Research On Indoor Location Technology Based On WiFi Location Fingerprint

With the continuous development of mobile devices,various wireless technologies can provide precise positioning methods for indoor positioning and navigation services.WiFi-based indoor positioning technology utilizes existing WLAN infrastructure,requires no additional hardware,and covers a wide range of indoor scenarios,making it compatible with smartphones and other mobile devices,thereby providing convenient location-based services and value-added services that have garnered significant attention.The current WiFi signal strength-based indoor fingerprint positioning faces several complex issues that affect positioning effectiveness,primarily divided into two aspects.Firstly,there is the issue of the labor-intensive traditional fingerprint database construction in indoor positioning.Secondly,common fingerprint positioning algorithms suffer from reduced accuracy due to complex factors.To address these problems,this paper presents the following contributions:(1)In response to the labor-intensive traditional fingerprint database construction during the offline database building phase of indoor positioning,along with the limited accuracy and substantial workload associated with general expansion methods,this paper proposes a Kriging interpolation method based on the mayfly optimization algorithm.This algorithm utilizes a small amount of fingerprint information to establish a variogram function and employs an improved mayfly optimization algorithm to fit the variogram function model,thereby generating an offline fingerprint database to enhance interpolation accuracy and efficiency.Experimental analysis demonstrates that when the sampling ratio reaches 50%,the improved Kriging interpolation achieves a positioning accuracy of 3.02 m,with only a 0.18 m difference in positioning accuracy compared to the original complete fingerprint database.This approach realizes a dual optimization of performance and efficiency,reducing the human and time costs associated with database construction.(2)In the online positioning phase,common fingerprint positioning algorithms in indoor environments are susceptible to the influence of complex factors,resulting in reduced positioning accuracy.This paper proposes an adaptive weighted Knearest neighbor(KNN)online positioning matching algorithm that combines Gaussian functions with Euclidean distance and propagation distance.It adapts the selection of K values based on the distribution characteristics of fingerprint data and introduces Gaussian functions to calculate weight coefficients for propagation distance and Euclidean distance separately.Finally,it combines the two weight coefficients to obtain the weighted KNN.Experimental analysis shows that the proposed algorithm significantly improves positioning accuracy compared to the existing weighted KNN algorithm,with a positioning accuracy improvement of0.49 m.It also demonstrates superior performance compared to other positioning algorithms.