Receive Spatial Modulation in Correlated Massive MIMO With Partial CSI

Spatial modulation is emerging as a promising signal transmission paradigm for future wireless communications relying on massive multiple-input multiple-output (MIMO) technology. In this paper, the receive spatial modulation (RSM) is applied to the downlink of multi-user MIMO systems with the base station equipped with a massive array of antennas. Due to the significant overhead associated with acquiring channel state information at the transmitter (CSIT) of frequency-division-duplex massive MIMO systems, herein we propose a preprocessing scheme for correlated channels that reduces the channel estimation overheads. The proposed preprocessing method relies on the partial CSIT and consists of two stages of beamforming and precoding. The beamforming is accomplished via block-diagonalization to combat multi-user interference while the precoding matrix is designed by formulating optimization problems aiming to eliminate inter-antenna interference and enhancing RSM transmission performance. The design criteria of the problems include bit-error-rate minimization and discrete-input continuous-output (DCMC) capacity maximization with a transmit power constraint. The arising non-convex problems are cast as semi-definite and second-order cone programs that can be solved efficiently. The proposed schemes are readily extended to generalized RSM as the most comprehensive case in RSM-MIMO scenarios. Simulation results demonstrate that the proposed schemes outperform the imperfect-CSIT based and two-stage zero-forcing procedures and offer similar or significant performance compared to conventional MIMO counterparts at low to mid signal-to-noise ratio regimes, depending on the configurations.