A Study Of Multipath Mitigation Technology In GNSS Dynamic Urban Positioning Environment

In complex urban environments,the Global Navigation Satellite System(GNSS)plays a key role in providing spatial and temporal datum,intelligent means and immediate dynamic location-related feedback.However,its receivers are affected by a large number of high-rise buildings,resulting in Multipath(MP)and Non-Line-of-Sight(NLOS)phenomena,which can cause serious deviations in positioning accuracy.Specifically,on the one hand,the presence of MP distorts the results of the receiver tracking loop and makes it more difficult to capture weak signals with low Carrier-to-Noise Ratio(CNR).On the other hand,for the dynamic relative positioning of driving vehicles in urban environments,the influence of MPinterfering signal sources and the residual iterations generated by the tracking and solving process make the receiver’s positioning accuracy low during movement.Therefore,mitigating the MP effect and reducing the influence of NLOS signals are key factors to improve the positioning accuracy and robustness of the receiver.To address the two major issues mentioned above,the main research of this thesis is divided into:(1)The algorithm based on Random Hough Transform(RHT)slope mitigation method and adaptive carrier signal tracking.In this thesis,from the perspective of traditional receiver loop multipath mitigation methods,the delayed impact of multipath effects on receiver tracking and capture is mitigated by correcting the code loop correlation curve distortion and tracking weak dynamic carrier signals by adaptive carrier signal update mechanisms at two levels.Firstly,the RHT slope mitigation method is combined with the identification of the slope of the straight lines on both sides of the auto-correlation function to reduce the error of the narrow correlator and to correct the code loop distortion.In addition,an adaptive carrier signal tracking algorithm is designed to assist the code loop in tracking and capturing weak signals in the low CNR region.Finally,the real-time co-ordinate positions of the receiver are solved and compared with real trajectories and performance analyses are made.The experimental results show that the average multipath error converges to a stable value around1 m after optimization with the RHT slope mitigation method under the condition of 25,and the receiver positioning accuracy is improved by 50%;and the carrier loop error value of the proposed method gradually converges to around 3～4 with the increase of code delay offset.In the D region,where the MP effect is strong,the final RMSE value is reduced by 6.3m.(2)The multipath mitigation model based on XGboost(e Xtreme Gradient Boosting)for double-layer consistency checks.In order to improve the usability of the healthy signal dataset and reduce the localization error caused by the dynamic localization process,this thesis firstly preprocesses the collected signal data sources in the first consistency check(CC1)process,using the chi-square fit test and the traditional LS method to estimate the position and eliminate the unhealthy signals.Secondly,the remaining available signals are then fed into the second consistency check(CC2)of the XGboost classifier,which is used to improve the detection rate of NLOS,LOS(Line-of-Sight,LOS)and MP signals,and then to exclude measurements that differ from the empirical pseudorange residuals.Experiments show that the classification accuracy of NLOS can reach 93.6%;and the results of two localization checks after a Weighted Least Squares(WLS)smoothing estimation fit show that the relative localization pseudorange error value converges to 1.668 m.