Research On Performance Of Synchronization And Measurement For NR Systems

New Radio(NR),as the fifth generation of mobile communication technology,supports three typical scenarios of services,including enhance Mobile Broadband(eMBB),massive Machine Type Communication(mMTC)and Ultra Reliable and Low Latency Communication(URLLC).In order to meet various service requirements,NR is very flexible in design.Cell search and measurement are key processes that NR terminals must go through before communicating with NR networks.Its design has some similarities with the Long Term Evolution(LTE)system,but it is improved and introduced on the basis of LTE,many new features have been included.In terms of cell search,the concept of Synchronize Signal Block(SSB)is introduced in the NR system,and the synchronization signals are bundled with the broadcast channel to speed up the synchronization and camping speed.The number of physical layer cell Identification(ID)in the NR system increases,which is twice as much as that in LTE system.In terms of system measurement,NR uses beamforming to reduce mutual interference between signals.Beam selection and report are based on beam measurement result.Channel State Information Reference Signals(CSI-RS)is introduced for connection state measurement,and SSB is used for idle state measurement.The cell detection mainly includes the detection of the Primary Synchronization Signal(PSS),the detection of the Secondary Synchronization Signal(SSS)and the frequency offset estimation.This thesis studies the performance of different sequence detection methods and different frequency offset estimation methods.Due to the increase in the number of cell IDs in the NR system and the increase in the reference signal length,the traditional cross-correlation detection method increases in complexity and detection delay.At the same time,the NR system includes high-frequency offset scenarios.Therefore,this article evaluates three other sequence detection methods,including mutual-mirror correlation,self-mirror correlation and difference conjugate methods.The complexity of the above four methods is analyzed,and the performance of these methods is further verified based on simulation results.The mutual-mirror correlation algorithm has lower complexity and better detection performance,and it can replace the cross-correlation algorithm.The self-mirror correlation algorithm has the lowest complexity but it is greatly affected by noise,which is suitable for the situation where the channel is simple and the signal-to-noise is relatively low.The difference conjugate method can eliminate the influence of frequency deviation,but its complexity is high,which is suitable for the situation where the channel is simple and the frequency deviation is relatively high.Secondly,this thesis studies the frequency offset estimation method based on Cyclic Prefix(CP).In order to improve the performance of frequency offset estimation,this thesis proposes an SSB-based frequency offset estimation method and evaluates the performance of this method through theoretical derivation and simulation.It is concluded that the SSB-based frequency offset estimation method has higher accuracy but smaller estimation range.This thesis recommends to use SSB and cyclic prefix to estimate frequency offset.Finally,the performance of frequency offset combining method based on complex average and phase average are studied.The performance advantages of frequency offset combining method based on complex average are verified through simulation and theory.The system measurement performance indicators mainly include Reference Signal Received Power(RSRP),Reference Signal Received Quality(RSRQ),and Signal-to-Interference-plus-Noise Ratio(SINR).First,this thesis evaluates RSRP estimation methods based on Least Squares(LS)and adjacent frequency cross-correlation.The traditional LS-based evaluation results are greatly affected by noise.The measurement error of the adjacent frequency cross-correlation method is small and the average measurement accuracy isn't affected by the signal-to-noise ratio and the number of samples.Then,the interference noise power estimation method based on difference and protection bandwidth is evaluated.The method based on the difference has low complexity and it is suitable for situations where the signal-to-noise ratio is relatively low.The method based on protection bandwidth is not affected by the signal-to-noise ratio,but the complexity is higher than the method based on the difference,which is suitable for the situation where the signal-to-noise is relatively high.Next,this thesis studies the impact of multiple factors(including resource block density,subcarrier spacing,and UE-side averaging method)on the accuracy of these three indicators in common scenarios.The measurement performance such as cumulative distribution function curve of measurement accuracy is analyzed,and reference parameters with engineering significance are given.Finally,in the beamforming scenario,the ability of the UE to resolve different beams under the conditions of using different numbers of transmit antennas and different reference signals at the base station are analyzed.