A mixed-integer programming approach to interference exploitation for massive-MIMO

Abstract

A novel low-complexity transmission scheme for Massive Multiuser Multi-Input Multi-Output (M-MU-MIMO) is proposed, where Transmit Antenna Selection (TAS) and beam-forming are jointly performed to exploit multiuser interference. Two separate solutions to the deriving optimization problem are proposed: a mixed-integer programming approach that can optimally solve the TAS-beamforming problem and a heuristic convex approach, based on the assumption of matched filtering beamforming. Numerical results prove that the proposed multiuser interference exploiting approaches are able to greatly outperform previous state-of-the-art schemes, where TAS and beamforming are disjointedly solved.