(A) study on beamforming and limited feedback for efficient usage of channel state information in multiple antenna systems

Abstract

This dissertation studies beamforming and limited feedback schemes for efficient usage of channel state information (CSI) in multiple antenna systems. First of all, dimension-optimized two-stage precoder design is proposed to reduce feedback overhead in massive multiple-input multiple-output (MIMO) systems with limited feedback environment. To maximize average sum rate, dimension optimization schemes respectively based on the lower bound on the average sum rate and machine learning framework with deep neural network are proposed. Also, dimension-optimized two-stage precoder to maximize the average sum signal-to-leakage plus noise power ratio is proposed. Next, novel codebook design for MIMO Rician fading channels that is adaptive to any Rician factor is proposed. By analyzing the distribution of the angle between the Rician channel and the LoS component, a band-structured non-homogeneous codebook is proposed. The design parameters of the proposed codebook are optimized to minimize the upper bound of the quantization error, and the construction of the proposed codebook with the determined parameters is elaborated on. Finally, multi-user energy beamforming to minimize the overall charging time for energy harvesting users with different energy requirements is proposed in radio frequency (RF)-enabled wireless energy transfer systems. To lessen the cost and power consumption by reducing the required RF chains, a method to figure out low-rank optimal energy beamforming is developed by analyzing the upper bound on the rank of a minimum-rank solution.