Research On Channel Estimation In Millimeter-Wave Massive MIMO Systems

With the rapid development of global mobile internet technology,the demand for data information is increasing.Solving the problem of scarce spectrum resources has become a major focus in current research on wireless communication technology.Due to its abundant bandwidth and potential spectrum resources,millimeter-waves have received widespread attention in wireless communication research.Meanwhile,by using Massive MIMO technology for beamforming of multiple antennas,it provides greater spatial freedom while effectively resisting severe path loss of millimeter-waves.Therefore,millimeter-wave massive MIMO has become a key technology for fifth-generation(5G)and even future sixth-generation(6G)wireless communication systems.However,due to the precision of current communication devices,millimeter-wave massive MIMO systems face a series of challenges such as high power consumption,high complexity,and expensive hardware costs.Traditional Analog-toDigital Converters(ADCs)with full resolution are no longer applicable to Massive MIMO technology.Studying how to balance system performance and hardware costs while designing efficient and reasonable channel estimation algorithms has attracted widespread attention from academia.This thesis focuses on the channel estimation problem in millimeter-wave massive MIMO systems for different system architectures and application scenarios.The specific research content is as follows:(1)To address the high power consumption and expensive hardware cost issues in millimeter-wave massive MIMO systems,this thesis studies the channel estimation algorithm under the mixed-ADC architecture.Specifically,most antennas are equipped with lowresolution ADCs,while the remaining few antennas use high-resolution ADCs.By leveraging the sparsity of millimeter-wave channels,the beam space channel estimation problem is converted into a sparse matrix recovery problem,and then the channel is recovered through compressed sensing(CS)technology.Simulation results show that compared with the traditional low-resolution ADC architecture,the algorithm’s performance under the mixedADC architecture is significantly improved.Furthermore,when the low-resolution ADC in the mixed-ADC architecture reaches 5 bits,the mixed-ADC can achieve performance close to that of high-resolution ADC.(2)Considering that lens antenna arrays can reduce system power consumption by reducing the use of Radio Frequency(RF)chains,this thesis studies the channel estimation algorithm in millimeter-wave massive MIMO systems equipped with lens antenna arrays at the base station.By using CS technology to convert the beam space channel estimation problem into an optimal l-norm problem,deep learning(DL)technology is introduced to optimize the variable parameters in traditional algorithms.Simulation results show that the DL-based network solution outperforms traditional estimation schemes,and when the Normalized Mean Square Error(NMSE)is about-23 d B,the rate loss of incomplete Channel State Information(CSI)under the mixed-ADC scheme is less than 5% compared to the complete CSI scheme.(3)To address the problem of direct link blocking between the base station and users,an intelligent reflection surface(IRS)is deployed at the base station to provide indirect communication services.This thesis proposes a row-structured sparsity-based Orthogonal Matching Pursuit(RS-OMP)algorithm based on the special row-column sparse characteristics of cascaded channels for an IRS-assisted millimeter-wave massive MIMO system.The algorithm needs to sequentially calculate the row-column support sets of angular cascaded channels,and then reconstruct the cascaded channel matrix through least squares(LS)algorithm to achieve lower computational complexity.Simulation results show that compared with the OMP algorithm,the proposed RS-OMP algorithm not only improves the channel estimation accuracy but also reduces more than 75% of the pilot overhead.