Maximizing the Geometric Mean of User-Rates to Improve Rate-Fairness: Proper vs. Improper Gaussian Signaling

This paper considers a reconfigurable intelligent surface (RIS)-aided network, which relies on multiple antenna array aided base station (BS) and an RIS for serving single downlink users. To provide reliable links to all users over the same bandwidth time-slot, proposes joint design of linear transmit beamformers programmable reflecting coefficients maximize geometric mean (GM) users’ rates. A new computationally efficient alternating descent algorithm is developed, based closed-forms only generating improved feasible points this nonconvex problem. We also consider widely further improve GM Hence another developed its solution, closed forms points. Numerical examples are provided demonstrate efficiency proposed approach.