基于信号强度指纹的无线定位算法改进与性能评估  

Improvement and performance evaluation of wireless localization algorithm based on signal strength fingerprinting

作　　者：甘新 冯准 王秀红 GAN Xin;FENG Zhun;WANG Xiuhong(Guigang High and New Technology Application Research Institute,Guigang 537100,China)

机构地区：[1]贵港市高新技术应用研究所,广西贵港537100

出　　处：《中国高新科技》2024年第24期11-13,共3页

基　　金：广西贵港市科技计划项目《无线高精度远距离室外畜牧业定位系统设计》(贵科攻2300045)。

摘　　要：文章设计并优化了基于无线信号的室内定位系统,旨在解决室内环境下的精准定位问题。离线训练阶段采用了信号强度采集、数据处理提高了位置指纹库的准确性。在线定位阶段通过K均值聚类和基于差值的加权K近邻法,实现了实时位置估计。实验结果表明,新算法相较传统方法具有更高的精度和更低的定位误差,为室内定位技术的进一步发展和应用提供了有力支持。This study designs and optimizes an indoor localization system based on wireless signals to address the challenge of precise indoor positioning.In the offine training phase,signal strength collection and data processing are employed to enhance the accuracy of the location fingerprint database.During the online localization phase,real-time position estimation is achieved through K-means clustering and a weighted K-nearest neighbor algorithm based on differences.Experimental results demonstrate that the new algorithm outperforms traditional methods with higher accuracy and lower positioning errors,providing robust support for further development and application of indoor localization technology.

关 键 词：室内定位系统 无线信号 位置指纹库 K均值聚类 

分 类 号：TN925[电子电信—通信与信息系统]