The Research On UWB Indoor Localization Algorithm

In recent years, wireless communication technology is developing rapidly. Withthe improvement of people’s life, A variety of new business of wireless communicationhas appeared one after another. The7.5GHz frequency band is now opened to publicuse that promote the development of ultra-wideband(UWB) wireless communicationtechnology. UWB technology can be used for accurate distance measurement andlocalization. It has the characteristics of high precision and low power consumptionand large system capacity which made this technology a inevitable choice for futurewireless localization. In this paper, based on the theory of UWB communication andthe goal of improve the accuracy of UWB indoor localization, we optimized thetraditional indoor localization algorithm, and verified the effect of optimization onlocation algorithm. The main research content is as follows:First, Through the analysis and comparison of some localization method, weevaluate each method in detail for the applicability to UWB system. Aiming at theseproblems of these basic algorithms, this paper selects several targeted improvementalgorithm. Before using these algorithms, we have to sufficiently understand the coreprinciple of these algorithms. And then made the LANDMARC indoor localizationalgorithm as the research platform for algorithm optimization. We analysis andsimulate the problems which LANDMARC algorithm have, summarize the problemsexisted in it.Then, we started to solve the different defect of LANDMARC algorithmrespectively. To solve the problem of large amount of calculation that LANDMARCalgorithm have, we adjust the parameters of the algorithm constantly, and simulateagain and again, and finally found the problem. In order to solve this problem, weconsult the information and seek solutions, after comparing various solutions, wedecided to use region segmentation algorithm to solve the problem. To solve theproblem of stability of LANDMARC system, we chose an algorithm called adaptive knearest select algorithm to optimize the ability of adaptability of LANDMARCalgorithm. To solve the problem of localization error of LANDMARC algorithm madeby hardware circuit, we chose kalman filter. Combined with a variety of optimizationalgorithms, we tried a variety of solutions, simulate and analyse the performance of each solution, finally chose an algorithm which include the function of kalman filterand BP network and adaptive k nearest and the smallest inclusive circle. It is called anadaptive algorithm which have the function of filter.Furthermore, we set up the UWB indoor localization model, simulate theoptimized algorithms above respectively, compared the accuracy of localization withtraditional LANDMARC algorithm, analyse the reason of the result in detail, andsummarize the reasons. Draw to the conclusion finally: After improved, the adaptivelocalization algorithm which has the function of filter can overcome the defects oftraditional algorithm, improve the accuracy of localization.Finally, this paper made a serious summary for the whole research work, and putforward some problems which should be solved in the future.