Research On Indoor Positioning Algorithm Of Location Fingerprint Based On WIFI

With the rapid development of wireless communication technology and the widespread popularity of mobile intelligent terminals,there is an increasing demand for location services based on indoor environments.Indoor positioning technology is urgently needed to provide high-quality location information services in medical,commercial and emergency rescue areas.Among various indoor positioning technologies,WiFi-based location fingerprint positioning method has the advantages of high positioning accuracy,easy implementation,strong scalability,wide coverage and low cost,and has gradually become a hot research topic.Therefore,this paper mainly studies the location fingerprinting method based on WiFi,the main contents are as follows:Firstly,the principle of location fingerprint positioning based on WiFi and traditional indoor location algorithm are introduced,and the advantages of location fingerprint location based on WiFi are proved by analyzing and comparing different location methods,which provides theoretical support for the selection of location methods in this paper.Secondly,in the establishment stage of fingerprint database,various factors affecting wireless signal propagation are introduced.An improved comprehensive filtering algorithm is proposed.The algorithm is applied to filter the collected data and build a simulation experiment platform.The results of data processing are compared with those of maximum filtering algorithm,Gauss filtering algorithm and mean filtering algorithm.The results show that the proposed algorithm is effective in processing data.The fluctuation range and variance of RSSI value were reduced,and the quality of fingerprint database is improved.Then,based on AP selection,the KPCA algorithm is proposed to extract the features of location fingerprint database.The simulation results are compared with LE,ICA,LDA and PCA algorithms.The experimental results show that KPCA algorithm has unique advantages of nonlinear feature extraction,which further improves the quality of fingerprint database.Finally,for the selection of positioning model,this paper introduces the traditional positioning algorithms such as NN,KNN,WKNN,probabilistic algorithm and machine learning-based positioning training model SVM,proposes the LSSVM indoor positioning model,and proposes an improved parameter optimization algorithm KV-PSO to optimize the parameters C and ? in the positioning model.Through experiments,the proposed positioning algorithm(KV-PSO-LSSVM)and the traditional positioning algorithm are tested in this paper.The test results show that the proposed algorithm reduces the positioning error and shows superior positioning performance.