CSI-Aware Adaptive PSO-Based Power Allocation for Enhancing SIC Reliability in Imperfect-CSI NOMA Systems
Future wireless networks must accommodate massive connectivity under stringent spectral- and energy-efciency requirements,thereby intensifying the demand for robust power-domain nonorthogonal multiple access (NOMA). The performance of NOMAcritically depends on reliable successive interference cancellation (SIC), whose efectiveness deteriorates under imperfect channelstate information (CSI) due to estimation errors, mobility, and hardware nonlinearities. To mitigate these challenges, this paperproposes an adaptive particle swarm optimization (PSO)-based power allocation framework that enhances SIC robustness bydynamically learning optimal power coefcients in the presence of CSI uncertainty. The problem is reformulated as a constrainedmultiobjective optimization task that incorporates decoding-order stability, interference propagation efects, user fairness, and rateperformance. The proposed adaptive PSO integrates CSI uncertainty directly into its update dynamics through error-awareadjustment of inertia and acceleration parameters. Extensive numerical evaluations conducted under realistic imperfect-CSIassumptions show that the proposed method signifcantly improves fairness, sum-rate stability, and SIC error resilience, achievingup to 98% successful SIC operations in medium-to-high signal-to-noise ratio (SNR) regimes. These fndings demonstrate thepracticality of lightweight, AI-driven optimization techniques for improving the reliability of NOMA-enabled 5G-and-beyondwireless systems.