Rate-Maximized Zero-Forcing Beamforming for VLC Multiuser MISO Downlinks

In this study, we aim to maximize the achievable rate for a visible light communication (VLC) downlink multiuser multiple-input-single-output (MISO) system deployed with zero-forcing (ZF) beamforming. By incorporating the ZF, transmit signal's nonnegativity, and light-emitting diode (LED) maximum permissible current constraints, we formulate the rate maximization problem of interest, which possesses a complicated nonconvex form. To handle this challenging problem, we first employ the powerful concave-convex procedure to develop an iterative algorithm that converges to a locally optimal solution. Subsequently, by exploiting the high signal-to-noise ratio (SNR) characteristic in VLC, we simplify the primal nonconvex design problem into a solvable convex one that admits a high-quality solution. As an alternative approach, we impose a channel pseudoinverse structure on the beamforming matrix followed by power allocation optimization. Finally, we carry out simulations to investigate the performance of the proposed schemes under a VLC scenario.