and the center node of the cluster is selected as the cluster head of the cluster, and the location concentration is set through the location concentration. The largest cluster head is the main cluster head group, the largest cluster head is the main cluster head, and then the initialization is completed. After that, an adaptive scoring function controlled by the number of dead nodes is used to update the cluster head and the main cluster head of each round. The cluster head is used only for fusion and transmission of data and is not responsible for the perception of environmental information. Simulation experimental results table Ming: this algorithm is compared with LEACH and the traditional K-means based clustering algorithm. It has been improved by 35% and 25%. in the lifetime of the whole network, respectively. It is proved that the feedback mechanism and the selection of the main cluster head are all beneficial to the improvement of network life.
【作者单位】：安徽大学电子信息工程学院计算智能与信号处理教育部重点实验室;
【基金】：安徽省科技攻关项目(1501b042205)
【分类号】：TN929.5;TP212.9

【相似文献】