基于余弦相似度的指纹匹配算法的室内定位方法  

Indoor Positioning Method Based on Cosine Similarity of Fingerprint Matching Algorithm

作　　者：刘冰[1,2] 李文书[2] Liu Bing;Li Wenshu(Digital Information Technology Department,Zhejiang Technical Institute of Economics,Hangzhou 310018,China;Lab of Intelligence Recognition and System,School of Information Science and Technology,Zhejiang Sci-Tech University,Hangzhou 310018,China)

机构地区：[1]浙江经济职业技术学院数字信息技术学院,杭州310018 [2]浙江理工大学信息学院智能识别与系统实验室,杭州310018

出　　处：《科技通报》2017年第3期198-202,共5页Bulletin of Science and Technology

基　　金：国家自然基金(60702069;31271102);浙江省自然基金(Y1080851;Y12H290045);浙江省钱江人才(B类)项目(2012R10054);浙江省高校访学教师发展项目(FX2014181)

摘　　要：介绍了基于位置指纹匹配的Wi Fi室内定位常用方法,提出了一种基于余弦相似度的指纹匹配算法的室内Wi Fi定位方法。该方法匹配算法复杂度低,降低了手机终端定位应用的功耗。本文也给出了Wi Fi定位系统的具体实现方法,通过在某室内环境将余弦相似度算法与常用方法进行对比测试,误差范围<1.5 m,准确度较高,满足室内定位精度需求。This paper introduces the common methods of WiFi indoor location based on position fingerprint matching,and proposes a new method of indoor WiFi location based on cosine similarity.This method has low complexity and reduces the power consumption of mobile terminal positioning applications.This paper also described a method to implement the WiFi positioning system.By comparing test between the cosine similarity algorithm and common method in an indoor environment,the error range is<1.5m and the accuracy is higher.The method can meet the indoor positioning requirements.

关 键 词：室内WiFi定位 RSSI 指纹匹配 余弦相似度 ANDROID 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]