Design And Implementation About Indoor Location-based Services Based On Finger Printing Method

Indoor location-based services have become popular because of the rapid growth of the IO T(Internet of Things) which causes much attentions of person. This develops a WIFI location system to provide indoor location services with high-accuracy, high-speed and high-stability to match this tendency.This dissert presents the development situation and tendency of wireless location technique, and uses receive diversity to improve the quality of signal innovative. This technique can increase the stability of received signal strength remarkable and maximum limit influence of the waviness of WIFI to positioning accuracy, meanwhile have little effect to the cost.Tthe finger printing method is chosen in this dissert which can get the best accuracy. In traditional K-nearest neighbor method, the calculate results are discrete distribution and make systemic error. This dissert develop the weight K-nearest nerghbor way to eliminate it. During the research, we find out a pre-estimate kernel method creatively. It adds a pre-estimate process before the kernel method, and can significant decrease the computation time. Because the error distribution is much more balance, our location system has higher speed and higher stability in this way, the location performance get better.Base on the theoretical research, this dissert introduces the whole project solution of the WIFI location system including software and hardware. Then we get the sample data in the testbed, analyse the result of every project, and pleasant to find the pre-estimate kernel method performs well in accuracy, speed and stability. At last, the dissert sums up experience and lessons in this work, meanwhile suggests research directions of the future work.