Equalization for MIMO ISI systems using channel inversion under causality, stability and robustness constraints

The problem of linear equalization for frequency selective multiple-input multiple-output wireless channels, using channel inversion, is considered. While in the flat-fading case, the solution with minimal noise enhancement is readily found and is automatically optimal in terms of energy norm, the problem becomes intricate when intersymbol interference is present. In order to analyze this behavior, a framework, based on the semidefinite programming technique, is employed in the paper. Additionally, connections with possible analytical solutions based on the common system block matrix representation are established. The proposed framework is easily adopted for other linear equalization methods, pre-equalization, and can also be used for the feedforward filter construction in successive interference cancellation strategies.