Research On WiFi Indoor Localization Arithmetic Based On Location Fingerprint

With the rapid development of wireless networks and wireless communications market share increase,WiFi deployment in the indoor environment has become particularly important and spread to all areas of life,and the Internet can be accessed only with a simple mobile device.Higher positioning accuracy based on the location of the fingerprint positioning algorithm using off-the-shelf WiFi wireless network,no additional increase or only a small increase in hardware costs,so based on location fingerprint positioning in recent years by the high concern.Location-based fingerprint localization algorithm mainly Nearest Neighbor(NN),K Nearest Neighbors(KNN)and so on.These classical positioning algorithms are sufficient for most of the use of the site,but the classical positioning algorithm has a large amount of computing,consuming more hardware resources and long positioning time.Especially,some large shopping malls,airport s and railway stations involve offline database large volume of the use of the scene.Firstly,this article describes the experimental environment and the division of the sample size of the offline database.According to the experimental environment,four APs are set up reasonably.The advantages and disadvantages of the establishment of a variety of off-line database are introduced,and the offline database process.Using the offline database collected in this paper,the positioning accuracy of several classical positioning algorithms is compared and verified.The positioning accuracy of the classical positioning algorithm is found.Can meet the general civilian occasions,for the improved positioning algorithm provides experimental basis.Secondly,based on the JAVA language,the theoretical basis of the time complexity calculation of online positioning algorithm is given.Based on this theory,the calculation method of WKNN and KNN algorithm is used to calculate the time complexity of the algorithm.Out of the corresponding calculation formula.Finally,this paper proposes an improved clustering algorithm based on the propagation model,which is based on the descending classification of the off-line database.In this paper,we propose an improved clustering algorithm based on the propagation model for the large time of WKNN and KNN online positioning algorithm.Can effectively solve this problem,reduce the computational complexity of the algorithm to reduce the positioning run time.