Optimizing Energy Efficiency in Large RIS-Aided Multi-User MISO Communication Systems

In this paper, we address a Multi-User Multiple-Input Single-Output (MU-MISO) communication system enhanced by a Large Reconfigurable Intelligent Surface (LRIS). The LRIS significantly improves beamforming gain and communication reliability, boosting system performance through increased signal manipulation flexibility due to its extensive number of reflecting elements. However, the deployment of a very large RIS brings challenges such as high power consumption and complexity in element management, necessitating a balance between size benefits and energy efficiency (EE). To mitigate these challenges, we propose partitioning the LRIS into smaller sub-RIS units. This approach, which involves beamforming and collaboration among multiple sub-RISs, aims to enhance scalability while maintaining EE. Unlike previous studies where all elements are continuously active, our method introduces a novel sub-RIS on-off mechanism for flexible activation and deactivation. We formulate an optimization problem to maximize total EE and use alternating optimization-based iterative algorithms, along with quadratic transform, variable substitution, and convex approximation methods, to derive sub-optimal solutions. Our analysis demonstrates that the sub-RIS architecture results in minimal performance degradation and significantly improves energy efficiency through flexible activation and deactivation of sub-RIS units.