Research On Channel Estimation And Beam Tracking Methods For V2X Millimeter Wave Communication

With the widespread application of massive multi-input multi-output in millimeter-wave(mmWave)communication systems,it is more difficult for the receiver and transmitter to obtain complete channel state information(CSI).Therefore,accurate channel estimation is required to ensure beamforming gain.Furthermore,the mobility of users in a highly dynamic time-varying channel will destroy the stability of the established data link,requiring repeated channel estimation to obtain real-time CSI,which causes additional delay and pilot overhead.Hence,in order to improve the reliability and efficiency of the mmWave communication system,the following schemes based on the hybrid precoding architecture are studied in this paper:1)A radar-aided multi-user time-varying channel estimation method is proposed.First,considering the characteristics of radar and communication modules,a joint transmission frame structure of uplink millimeterwave channel is designed,and the time-varying channel estimation process is decomposed into static angle estimation problem and dynamic gain estimation problem.Considering the imperfection of radar array elements in practical applications,an angle estimation algorithm based on subspace reconstruction of transmitted signals is proposed.Then the angle in timevarying channel is estimated through the reconstructed covariance matrix.In order to solve the performance loss caused by the high-speed movement of users,a millimeter-wave path gain estimation algorithm based on denoising and angle prior is proposed in this paper,constructing a denoising auto-encoding preprocessor based on supervised deep learning,which aims to optimize the traditional method and improve the anti-noise and antimovement robustness of the algorithm.Finally,the theoretical derivation of the upper bound and the simulation of the channel estimation performance are carried out.Simulation results show that the proposed algorithm can improve the accuracy of channel estimation.2)A beam tracking method based on decision fusion is proposed and a model decision fusion strategy is designed in this paper.First,the interframe correlation model of the angle of arrival is established and two predictors based on(Recurrent Neural Network,RNN)are built,which perform predictions based on the temporal correlation characteristics of the channel sequence and the correlation characteristics between adjacent frames,respectively.Specially,the fused coefficents of the two prediction models are depended on the performance of the prediction model in the previous frame.Due to the mm Wave is easily blocked,which causes abrupt changes in the channel,the correlation model of the path gains is constructed and a path gain tracking algorithm based on abrupt detection is proposed with reference to the above decision fusion strategy.The abrupt detection is performed on the received signal energy,and the channel information is re-obtained when the abrupt changes occur,avoiding the accumulation of errors in the tracking process.Finally,the proposed algorithm is simulated and analyzed.The results show that this method has higher tracking accuracy than existing methods.Finally,the research content of this paper is summarized and the future research topics are prospected.