Performance Analysis And Detection Design Of Massive MIMO Systems With Superimposed Pilots And Low-Resolution ADCs

With the rapid increase of mobile communication traffic,future mobile communication systems need to support higher data transmission rates.In the case of increasingly tight spectrum resources,large-scale multiple input multiple output(MIMO)technology can fully exploit spatial degree of freedom,significantly improve spectrum efficiency,and energy efficiency,reduce delay and improve transmission reliability,etc.,and thus become one of the key technologies of the 5th Generation of Mobile Communication Systems(5G).However,the use of large-scale antenna arrays will cause problems such as expensive hardware costs,high system power consumption,and complex data processing.Using low-resolution analog-to-digital converters(ADC)can effectively solve the above problems.However,low-resolution ADCs introduce serious quantization noise,making accurate channel estimation difficult.Under the superimposed pilots(SP)method,users send pilot and data signals at the same time,which can effectively overcome the low bandwidth utilization and low data transmission efficiency faced by traditional timemultiplexed pilots(TP).And the length of the superimposed pilot is significantly increased,which helps to improve the quality of channel estimation.This thesis researches the massive MIMO system with low-resolution ADCs and superimposed pilots,focusing on the impact of superimposed pilots on the performance and channel estimation of low-resolution quantized massive MIMO systems.On the one hand,the performance of massive MIMO systems with low-resolution ADCs and superimposed pilots is researched.Consider removing the pilot in the data detection stage,the approximated expression of the uplink achievable rate is derived when the BS adopts Minimum Mean Square Error(MMSE)channel estimation and Maximum Ratio Combining(MRC)detection.optimal pilot power factor which maximizes the achievable rate is deduced.Consider the power consumption of the system and the expression of Energy Efficiency(EE)is derived.In addition,the achievable rate and optimal power allocation strategy of the system under various asymptotic limits are analyzed.Results show that systems with higher-resolution ADCs or larger number of base station(BS)antennas need to allocate more power to pilots.In contrast,more power needs to be allocated to data when the channel is slowly varying.The achievable rate and energy efficiency of SP and TP under different quantization bits are compared.Numerical results show that in the low signal-to-noise ratio(SNR)region,for 1-bit quantizers,SP outperforms time-multiplexed pilots(TP)in most cases,while for systems with higher-resolution ADCs,SP scheme is suitable for the scenarios with comparatively small number of BS antennas and relatively long channel coherence time.On the other hand,the problem of data and pilot interference in the SP method is studied to further improve the channel estimation and data detection performance.Through theoretical derivation and analysis,it is shown that when the traditional data-assisted iterative channel estimation method uses the LS algorithm for channel estimation,the iterative algorithm will face the problem of non-convergence,resulting in the estimation performance of the channel and data unable to increase with the number of iterations.Based on this,this thesis optimizes the design of the algorithm,and proposes a joint channel estimation and data detection algorithm.The simulation results show that this algorithm has the best number of iterations in special scenarios,such as 1-bit quantization scenarios,high-speed moving scenarios,and low SNR scenarios.By analyzing the influence of SNR,quantization resolution,channel coherence time,and pilot power factor on the performance improvement degree of the optimization algorithm,it can be found that the algorithm can effectively improve the channel estimation and data detection performance under the SP method.The specific conclusions are as follows: In high SNR scenarios,the performance advantage of this algorithm is more significant,which is beneficial to improve the performance of SP relative to TP;the performance gap between optimized and unoptimized algorithms is gradually obvious with the increase of quantization resolution,and finally keep unchanged;the increase of the channel coherence time can gradually improve the advantages of the optimization algorithm in channel estimation;for the data estimation performance,there is an optimal pilot power factor that can maximize the advantages of the algorithm.