Low Complexity Beam Selection for Sparse Massive MIMO Systems

International audience; In this work, we propose a beam selection scheme that exploits the geometric sparsity of the multi-user (MU) massive multiple input and multiple output (MIMO) channel, using its beamspace representation. Assuming knowledge of the beamspace channel, the beamspace precoder consists in selecting and weighting the beams steered to the users, in order to maximize the signal-to-interference-and-noise ratio (SINR) at the user equipment (UE). Low complexity solutions are proposed through three simple heuristics. In a first step, the heuristics use the maximal ratio transmission (MRT) principle to determine the beam gains. In a second step, the beams are selected to enhance the SINRs. For all heuristics, we solve this problem in two steps i) selection of beams ii) power allocation. We show that using MRT as base, adding and/or removing some beams improve the system performance. Simulation results show that, compared to the linear MRT precoder, the proposed heuristics can improve the performance under a scenario with channel sparsity.