Design And Implementation Of Indoor Positioning System Based On WIFI Location Fingerprint

Studies have found that people spend most of their time indoors,but the current Global Positioning System(GPS)can not provide indoor positioning,so relevant scholars have proposed many methods to achieve indoor positioning.Among the many indoor positioning technologies,WIFI positioning has the characteristics of no need to install equipment and is easy to popularize,therefore,it has become a popular research direction for indoor positioning.After analyzing the current commonly used indoor location algorithm,the thesis chooses the location fingerprint method based on received signal strength(RSS)to realize WIFI location,and improves some problems in the offline and online stages of the positioning process,and then implements the improved algorithm.First of all,in the offline phase,the selection algorithm of WIFI access points(AP)is improved,combining the probability of AP occurrence with the RSS mean and average difference of AP,and the maximum information coefficient(MIC)is introduced to select AP;in terms of filtering,the mean filtering is combined with limiter filtering,and the problem that limiter filtering can not filter outliers that appear continuously at the beginning of sampling is improved;aiming at the problem of reduced positioning accuracy due to heterogeneous equipment,a zero score(Z-Score)standardization is proposed to process fingerprint data.Secondly,in the online stage,aiming at the problem that the clustering of RSS values is likely to produce discrete points,it is proposed that after the use of weighted k-nearest neighbor(WKNN),the coordinates are not directly estimated,but are clustered,and the coordinates are divided into useful and useless classes,and the coordinates are estimated in the useful classes to achieve the elimination of discrete points.Finally,the algorithm is implemented by WeChat and tested.Through the WeChat applet,the client/server(C/S)model is used to collect RSS values on the applet side,display the positioning results,and store the RSS values on the server side for training and matching.