Positioning And Tracking Of High-speed Trains Based On 5G NR PRS

With the advancement of railway,high-speed trains(HST)have brought great con-venience to our travel and logistics.Real-time location information of high-speed trains is not only significant in location-based services(LBSs),but also facilitates the imple-mentation of wireless connectivity,channel estimation and beamforming in railway com-munications.Therefore,the positioning and tracking of high-speed trains are particularly important in railways.With the deployment of 5G new radio(NR)networks in railway systems,its characteristics of continuous wide area coverage,high capacity hotspots,low power consumption,large connections,and high reliability and low latency put forward higher requirements for positioning and tracking accuracy of HST.This thesis studies the positioning and tracking of high-speed trains based on 5G NR positioning reference signals(PRS).In a 2D HST scenario,it is assumed that the train runs along the railroad at a fixed velocity and receives PRS signals from the base stations(BSs)deployed on one side of the railroad.Aiming at the condition that the traditional method requires multiple base stations to participate in the positioning at the same time,this thesis considers that the train is only associated with one BS at each positioning time.The main contents in this thesis are as follows:(1)In terms of positioning of HST,both the nonlinear deployment and strictly linear deployment of BSs are studied in this thesis.For HST scenario with nonlinear distribu-tion of BSs,this thesis proposes an iterative two-phase weighted least squares(I2WLS)method based on range difference of arrival(RDOA)to estimate train position.And then,the I2 WLS positioning algorithm is discussed when the train speed is known and the speed is unknown.The research shows that the gap between root mean square error(RMSE)of the proposed approach and Cramer-Rao lower bound(CRLB)is small when the RDOA measurements noise level is sufficiently small,and centimeter level accuracy can be achieved.(2)For high-speed railway scenario where BSs are strictly linearly distributed,this thesis proposes an iterative weighted least squares(IWLS)method to solve the localiza-tion problem.Compared with the scenario of nonlinear deployment of BSs,the position-ing algorithm proposed in this aspect only needs to do one step WLS.The simulation results show that a smaller PRS period,train speed and base station distance(BSD)can improve positioning performance.(3)In the research of HST tracking,we propose an extended Kalman filter(EKF)algorithm based on the train movement model and the RDOA measurement model,which utilizes prediction and update processes to obtain HST tracking results.The simulation results show that the performance of tracking will be reduced by an order of magnitude compared with the noise level.