Cooperative Beamforming Design for Multiple RIS-Assisted Communication Systems

TL;DR

This paper proposes a cooperative beamforming algorithm for multi-RIS-assisted communication systems, optimizing sum rate by effectively coordinating multiple RISs, and demonstrates that distributing reflecting elements among multiple RISs outperforms single-RIS deployment.

Contribution

It introduces a novel cooperative beamforming design for multiple RISs, including a low-complexity algorithm and insights into optimal element distribution.

Findings

Distributed RISs outperform single RIS in sum rate.

The proposed algorithm achieves effective beamforming with low complexity.

Simulation results confirm the advantages of cooperation among multiple RISs.

Abstract

Reconfigurable intelligent surface (RIS) provides a promising way to build programmable wireless transmission environments. Owing to the massive number of controllable reflecting elements on the surface, RIS is capable of providing considerable passive beamforming gains. At present, most related works mainly consider the modeling, design, performance analysis and optimization of single-RIS-assisted systems. Although there are a few of works that investigate multiple RISs individually serving their associated users, the cooperation among multiple RISs is not well considered as yet. To fill the gap, this paper studies a cooperative beamforming design for multi-RIS-assisted communication systems, where multiple RISs are deployed to assist the downlink communications from a base station to its users. To do so, we first model the general channel from the base station to the users for arbitrary number of reflection links. Then, we formulate an optimization problem to maximize the sum rate of all users. Analysis shows that the formulated problem is difficult to solve due to its non-convexity and the interactions among the decision variables. To solve it effectively, we first decouple the problem into three disjoint subproblems. Then, by introducing appropriate auxiliary variables, we derive the closed-form expressions for the decision variables and propose a low-complexity cooperative beamforming algorithm. Simulation results have verified the effectiveness of the proposed algorithm through comparison with various baseline methods. Furthermore, these results also unveil that, for the sum rate maximization, distributing the reflecting elements among multiple RISs is superior to deploying them at one single RIS.