Millimeter Wave Massive MIMO Channel Estimation And Its Hybrid Precoding

In this paper,millimeter wave(mm Wave)massive multiple input multiple output precoding and channel estimation are studied.The detailed work is concluded as follows.A variable-phase-shifter(VPS)architecture with hybrid precoding for mixture use of phase shifters and switches,is proposed considering a large number of phase shifters are used in the fully-connected architecture.For the VPS architecture,a hybrid precoding design(HPD)scheme,called VPS-HPD,is proposed to optimize the phases according to the channel state information by alternately optimizing the analog precoder and digital precoder.To reduce the computational complexity of the VPS-HPD scheme,a low-complexity(LC)HPD scheme for the VPS architecture called VPS-LC-HPD including alternating optimization in three stages is then proposed,where each stage has a closed-form solution.Except for the reduction of computational complexity,decreasing the hardware complexity is also considered.By further reducing the number of used switches,a group-connected VPS architecture is studied and its HPD scheme is proposed,where the HPD problem is divided into multiple independent subproblems with each subproblem flexibly solved by the VPS-HPD or VPS-LC-HPD scheme.Simulation results verify that the proposed schemes can achieve better spectral efficiency performance than the fully connected architecture on the basis of significantly reducing the computational complexity and hardware complexity.The use of channel estimation schemes to obtain channel state information under the VPS architecture is also considered.A channel estimation scheme based on sequential and adaptive subspace estimation(SASE)is mainly studied,which transforms the channel matrix to be estimated into its column subspace estimation and row subspace estimation,respectively.In the column subspace estimation stage,by optimizing the design of the hybrid precoder and the hybird combiner,the number of channel measurements can be reduced,and the hybrid combining matrix can be optimized after estimating the column subspace to improve the estimation performance.In the row subspace estimation stage,the designed receiver hybrid combining matrix is used to generate the receiver signal,and after the row subspace is obtained,the design of the hybrid precoding matrix is optimized.Finally,a low-dimensional matrix can be solved to obtain the channel matrix.The simulation results show that the SASE scheme in this paper can achieve better normalized mean squared error performance than the least squares channel estimation scheme with lower computational complexity and smaller number of channel measurements.