Research On Limited Feedback And Its Related Key Techniques In MIMO Precoding Systems

Employing a variety of channel adaptive transmission techniques can achieve large improvements in almost any performance metric,when the transmitter has some form of the channel state information for multiple-input multiple-output(MIMO)wireless communication systems.In frequency division duplexing MIMO systems,limited feedback and its related techniques are important factors for improving systems performance,which have been extensively studied and well developed.The recent research in this area has begun to focus on temporally or spatially correlated channels and emerging hardware architectures in MIMO systems.This thesis mainly investigates the limited feedback and its related techniques both in multiple-input single-output(MISO)systems with temporally correlated channels and in millimeter wave(mmWave)massive MIMO precoding systems.The main contributions of the investigation can be summarized as follows.(1)To design efficient channel estimations for mm Wave massive MIMO systems with hybrid architectures,a chained-block sparse channel representation model is formulated by vectorizing the channel matrix in the virtual angles domain,when considering the real-world angular speads of scattering clusters.Then,we propose a matrix decomposition-based preprocessing algorithm to further formulate the training-based channel estimation as a block-sparse signal reconstruction problem.Furthermore,a novel channel estimation algorithm based on block sparse Bayesian Learning is proposed,and an estimation method for relevant parameters of each nonzero block is designed.The simulation results show that this channel estimation algorithm is more suitable for low signal-to-noise ratio scenarios than the existing algorithms,i.e.,the structured sparsity of mmWave massive MIMO channels can be utilized to obtain better systems performance,where using the normalized mean squared error and theoretical spectral efficiency as performance metrics.(2)To design the enhanced quantization codebooks for hybrid precoding in mmWave massive MIMO systems,it is shown that optimal precoders for all data streams exhibit a hidden joint sparsity structure caused by the limited number of scattering clusters with low angular spreads of mmWave MIMO channels.Then,the codebooks design is formulated as a joint optimization problem,and an iteration algorithm is proposed based on dictionary learning methods,where all codewords can be optimized in turn in every iteration step.Furthermore,a jointly sparse hybrid precoding is proposed as the pursuit method for the proposed codebooks design algorithm.The simulation results demonstrate that the near-optimal performance can be achieved by the joint utilization of the proposed hybrid precoding algorithm and codebooks even when the low-resolution phase shifters are used.(3)For hybrid precoding technique based on the analog codebook composed of array response vectors in mmWave MIMO systems,one effective way for better performance is to increase the number of RF chains.With the increase of radio frequency chains,the optimal precoder exhibits the block sparsity structure over the analog codebook.Consequently,we formulate hybrid precoders design as a block-sparse reconstruction problem,and propose a greedy clustering-based precoding algorithm.Moreover,an estimation method for the number of the nonzero blocks and RF chains is presented based on the the relationship among all scattering clusters.The simulation results demonstrate that the proposed algorithm can save the feedback overhead while achieving the near-optimal system performance.(4)For the adaptation scheme in temporally correlated MISO channels,two design methods of differential spherical cap codebooks are presented.For the method one,each spherical cap codebook is generated by linear operations and rotation of codewords in the basis codebook such as Grassmannian codebooks.For the method two,each spherical cap codebook is composed of the center vector and codewords distributed on two polar-caps corresponding to the different radii,so called double polar-caps codebook.Moreover,two methods of estimating the radii of differential codebooks are proposed.For the first codebooks method,we design the calculation method of the spherical cap radius for all differential codebooks based on their generation approaches,which can tend to stable more quickly.The simulation results show that the adaptive differential feedback strategy based on the proposed codebooks can achieve improved performance compared with existing schemes.