Research On Key Techniques In High Precision Wireless Local Positioning

The high precision wireless local positioning is a hotspot in the wireless positioning area. The ultra-wideband(UWB) technique is the first selection for it because UWB signals have the high time resolution and strong ability to penetrate the obstacles. At the same time, the large relative bandwidth of UWB signals bring many problems and challenges to high precision localization. With the consideration that the high precision distance measurements ability is the prominent advantage and important premise of UWB positioning, this dissertation carries out the research mainly on two key technologies. One is UWB ranging and the other is target position estimating under the non-line-of-sight(NLOS) conditions. It has achieved important results as follows:Firstly, on the basis of intensively analysing the UWB channel model in the multipath environments, we give a modified peak detection algorithm. As the simulation showing, it is more outstanding when in the non-separable channel. Moreover, we introduce the time reverse(TR) technique into the UWB high precision ranging, and propose a time of arrival(TOA) estimation method basing on the TR processing. It has realized the matching of the multipath components and the accurate estimation of TOA through retransmitting the TR signals which containing the channel impulse response information into the UWB channel. The simulation results show that compared with the traditional method of generalized maximum likelihood, this method has a lower TOA estimation error and a lower input signal-to-noise ratio(SNR) ranges.Secondly, aiming at the non-Gaussian UWB channels and the existing of the outliers, we propose the TOA estimation methods basing on the rank statistics. According to the properties of multi-pulse(MP) ranging, we design two different kinds of rank detection forms including row rank and column rank. Then we propose two TOA estimation methods basing on the row rank and column rank respectively. These two methods all suppress the effects of heavy tailed non-Gauss noise and outliers through using the rank of the signals, and at the same time, they can protect the information of the first path(FP) effectly. The simulation results show that the methods are more effective and robust on both Gaussian and non-Gaussian noise conditions, and they also have the advantage of simple calculation.Thirdly, aiming at the high precision localization problem in NLOS environment, after the intensively studying of the positioning methods using weighted least square(WLS) method, and according to the relation between the signal-to-noise ratio of the received signals and the distance measurements reliability, we point out that when designing the weighted value basing on the distance information, it can avoid imbalance problems bringing by using distance estimation variance as the weighted value. We propose a WLS positioning method which combines the distance information weighting with the result of non-line-of-sight identification. This method designs the WLS optimal weighted value using the distance between the measurement nodes and the location node. Further more, it combines the results of NLOS identification with the design of distance weighted array in order to further improve the positioning accuracy. The simulation results show that this method has the significantly better positioning performance than those existing WLS methods.Fourthly, using indoor non cooperative target positioning as the background, and aiming at the one-transmitting and multi-receiving mode, and the self-transmitting and self-receiving mode, we provide the positioning system models for them under line-of-sight(LOS) and through-the-wall(TTW) environments respectively. Using this as foundation, we build a modified ranging model basing on the analysis of the TTW propagation. The ranging model rebuilds the delay of the receiving signal which is seriously affected by the wall according to the relation between the range and the wall parameters, and also the incidenting angle of the propagation path. So we can eliminate the ranging error due to TTW propagation, and improve the accuracy of the location estimation. Basing on the modified ranging model, we propose a new TTW geometric curve positioning method. This method computes the actual target position using the TOA measurements of the positioning reference spots and the modified ranging model, so it has higher positioning accuracy. The dissertation has also deduced the CRLB expression considering of TTW propagation. The simulation results show that this positioning method has reduced the localization error causing by TTW propagation effectly. And it has advantages of both high precision target location estimation and low computational complexity.