# Multi-user Downlink with Reconfigurable Intelligent Metasurface Antennas (RIMSA) Array

arXiv: https://arxiv.org/abs/2506.18418 · HTML: https://mlsanity.com/paper/2506.18418/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 23 Jun 2025.

## Built on by (most cited first, 1 of 1)

- [Anti-Jamming Sensing with Distributed Reconfigurable Intelligent Metasurface Antennas](https://mlsanity.com/paper/2508.04964/) (2025)

## Abstract (arXiv)

Reconfigurable Intelligent Surfaces (RIS) is a transformative technology with great potential in many applications in wireless communications and realizing the Internet of Everything at sixth generation (6G). In this study, we propose a wireless system where the RIS acts as an antenna, which we call Reconfigurable Intelligent Metasurface Antennas (RIMSA). In particular, the base station (BS) equipped with a RIMSA array performs downlink transmissions to multiple users, where each user has a single or multiple RIMSA/RF links, and we aim to solve the sum-rate maximization problem by jointly optimizing the digital processing matrix of the transceivers and the phase responses of RIMSA array at both BS and users. For the multi-user multiple-input single-output (MU-MISO) scenario, we develop an alternating optimization algorithm to slove the problem, where a fractional programming (FP) is used to optimize the digital processing matrix and a product manifold optimization (PMO) is proposed to provide the optimal phase responses of the RIMSA array at both BS and users. For the multi-user multiple-input multiple-output (MU-MIMO) scenario, we equate it to a weighted sum of mean square errors minimization problem, which can be solved by three subproblems iteratively. Both the optimal digital precoder subproblem and the optimal digital combiner subproblem have closed-form solutions, and the subproblem of RIMSA configuration is solved by the PMO algorithm as well. Simulation results demonstrate that the proposed algorithms achieve significant performance gains over conventional algorithms.