Title:Dynamic Energy-Saving Design for Double-Faced Active RIS Assisted Communications with Perfect/Imperfect CSI

Abstract:Although the emerging reconfigurable intelligent surface (RIS) paves a new way for next-generation wireless communications, it suffers from inherent flaws, i.e., double-fading attenuation effects and half-space coverage limitations. The state-of-the-art double-face active (DFA)-RIS architecture is proposed for significantly amplifying and transmitting incident signals in full-space. Despite the efficacy of DFA-RIS in mitigating the aforementioned flaws, its potential drawback is that the complex active hardware also incurs intolerable energy consumption. To overcome this drawback, in this paper we propose a novel dynamic energy-saving design for the DFA-RIS, called the sub-array based DFA-RIS architecture. This architecture divides the DFA-RIS into multiple sub-arrays, where the signal amplification function in each sub-array can be activated/deactivated dynamically and flexibly. Utilizing the above architecture, we develop the joint optimization scheme based on transmit beamforming, DFA-RIS configuration, and reflection amplifier (RA) operating pattern to maximize the energy efficiency (EE) of the DFA-RIS assisted multiuser MISO system considering the perfect/imperfect channel state information (CSI) case. Then, the penalty dual decomposition (PDD) based alternating optimization (AO) algorithm and the constrained stochastic majorization-minimization (CSMM) based AO algorithm address non-convex problems in the perfect/imperfect CSI case, respectively. Simulation results verified that our proposed sub-array based DFA-RIS architecture can benefit the EE of the system more than other RIS architectures.