Fairness-aware resource allocation for FSO/RF non-terrestrial networks

Non-Terrestrial Networks enable wide-area connectivity in remote and underserved regions. However, the rapid growth of Internet of Things applications poses challenges in achieving high transmission capacity while maintaining fairness under heterogeneous channel conditions. Conventional sum-rate maximization prioritizes users with strong channels, potentially causing user disconnection and a capacity–connectivity trade-off under severe impairments. To address this, this paper investigates transmission capacity enhancement and fairness-aware resource allocation in a hybrid space–air–ground NTN architecture, ensuring fairness in terms of user connectivity under an SNR constraint. Sparse Code Multiple Access (SCMA) and Rate-Splitting Multiple Access (RSMA) are employed to improve spectral efficiency and support massive connectivity. In addition, a Fairness-Aware Geometric Water-Filling (FAGWF) algorithm is proposed, extending conventional GWF with an SNR-based minimum power constraint to ensure user connectivity while maintaining near-optimal capacity. The system is evaluated under realistic radio frequency and free-space optical channel models, including Nakagami-m fading and path loss for RF links and Málaga turbulence with attenuation and pointing errors for FSO links. Simulation results demonstrate that Rate-Splitting Multiple Access outperforms Sparse Code Multiple Access in terms of energy efficiency and sum data rate, achieving up to 19% improvement in RF links and approximately 15% higher transmission capacity in FSO links, due to more efficient interference management via rate-splitting. Furthermore, the Fairness-Aware Geometric Water-Filling algorithm achieves performance close to GWF while ensuring user connectivity under SNR constraints, mitigating user outage under severe fading with marginal performance loss. These results show that near-optimal capacity can be maintained without user disconnection, even under severe fading.