Research On AP’s Location To Improve The Accuracy Of Wi-Fi Indoor Positioning System

The subject comes from the project "R & D-based Wi-Fi indoor location system in Chongqing International Expo Center" which constructed by Chongqing Institute of Green and Intelligent Technology. Studies have shown that AP’s location has a great influence on the positioning accuracy of Wi-Fi-based fingerprinting location system.In this paper, based on the results of the field tests we established RSS model and a theoretical model of fingerprint-based positioning method. And we proposed a new concept of fingerprint vector discrimination. In the certain positioning section we discussed the AP’s number and shape to find the maximum fingerprint vector discrimination. And by finding the maximum fingerprint vector discrimination, the AP’s best position can be found.Firstly, in this paper, we introduced the basic concepts and related content about the indoor positioning, Wi-Fi-based indoor positioning and Wi-Fi-based fingerprinting positioning at home and abroad. And also we found that, currently, no one discussed the impact of the AP’s location to the positioning accuracy of Wi-Fi-based indoor location system.Secondly, we introduced several typical RSS models. According to the real positioning environment and the results of the field tests, we chose the suitable RSS model and calculated the path loss factor n at the same time.To optimize the AP’s location we proposed the concept of fingerprint vector discrimination. And we build a theoretical model of fingerprint-based positioning method to do the simulation tests. Test results showed that the AP’s number and position have a great influence on the fingerprint vector discrimination. On certain positions we can find the maximum fingerprint vector discrimination. And the higher fingerprint vector discrimination, the better positioning accuracy. To further improve the positioning accuracy, we reduced the spacing of reference points and other conditions equal.Field tests showed that the fingerprint vector discrimination has been raised after optimized the AP’s location. Correspondingly, the positioning accuracy has been improved.In short, the RSS model has been established in this paper. And a new concept of fingerprint vector discrimination has been proposed. The method discussed in this paper gives an idea for AP’s location optimization. Reference points are distributed in the field under certain preconditions to establish a fingerprint database with larger fingerprint vector discrimination. And it becomes feasible to improve the positioning accuracy of fingerprint Wi-Fi-based indoor positioning. In the foreseeable future, the positioning accuracy will be improved along with the better hardware conditions and the new algorithms.