Research On Fingerprint Location Technology Based On Compressed Sensing

In recent years, with the development of mobile application, the demand for wireless location service gradually appear and make the research for location technology become a hot topic.For location service, because of some disadvantages as multipath effect and NLOS, positioning accuracy is not high. However, the fingerprints location technology could utilize these adverse factors as the characteristic of place to minimize the error, so this method is an ideal localization way.Compressed sensing (CS) is a novel theory that proposed a few years before. CS could recover the original signals accurately from only a small quantity of sample values as long as the signals are sparse. Fortunately, this feature is suit for location system because the positioning process also possesses sparse property, and this is to say that the position of one MS is unique at one moment, compared to the whole area, it is sparse. So CS theory could be used for locating.In order to improve the positioning accuracy in mobile network, this paper made a research on CS theory and fingerprints location technology. First of all, we introduce the common location technology and elaborate the fingerprints location technology. Then the concept and principle of CS is described, since the reconstruction algorithm is the core of CS, we mainly focused on them, and many algorithms have been summarized, and an improved CMP algorithm is proposed, the simulation results shows that the new algorithm could obtain better quality than some algorithms; Also, in order to make the reconstruction time shorter, an algorithm that based on the optimal inner product is proposed, and the results prove that the new algorithm could improve the reconstruction speed without compromising on the quality, so it establishes the basis of the application of CS. Finally, a location system based on the CS is designed and the localization result indicates that this scheme could enhance the positioning accuracy.