Low-Resolution RIS-Aided Multiuser MIMO Signaling

A multi-antenna aided base station (BS) supporting several multi-antenna downlink users with the aid of a reconfigurable intelligent surface (RIS) of programmable reflecting elements (PREs) is considered. Low-resolution PREs constrained by a set of sparse discrete values are used for reasons of cost-efficiency. Our challenging objective is to jointly design the beamformers at the BS and the RIS's PREs for improving the throughput of all users by maximizing their geometric-mean, under a variety of different access schemes. This constitutes a computationally challenging problem of mixed continuous-discrete optimization, because each user's throughput is a complicated function of both the continuous-valued beamformer weights and of the discrete-valued PREs. We develop low-complexity algorithms, which iterate by directly evaluating low-complexity closed-form expressions. Our simulation results show the advantages of non-orthogonal multiple access-aided signaling, which allows the users to decode a part of the multi-user interference for enhancing their throughput.

Automated summary

This paper considers a multi-antenna aided base station (BS) that supports multiple multi-antenna downlink users using a reconfigurable intelligent surface (RIS) with programmable reflecting elements (PREs) to improve throughput. The objective is to jointly design the beamformers at the BS and the RIS's PREs by maximizing the geometric-mean of all users' throughput under various access schemes. The paper develops low-complexity algorithms and shows the advantages of non-orthogonal multiple access-aided signaling.