Research On High Accuracy Estimation Algorithm Based On 5G NR Downlink Channel

As the latest mobile communication system,5G NR(New Radio)has developed rapidly in recent years and gradually started to be put into commercial use,attracting extensive attention from all walks of life.Among them,channel estimation is an important module of downlink demodulation of NR system,because the performance of wireless communication system is greatly affected by the fading in the wireless channel,and the accuracy of channel estimation plays a decisive role in the quality of signal detection.Most of existing channel estimation methods have their own limitations.For example,when the SNR is low,the estimation accuracy of LS(least-squre)algorithm is poor.And MMSE algorithm complexity is high,which is not conducive to the demodulation efficiency of 5G devices.In addition,in the physical layer design of NR system,less Demodulation Reference Signal(DMRS)is allocated to SSB(Synchronization Signal Block),and the distribution of reference signals is not as uniform as that in LTE(Long Term Evolution)system,which results in the performance decline of NR channel estimator.Aiming at the above two problems,this thesis designs a high-precision channel estimation scheme suitable for 5G according to the characteristics of NR system.The main work contents are as follows:A low SNR and low complexity channel estimation algorithm for PBCH downlink channel is proposed.In this thesis,based on the distribution characteristics of demodulated reference signal in SSB,a noise-removing algorithm with low computational complexity is designed to reduce the influence of noise and inter-carrier interference on channel estimation,helping estimator restore channel fading more accurately.In the case of low signal-to-noise ratio(SNR)of 0dB,the proposed algorithm can reduce the bit error rate by 12%compared with the LS algorithm with the same time complexity.It is proved that this channel estimation method has the advantages of low complexity and high precision.An optimization scheme of channel estimation using demodulation reference signal and synchronization signal is proposed.Based on the original channel estimation algorithm,the synchronization signal is used to estimate the channel fading for the second time in order to optimize the channel estimator in the NR system.The simulation results show that the proposed scheme has performance gain for LS,LMMSE and the proposed channel estimation method,the mean square error of channel fading decreases by 1.7%,2.5%and 1.6%,respectively when SNR is 0dB.It proves that the proposed scheme is universal and connected with the first work in this thesis.The measured platform was built,and the NR downlink demodulation process from cell search to Cyclic Redundancy Check(CRC)was completed.The channel estimation method and optimization scheme proposed in this thesis were tested with NR actual signals,which proved that the research content in this thesis has practical application significance.