Research On Beamspace Channel Estimation And Hybrid Beamforming In Massive MIMO Systems

With the development of communication technology,millimeter wave(mm Wave)has become a key research direction in modern wireless communications.The penetration ability of mm Wave is relatively weak,so it attenuates seriously in the transmission process.To resist the signal fading,the multiple input multiple output(MIMO)technology is applied to mm Wave systems.However,it is unaffordable to configure radio frequency(RF)chain for each antenna in MIMO system in practice.In order to reduce the number of RF chains in MIMO system,lens antenna array has become a research focus.The lens antenna array is composed of an electromagnetic lens and an antenna array located on the focal plane of the lens.The lens can concentrate the incident electromagnetic wave from different directions on several antennas near the focus,thus it is only necessary to configure RF chains for these antennas with large signal power.Due to the small number of paths in the mm Wave physical channel,the number of incident angles of the signal is also small.Thus the beamspace channel converted to the angle domain by the lens is sparse.To recover useful data from the received signals,it is necessary to estimate the beamspace channel.Aiming at the problem of broadband channel estimation of orthogonal frequency division multiplexing(OFDM)in millimeter wave beamspace MIMO systems,a two-stage channel estimation scheme is proposed in this thesis.In the first stage,the multi-task sparse Bayesian learning(MT-SBL)is utilized to coarsely estimate the beamspace channel vectors.In the second stage,the channel estimation problem is reformulated by exploiting the sparsity of beamspace channels.And all parameters involved are estimated based on the expectation maximization(EM)algorithm to improve the estimation accuracy.Compared with the existing beamspace channel estimation algorithms,simulation results show that the proposed scheme can achieve much better performance while requiring a lower training overhead.In addition,compared with the traditional digital beamforming technology,hybrid beamforming technology can reduce the number of RF chains through partial analog precoders and combiners.For multi-user mm Wwave communication systems,this thesis designs a hybrid beamforming scheme aiming to minimize the bit error rate(BER)in the downlink data transmission.Because the BER performance is closely related to the mean square error(MSE)between the transmitted and received data,this thesis selects the sum of MSE of all users' transmitted and received data as the objective function.Then constructs precoding matrix and merging matrix respectively in the iterative process based on the minimum mean square error(MMSE)criterion.Simulation results show that this algorithm can obtain better BER performance compared with the existing hybrid beamforming algorithms.