Research On Hybrid MIMO Uplink Wireless Transmission Technology In 5G NR

In order to enhance the wireless data transmission rate in wireless communication networks,the physical layer primarily relies on increasing the number of operating frequency points,such as millimeter wave and terahertz frequency bands,to obtain a broader communication bandwidth.Additionally,increasing the number of base station antennas and the number of base stations can provide high multiple-input multiple-output(MIMO)multiplexing diversity gain.However,the massive data of the fifth generation(5G)/sixth generation(6G)mobile communication has led to significant concerns regarding power consumption.Consequently,the deployment of 5G/6G applications presents significant challenges in terms of power consumption,operating expenses,and capital expenditure(OPEX/CAPEX).The use of low-power and low-cost nonlinear devices,such as envelope detectors,phase detectors,and low-precision analog-to-digital converters(ADCs),in the design of nonlinear MIMO receivers can effectively reduce the circuit power of the base station.This approach can alleviate the power consumption and OPEX/CAPEX issues faced by 5G/6G mobile communications.However,the introduction of nonlinear devices can cause the traditional ideal In-phase/Quadrature(I/Q)complex baseband observation to suffer from various nonlinear/linear distortions,including the loss of phase/amplitude observation information and ADC quantization distortion.Signal processing algorithms for traditional communication are typically designed for classical I/Q complex baseband observations and cannot be applied to nonlinear MIMO scenarios.As a result,the development of a new communication signal processing algorithm is necessary for the latter.This paper mainly focuses on the following research on the signal processing issues in two heterogeneous MIMO schemes including only periodic phase(Halved Phase-Only,HPO-)MIMO and amplitude-only(Magnitude-Only,MO-)MIMO:1.Aiming at the problems of complex receiving antenna circuits and high power consumption under the commercial background of 5G New Radio(NR),a heterogeneous MIMO receiving antenna array is designed.The heterogeneous(hybrid-)MIMO receiving antenna array combines traditional MIMO antennas and nonlinear MIMO antennas.The simulation results verify the effectiveness of the heterogeneous MIMO system.The performance of the proposed hybrid-MIMO system can be greatly improved with the increase of the number of nonlinear MIMO antennas,and the circuit power consumption is significantly reduced compared with the traditional MIMO receiving antenna architecture.2.Aiming at the channel estimation(CE)problem in heterogeneous MIMO systems in 5G NR,the sparse time domain structure of wireless channels is applied to this problem,thus improving the performance of channel estimation in uplink communication systems.Firstly,the problem of structured sparse channel estimation in heterogeneous MIMO systems is summarized as a generalized linear mixed problem under mixed phase/amplitude observation,which is solved by the generalized approximate message passing(GAMP)algorithm.An EM-BP-GAMP channel estimator is proposed to improve the GAMP algorithm by combining the expectation maximization(EM)mechanism and the belief propagation(BP)algorithm.The prior distribution of the sparsely structured channel is modeled as Bernoulli Gaussian(BG)model with unknown parameters.The channel response and unknown parameters(including BG parameters and noise variance)will be updated alternately through GAMP and EM,respectively.The simulation results verify that the EM-BP-GAMP channel estimator can improve the performance of channel estimation by taking advantage of the sparsity of the time domain structure of the wireless channel.3.Aiming at the problem of coherent multiuser detection(Multiuesr Detection,MUD)in the heterogeneous MIMO system in 5G NR,an improved GAMP algorithm is proposed.First,the MUD problem is transformed into a problem of recovering user data from generalized linear mixed models under mixed phase/magnitude observations.Then this problem is solved by the improved GAMP(DCD-GAMP)algorithm with Dropout(D)mechanism and Constellation deletion(CD)mechanism.Considering the limited pilot resources in 5G NR,the limited number of receiving observations,and the inaccurate channel estimation,since the performance of coherent detection is sensitive to the performance of channel estimation,we adopt the parameter bilinear generalized approximate message passing based on the D and CD mechanism(Parametric Bilinear Generalized Approximate Message Passing,P-Bi-GAMP)joint channel estimation and multi-user detector(Joint Channel-and-data Estimation,JCDE)to ensure multi-user detection performance.Simulation results show that the proposed coherent and joint multiuser detection can effectively deal with amplitude/phase quantization loss for reconstructing user data.