Research On High-performance Localization Algorithm For NLOS Environment Based On Optimization Theory

With the rapid development of information industry,People put forward higher demands on the accuracy of location information.Non-Line-of-Sight(NLOS)transmission is the main reason for the deviation of target location.Therefore,how to reduce the localization error caused by NLOS propagation to improve the positioning accuracy of mobile terminal is a key problem to be solved urgently.In this thesis,the anti-NLOS localization algorithms of the time difference of arrival(TDOA)and the received signal strength(RSS)based on the optimization theory are studied in depth,and this thesis mainly studies how to fully apply the optimization theory to further improve the localization performance of the algorithm.The main contents of this thesis are as follows:Firstly,after consulting a large number of domestic and foreign literatures,the research status of anti-NLOS high performance optimization localization algorithm is summarized in this thesis..Secondly,the quadratic programming is added to chan algorithm which is the classical localization algorithm in Line-of-Sight(LOS)environment,and a TDOA location algorithm based on quadratic programming is proposed.The simulation results show that the classical localization algorithm in LOS environment is improved by using the optimization theory,and it has a good effect on the reduction of localization error caused by NLOS propagation.In view of the severe transmission of NLOS,the data preprocessing and iterative improvement are used to optimize the TDOA location algorithm based on quadratic programming,and the new data processing method and iterative idea are reintegrated into the original algorithm.An improved TDOA algorithm based on quadratic programming is proposed to further reduce the NLOS error.The simulation results show that the positioning accuracy of the algorithm is further improved by adding data preprocessing and iterative improvement on the basis of quadratic programming.Then,aiming at the problem of RSS location,a method of using UT transform to approximate RSS location problem to WLS problem is proposed.Then,the approximate problem of WLS is equivalent to the semidefinite programming problem(SDP)by taking the form of mirror diagram as the intermediate form.The original RSS localization problem is simplified and the algorithm is coordinated in terms of computation complexity and location accuracy.The simulation results show that the relative position of the target and base station is the main factor affecting the localization performance of the optimization algorithm.The location accuracy of the algorithm is relatively high when the mobile terminal is located within the surrounding area of the base station.Next,the calculation of the Cramer-Rao Lower Bound(CRLB)in NLOS environment is carried out.Under the condition of LOS/NLOS propagation path identification,the relationship between CRLB and LOS and NLOS base station parameters and the statistical characteristics of localization error caused by NLOS propagation is analyzed,and the variation of CRLB with the above parameters under different location network structure is analyzed by simulation.Finally,the research content is collated and summarized,and the future research focus is planned.