MIRROR: Optimal RIS Deployment for Multi -Ris, -ReflectiOn, and -useR 6G Communications

Abstract

Reconfigurable intelligent surfaces (RISs) are promising for revolutionizing 6G networks by enhancing coverage, capacity, and reliability.However, the full potential of RISs can be realized only through their optimal deployment, as they directly affect channel conditions and overall network performance. Instead, suboptimal RIS deployment may result in inefficient resource utilization, diminished user experience, and reduced network efficiency. Moreover, it is crucial to consider scenarios with multiple users and RISs, as such complex environments are increasingly prevalent in modern wireless networks. This work addresses the challenge of optimizing RIS deployment in multi-user, multi-RIS, and multi-reflection scenarios. We propose a novel framework for RIS deployment, called MIRROR (MultI-Ris, -ReflectiOn, and -useR), which is designed to optimize channel gain and, consequently, improve data transmission rates. We then design a graph-based heuristic that reformulates the deployment problem as a shortest-path optimization, thereby balancing computational complexity and performance. Finally, we leverage a sensor fusion approach to determine RIS deployment that performs well across various user distributions. Our results demonstrate that our proposal significantly outperforms the state-of-the-art, achieving substantial data-rate gains under the considered settings.