Research And Implementation Of Indoor Positioning System Based On WIFI

With the economy developing rapidly,science and technology has a trend of booming.In recent years,Location Based Service has attracted more and more people’s attention.Outdoor positioning systems are mainly with assist of cellular localization and global positioning system(GPS),which can provide high positioning accuracy in outdoor positioning.However,block of buildings and multipath propagation etc.for the signal makes it difficult to meet the needs of positioning accuracy in complex indoor environments.As for indoor positioning technology,WIFI positioning technology has become a research hotspot for its relatively high positioning accuracy and low deployment cost.But the results of indoor location algorithm based on WIFI positioning technology nowadays have greater volatility and worse positioning accuracy,thus still having need of further optimization.In this thesis,we propose an improved indoor positioning method based on the weighted K nearest neighbor algorithm as well as algorithm based on curve fitting(CF)to obtain better accuracy.Furthermore,we implement an indoor positioning system which can achieve a higher positioning precision and better robustness based on the proposed algorithm.The research results are displayed as flows:1.Analyzing advantages and disadvantages of K-nearest neighbor algorithm,weighted K-nearest neighbor algorithm and algorithm based on CF,and proposing an improved algorithm WSKNN based on weighted K-nearest neighbor algorithm and the CF algorithm.WSKNN algorithm makes use of filtration method for removing outliers,and integrates advantages of exhaustive location search algorithm(ELS)and gradient-descent-based location search algorithm(GLS),improving positioning accuracy.We have conducted a test in the 15 m × 15 m lab environment and the results show that the improved algorithm can obtain improvement of the mean error distance of 22.54%and 7.17%,compared with traditional KNN method and the GLS-CF method,respectively.Also,the standard deviation distance is also smaller comparing to the GLS-CF algorithm.Thus,the results of the improved positioning algorithm fluctuate smaller and positioning more precise.2.Designing and implementing an indoor positioning system based on WIFI applying the proposed improved algorithm.We have designed and implemented functions for three parts which are the server,client and database,including two stages of offline training and online positioning.Tests in the real environment show that the indoor positioning system applying WSKNN algorithm has high positioning accuracy and good stability,and can achieve the purpose of positioning the mobile device in the actual environment.