Adaptive Energy-Efficient Power Allocation in Downlink NOMA Systems with Imperfect CSI: A PSO-Based Approach for Improving SIC Accuracy and System Performance

Non-Orthogonal Multiple Access (NOMA) has emerged as a critical facilitator of downlink communications in next-generation wireless networks, where user connectivity and spectral efficiency are paramount. The present paper examines dynamic and energy-efficient power allocation in the NOMA systems with imperfect Channel State Information (CSI). To overcome the issues caused by CSI uncertainty and to ensure maximum performance of the system, an Adaptive Particle Swarm Optimization-Dynamic Power Allocation (APSO-DPA) algorithm is proposed. The algorithm has a multi-objective optimization framework that takes into account energy-efficiency (EE), user fairness and the reliability of Successive Interference Cancellation (SIC) decoding in a combined manner. Simulation results reveal that APSO-DPA has always shown better results in comparison with the conventional fixed and equal power allocation scheme at different signal-to-noise ratio (SNR) levels. Its maximum sum rate at 40 dB is 17.7367 bps/Hz, with an energy-efficiency of 0.00177 bits/joule and a Jain’s Fairness Index of 0.5971. Although EE is expected to decline with elevated SNR as more power is consumed, APSO-DPA keeps the balance between throughput and fairness very high. It also provides better SIC robustness and stable convergence to verify that it is resilient to CSI imperfections. In conclusion, the implementation of APSO-DPA in 5G and next-generation wireless communications is highly suitable as it provides a scalable, efficient, and robust power allocation approach.