Research On Estimation And Compensation Methods Of Phase Noise In High Frequency Millimeter Wave OFDM System

The fifth generation(5G)mobile communication system exploits the millimeter wave frequency band to obtain more available bandwidth to support the rapid growth of mobile traffic.However,as the frequency increases,large phase noise caused by hardware defects can impair system performance.The 3rd Generation Partnership Project(3GPP)is studying high-frequency millimeter wave technology above 52.6GHz,and one of the key issues is how to deal with phase noise.Different from low-frequency millimeter wave,there are two new characteristics of the influence of phase noise on the signal in highfrequency millimeter wave orthogonal frequency division multiplexing(OFDM)system.One is that serious phase noise can easily lead to inaccurate estimation of frequency offset,resulting in residual frequency offset(RFO),and the RFO will lead to phase deflection as time accumulates,making high-frequency millimeter wave phase noise correlated in the time domain;Second,compared with the central local oscillator(LO)architecture commonly used in the low-frequency millimeter wave multiple input multiple output(MIMO)OFDM system,the high-frequency millimeter wave MIMO-OFDM system typically use a distributed LO architecture with independent phase noise on each antenna.In this way,there are new interference items of phase noise in addition to the original common phase error(CPE)and inter-carrier interference(ICI).Aiming at the correlation of high-frequency millimeter wave phase noise caused by RFO in the time domain,a joint estimation algorithm of phase noise and RFO based on kalman filter is proposed in this thesis.This method uses the phase noise and RFO as the second order state value of the system,the estimated phase noise obtained from the mathematical model as the prior estimate of the real state of the system,the phase noise estimated by the Phase Tracking Reference Signal(PTRS)as the observed value of the real state of the system.The observed value is then used to correct the prior estimate to obtain a more accurate posterior estimate to compensate for the phase noise in the system.The simulation results show that the proposed method can significantly reduce the system block error rate and improve the system performance under various signal-to-noise ratio scenarios compared with the traditional phase noise estimation and compensation algorithm.In view of the independence of phase noise of different radio frequency channels under distributed LO architecture,this thesis proposes a phase noise estimation and compensation algorithm based on timefrequency domain hybrid processing.Firstly,PTRS is used for CPE estimation and compensation in the frequency domain,and then time domain processing is performed.Since the impact of new interference terms and ICI is reflected in the difference between instantaneous phase noise and CPE,the received signal after channel estimation and frequency domain equalization is divided into several sub-blocks after N-point Inverse Discrete Fourier Transform(IDFT),and the difference between the average phase value of each sub-block and the CPE is used as the estimated value of the new interference term and ICI,then secondary phase compensation is performed.Finally,the effectiveness of the proposed algorithm is verified by simulation.The simulation results show that under the distributed LO architecture,the proposed algorithm not only compensates for CPE,but also effectively compensates for the impact of new interference items on the signal,so its performance is better than the traditional algorithm.