Fairness-aware resource allocation in PD-NOMA UFMC systems: a hybrid metaheuristic approach

Abstract

Power-domain non-orthogonal multiple access (PD-NOMA) combined with universal filtered multicarrier (UFMC) offers a promising waveform architecture for 5G and beyond, providing high spectral efficiency and improved out-of-band emission suppression. However, the performance of PD-NOMA UFMC systems critically depends on optimal power allocation, which poses a challenging non-convex optimization problem with fairness constraints. Although metaheuristic methods have been explored in PD-NOMA, their application to UFMC-based systems remains limited. This study proposes a phase-controlled hybridization mechanism that integrates the exploration dynamics of moth-flame optimization (MFO) with the exploitation capability of the whale optimization algorithm (WOA) for fairness-driven power allocation in a two-user PD-NOMA UFMC system. A composite weighted objective formulation is designed to jointly minimize bit error rate (BER), maximize achievable rate, and enhance user fairness. The simulation results show that the hybrid WOA–MFO approach converges faster and achieves approximately 12.16% improvement in the best fitness compared to the conventional WOA. In terms of communication performance, the hybrid framework significantly reduces the transmit power required to achieve the target performance levels. For the BER benchmark, the required transmit power is reduced by 6.33 dBm for the Far user and 4.66 dBm for the Near user. For the achievable rate target, the required transmit power is reduced by 2.29 dBm and 4.81 dBm for the Far and Near users, respectively. Similarly, reductions of 2.5 dBm and 4.8 dBm are observed for the Far and Near users at the outage probability threshold. For the fairness requirement based on Jain’s fairness index, the hybrid approach achieves the target level with approximately 2.4 dBm lower transmit power compared with the baseline WOA. These results indicate that the hybrid WOA–MFO approach provides an efficient and balanced solution for computational performance and fairness-driven power allocation in PD-NOMA UFMC systems.