Research On Wireless Location Algorithm Based On NLOS Identification And Error Mitigation

With the rapid development and widespread application of location services,people's demand for positioning accuracy is getting higher and higher,but NLOS(Non-Line-of-Sight)error is the most important factor leading to the decline of positioning accuracy.The NLOS identification method can effectively eliminate the influence of NLOS error on positioning accuracy,first identify NLOS BSs,and then use the LOS BSs to participate in the positioning.This thesis is based on the NLOS identification technology to eliminate NLOS error and do the following work:Firstly,the research status NLOS positioning is summarized by reviewing a large number of related literature.The NLOS error elimination technology is mainly divided into the direct NLOS error suppression positioning and use of LOS measurements for positioning after identifying NLOS.Secondly,the model of NLOS propagation and the influence of NLOS error on positioning accuracy are analyzed.Rwgh algorithm,IMR algorithm,SRwgh algorithm and LCC-Rwgh algorithm for NLOS error suppression are studied,the computational complexity of the above four algorithms is compared and analyzed.The simulation results show that the accuracy of the existing localization algorithms is drastically reduced in the scene with more NLOS BSs and larger NLOS error.Then,the NLOS identification methods are discussed.Considering from the perspective of signal detection of NLOS identification,NLOS identification algorithms based on Neyman-Pearson(N-P)and bayesian sequential test are studied.According to the N-P criterion,given the false alarm probability,a judgment threshold is obtained to identify NLOS.Based on the false alarm probability and missed alarm probability,the maximum likelihood estimation method is used to improve the judgment expression of the bayesian sequential test in the unknown NLOS probability density function.The simulation results show that the positioning accuracy of LOS BSs is significantly improved after identifying NLOS BSs by bayesian sequential test.After that,in the case of unknown prior information,considering from the perspective of residual detection of NLOS identification,the NLOS identification algorithms based on distance residual identification and positioning location residual identification are studied.The distance residual identification algorithm is improved by using TOA/AOA hybrid localization algorithm to estimate mobile middle position and then identifying the NLOSBSs,it can be identified in the scene of only two LOS BSs.The simulation results show that the identification algorithm based on TOA/AOA hybrid localization algorithm has better identification performance.Finally,the research work of this thesis are summarized,the deficiencies of the research are pointed out,and the future work are prospected.