Design And Optimization Of WiFi Indoor Positioning System Based On Fingerprint Contrast Technology

There is more and more research and attention observed in the domain of indoor localization, which is a critical technology in the Internet of Things and O2 O. The target of indoor localization is to provide location service in the indoor environments where the traditional position system such as GPS cannot work well any more. However, due to the extremely complexity and uncertainty of the indoor environments, there does not exist a uniform, general, and scalable indoor localization solution that could work well anywhere. As an exploratory in the domain of indoor localization, this thesis studied that how to improve the accuracy, delay, and visualization performance of localization by proposing a new localization algorithm, and this work also designed a complete indoor localization system which is tested by a series of experiments to demonstrate the correctness and robustness of the system. In contrasted with previous works, the advantages of this work are four folds: firstly, the localization accuracy is improved by combing the closest-AP localization and fingerprint-based localization; secondly, the system cost less because there is no hardware required given the general wireless infrastructure for the wifi convergence, and count of the sample points is also small leading to the save of manpower; thirdly, the delay of localization is short because the algorithm is specially designed to be compatible with fast refresh rate, which guarantees the real real-time localization; and finally, the system address the problem of how to visualize the trajectory of mobile device moving from a point to another point across a wall or some other obstructs by including artificial geographic information, which could also be used to avoid the jitter of localization results. In conclusion, the proposed localization solution is effective and practical to be deployed in the real world.