基于高斯过程的无线信号指纹定位算法  

FINGERPRINT LOCALIZATION ALGORITHM FOR WIRELESS SIGNAL BASED ON GAUSSIAN PROCESS

作　　者：韦春玲[1] 王步飞 Wei Chunling;Wang Bufei(School of Physics and Electronic Information,Huanggang Normal University,Huanggang 438000,Hubei,China)

机构地区：[1]黄冈师范学院物理与电子信息学院,湖北黄冈438000

出　　处：《计算机应用与软件》2019年第5期91-95,125,共6页Computer Applications and Software

基　　金：湖北省教育厅科研项目(B2018189);黄冈师范学院高级别培育项目(201812083);湖北省电工电子示范中心科研项目(zx201962);湖北省虚拟仿真示范中心科研项目(zj201928)

摘　　要：基于Wi-Fi接收信号强度的室内定位技术因其良好的定位性能获得广泛关注。为了提高定位精度,需要构建高密度信号指纹数据库。但是创建与维护数据库需要投入大量人力物力,高密度信号指纹数据库还会增加定位所需时间,实时性难以得到保证。针对上述问题,提出一种基于高斯过程的两阶段信号指纹定位算法:利用高斯过程,在训练数据库基础上,生成高密度虚拟信号指纹数据库;定位过程中首先使用训练数据库确定大致范围,然后使用虚拟数据库进行精确定位。仿真结果显示,该算法与K近邻算法相比定位精度平均提高了81.9%,定位时间平均减少86%,从而验证了算法的有效性。Indoor localization based on Wi-Fi received signal strength(RSS)has attracted wide attention because of its good positioning performance.In order to improve the localization accuracy,it is necessary to build a high-density fingerprint database.However,the creation and maintenance of database requires a lot of manpower and material resources.High-density fingerprint database increases the time required for location,and it is difficult to guarantee realtime.For the above problems,we proposed a two-stage signal fingerprint localization algorithm based on Gaussian process.Based on the training database,a high-density virtual signal fingerprint database was generated by using the Gaussian process.In the localization process,the training database was used to determine the approximate range,and we used the virtual database to locate accurately.The simulation results show that the algorithm improves the localization accuracy by 81.9%and reduces the localization time by 86%on average compared with K-nearest neighbor algorithm,thus verifying its effectiveness.

关 键 词：室内定位 信号指纹 高斯过程 虚拟数据库 

分 类 号：TP929.5[自动化与计算机技术]