Research Of3D Positioning Scheme Based On Fingerprinting Matching

With the development of economy and society, users can use mobile computing devices to access the Internet anytime, anywhere. Therefore the Location-Based Service (LBS) in indoor environments has recently received unprecedented attention. Nowadays, the demands for indoor location and navigation are increasing enormously, especially in complex indoor environments, so it is necessary to research three dimensional positioning scheme, which is high-precision, low-cost and highly adaptable interior scene. Then we apply it to wireless network refine optimization and problem analysis.This thesis focuses on fingerprinting positioning technology, and makes research on positioning algorithm, the method to generate pattern database and the way to improve the accuracy of database in cellular networks, designs fingerprinting positioning scheme, construct of database, and the skill to revise simulation pattern database by engineering parameters and drive test of outdoor environments. In addition, the methods to improve positioning efficiency and accuracy are proposed in the case of multi-dimensional information and related fingerprints.Firstly, this thesis reviews the existing common indoor positioning technology and fingerprinting matching positioning algorithm, and makes comparision and analysis on their applicability. In cellular networks, a new3D fingerprinting positioning scheme is proposed by using implicit information of the location around geographic environments. Then, we elaborate its principles and specific processes. Compared with the maximum likelihood method and the weighted k nearest neighbors method, the promising numerical results are also presented. Afterwards, the paper describes the three-dimensional positioning system platform, its functional modules, and expounds database generation, including the actual and simulation database. Given that the simulation database is not accurate enough, it is revised by taking advantage of outdoor drive test data and other information. For different building height scenarios, two amended methods are given and validation results show that the proposed methods greatly improve the positioning accuracy and matching efficiency. Finally, in certain scenarios, positioning efficiency and accuracy are enhanced through multi-dimensional measurements and the relationship of every fingerprint. Multi-dimensional measurements include best cell, time of arrival, and angle of arrival, receive signal time difference, received signal code power, timing advance and so on. Compared with traditional matching method (such as maximum likelihood algorithm) and cellsense positioning model based on Bayesian probability and statistics, the proposed methods have better results in terms of efficiency optimization and accuracy improvement.