The Research And Implementation Of WiFi Indoor Location Algorithm Based On Location Fingerprint Recognition

Traditional GPS and other location technologies have been able to achieve high accuracy on meter level in outdoor environments.But for indoor environment,they can’t achieve that accuracy.With the popularity of mobile intelligent terminal and WiFi provides the possibility for low-cost,high-precision indoor positioning technology. Wherein the WiFi indoor location algorithm based on location fingerprint recognition its simple, low cost, high precision positioning has become a hot topic of indoor positioning technology.Through in-depth study of this algorithm, an algorithm to further improve the positioning accuracy and speed is improved.Offline phase,Indoor positioning environment becomes increasingly large and fingerprint collection points also increase.Taking into account real-time positioning,this paper handles the database by clustering. The paper improves the binary K-means.The similarity of the clustering algorithm is the product of the signal strength’s and the coordinate’s euclidean distance. After clustering, the classification accuracy and the positioning accuracy are improved. Online stage, an improved WKNN is proposed.Firstly, we should detect the AP signal in the spot to be located and weight the signal strength of AP. secondly, we filter K neighboring points by the weighted euclidean distance and then remove discrete points.Lastly, the remaining points are weighted averaged.Then the paper further improves the algorithm to improve positioning accuracy of clustering boundary anchor points.Experimental results show that the proposed algorithm effectively improves the positioning accuracy and reduces the amount of computation targeting. The average positioning error of the improved algorithm is 1.24 m.Compared to the traditional WKNN, the average positioning erorr decreases 0.47 m and can improve more than 27.4%. The amount of matching fingerprint data reduces by Q-1/Q and the positioning time is shortened by Q-1/Q accordingly. Finally, the proposed algorithm is applied to the garage navigation system, wherein the positioning system module using C / S architecture.The paper introduces the client, server, database module, and then introduces the design and implementation of each module.This study can provide theoretical support for further research on WiFi fingerprint location algorithm and provide the appropriate algorithm support to determine the current location for major positioning and navigation systems.