On beamformer design for MIMO interference channels

Abstract

Multi-input multi-output (MIMO) architecture is known to increase the capacity of the single-user communication systems significantly through spatial multiplexing. In the case of multi-user scenarios, multiple antennas provide various ways of boosting the total throughput of the system. Specifically, coordinated beamforming utilizing multiple transmit and/or receive antennas is a promising technique to achieve the maximum benefit from MIMO interference channels. However, it still remains unknown how one can extract the full advantages provided by multiple antennas under the MIMO interference channels. In this thesis, we have investigated various beamformer design strategies and its optimality under MIMO interference channels. In the first part of the thesis, several linear beamformer design paradigms were considered. Notably, interference alignment and sum-rate maximizing algorithms such as the maximum signal-to-interference-plus noise (max-SINR) algorithm are considered. Optimal linear beamforming under interference alignment consists of two layers; an inner precoder and decoder (or receive filter) accomplish interference alignment to eliminate inter-user interference, and an outer precoder and decoder diagonalize the effective single-user channel resulting from the interference alignment by the inner precoder and decoder. The relationship between this two-layer beamforming and the max-SINR algorithm is established at high signal-to-noise ratio. Also, the optimality of the max-SINR algorithm within the class of linear beamforming algorithms, and its local convergence with exponential rate, are established at high signal-to-noise ratio. In the second chapter of the thesis, the Pareto-optimal beam structure for multi-user MIMO interference channels is investigated and a necessary condition for any Pareto-optimal transmit signal covariance matrix is presented for the $K$-pair Gaussian MIMO interference channel. It is shown that any Pareto-optimal transmit...