Energy-Efficient OIRS-Aided VLC Systems Employing ML-Based User Orientation and Obstacle Awareness

Abstract

Visible light (VL) for optical wireless communication (OWC) in indoor environments is a promising approach to achieving high-speed and energy-efficient (EE) wireless connectivity and sensing. Unlike traditional radio frequency (RF) systems, the OWC channel is not isotropic, making the orientation of the user equipment (UE) and the existing obstacles significant factors affecting the link gain and EE in VL communication (VLC) systems. Specifically, VLC performance degrades when users or obstacles block the line-of-sight (LoS) link, and the received power fluctuates due to random UE orientation in non-LoS (NLoS) scenarios. This article tackles these challenges by proposing an EE framework that integrates optical intelligent reflecting surfaces (OIRSs) in indoor VLC systems while considering users as obstacles causing link blockage. The proposed framework introduces a machine learning (ML)-based mechanism to align OIRS elements with UE directions. To this end, we mathematically formulate the EE maximization problem in OIRS-aided indoor VLC systems in the presence of obstacles while considering a set of design and quality of service (QoS) constraints. The transmit power, the offset management, and the modulation size are jointly optimized under the transmit signal nonnegativity, the optical and electrical powers, the minimum rate demand, and the bit error rate (BER) constraints. The formulated EE optimization problem is demonstrated in a semiclosed-form solution. The simulation results confirm the performance superiority of the proposed design compared to existing designs.