摘要
针对传统聚类算法仅依靠信号强度进行判别以及K值与聚类联系不大的问题,提出了改进聚类与加权K临近算法,对Wi-Fi室内定位问题进行研究。从离线阶段和在线阶段出发,分析基于位置指纹定位的一般原理。为了保证一个聚类里的采样点能够在地理空间上也相邻,对样本点进行聚类,以聚类结果为基础,通过使用信号强度与实际建筑环境相结合的聚类算法进行聚类,然后根据每个聚类的特性来确定每一个聚类的K临近算法的K值,实现Wi-Fi室内定位。实验结果表明,改进的聚类与加权K临近算法比起传统算法的定位精度更高。
Aiming at the problem that traditional clustering algorithm only relies on signal strength to distinguish and the K value has little connection with clustering,an improved clustering and weighted K proximity algorithm is proposed to study the problem of Wi-Fi indoor positioning.Starting from the offline stage and the online stage,the general principle of location-based fingerprint positioning was analyzed.In order to ensure that the sampling points in a cluster were also adjacent in geographic space,the sample points were clustered.Based on the clustering results,clustering was performed by using a clustering algorithm that combines signal strength with the actual building environment.Then the K value of each cluster's K proximity algorithm was determined according to the characteristics of each cluster to realize Wi-Fi indoor positioning.Experimental results show that the improved clustering and weighted K proximity algorithm has higher positioning accuracy than traditional algorithms.
作者
叶瀚云
何军
YE Han-yun;HE Jun(College of Computer Science,Sichuan University,Chengdu Sichuan 610065,China)
出处
《计算机仿真》
北大核心
2023年第5期408-412,共5页
Computer Simulation
基金
四川省科技重点研发项目(18ZDYF2039)
四川省重大科技专项(2017GZDZX0002)。
关键词
室内定位
聚类算法
位置指纹
Indoor positioning
Clustering algorithm
Locationfingerprint
