OIRS-enhanced NOMA-VLC systems: optimization and experimental validation of sum data rate and user fairness

Optical intelligent reflecting surfaces (OIRS) provide a promising solution to enhance the performance of visible light communication (VLC) systems by dynamically adjusting the reflection paths. OIRS-assisted VLC systems can optimize signal strength and reduce interference, making them highly suitable for integration with non-orthogonal multiple access (NOMA) technology. This paper presents an OIRS-assisted NOMA-VLC system aimed at balancing achievable sum data rate and user fairness. A mathematical model is developed under the constraints of OIRS configuration, successive interference cancellation (SIC), and individual QoS requirements. To address this non-convex problem, we decompose it into two joint optimization subproblems using an alternating optimization algorithm: (i) OIRS-side allocation matrix optimization for user channel states, and (ii) transmitter-side LED power allocation matrix optimization for user power distribution. For the first subproblem, we propose a non-dominated sorting genetic algorithm (NSGA) enhanced with simulated annealing and an adaptive crossover and mutation strategy (NSGA-SACM). For the second subproblem, we introduce a decomposition-based multi-objective evolutionary algorithm (MOEAD) combined with a hill-climbing algorithm and an adaptive neighborhood weights strategy (MOEAD-HCANW). Numerical evaluations of system performance in an indoor environment demonstrate that the proposed algorithm significantly outperforms baseline methods in terms of bit rate and fairness, effectively improving user performance regardless of channel disparities. Notably, this work presents the first experimental validation of OIRS-assisted NOMA-VLC systems, and the experimental results align well with the simulations, collectively validating the effectiveness of OIRS in enhancing NOMA-VLC systems.