Adaptive Interference Management for Enhancing RIS-Assisted NOMA Systems Through Satellite–Terrestrial Links

This paper proposes an elevation-aware interference management framework for reconfigurable intelligent surface (RIS)-assisted downlink non-orthogonal multiple access (NOMA) in integrated low Earth orbit (LEO) satellite–terrestrial networks. In this setting, the strength of satellite interference varies strongly with the elevation angle due to path loss and shadowing. We model this effect via an elevation-dependent weighting factor$\alpha (\theta)$, expressed as a sigmoid function and updated in real time by a lightweight proportional–integral (PI) controller driven by a scalar performance metric. The resulting$\alpha (\theta)$enables an adaptive successive interference cancellation (SIC) strategy that prioritizes decoding of satellite or terrestrial components according to the instantaneous interference level and omits satellite decoding when its contribution is weak, thereby reducing unnecessary SIC stages. In parallel, a three-mode RIS codebook (terrestrial-priority, balanced, satellite-priority) is selected based on$\alpha (\theta)$and refined online using low-complexity gradient-based phase updates. Simulations under 3GPP-compliant parameters show that the proposed scheme achieves up to 20% higher throughput than a conventional static RIS–NOMA baseline with non-adaptive RIS phases and a fixed channel-gain-based SIC order that does not exploit the elevation-dependent interference weight$\alpha (\theta)$, while maintaining robust performance across a wide range of elevation angles. These results highlight the benefits of combining elevation-aware RIS control, adaptive SIC ordering, and low-overhead online RIS refinement in next-generation satellite–terrestrial NOMA networks.