Research On Ultra-wideband Signal Fingerprint Positioning Algorithm

With the increasing application of wireless communication technology in our lifeand work, the demand for wireless communication with better performance hasbecome especially urgent. As a new type of wireless communication technology,Ultra-wideband wireless communication technology has aroused great attention. Forits apparent advantages of the high accuracy, low power consumption, better signalconcealment, less interference to other systems and so forth, UWB positioningtechnology has become the preferred technique for wireless positioning, which issuperior to the traditional methods.This paper is organized as following: first we introduces the basic concept ofUWB technology, along with the background and the significance of the study. Thenbased on the presentation of the basic knowledge of UWB positioning, we give a briefintroduction of the current development state of this technology in both overseas anddomestic area. Secondly, the main methods of wireless positioning were studied, andboth of their advantages and disadvantages are compared and analyzed. Besides,several factors which affect the accuracy of positioning are summarized andcorresponding solutions are proposed. Furthermore, the UWB wirelesscommunication system is studied in detail, including the UWB pulse waveforms,modulation technology, channel model, receiver technology and so forth, whichprovide the basis for the simulation of the following proposed algorithm. In this paper,we adopt Gaussian pulse and PPM-TH-UWB modulation for our UWB signal in thesimulation.The NLOS, multipath, signal fading and other problems of the wirelesscommunication environment have a great impact on signal propagation. In order toreduce the environmental impact on the positioning accuracy, we need to create awireless positioning algorithm which can adapt to the complex environmentappropriately. UWB signal fingerprint positioning algorithm is such a method that it establishes an offline database with the fingerprint data measured under the complexcommunication environment, by utilizing the fingerprint information of UWB signal.By training the offline data with neural network algorithm, these data can perfectlymatch the actual environment, thus reducing the negative impact form the complexenvironment on UWB signal and improving the positioning accuracy at the sametime.Finally, simulation of performance of both the UWB signal fingerprint positioning algorithm and LES positioning algorithm are conducted and compared.Results show that the UWB signal fingerprint positioning algorithm can achievehigher precision and better stability.