Joint subcarrier assignment and power allocation for UAV-assisted air-ground integrated full-duplex OFDMA networks

Abstract

The self-interference caused by simultaneous uplink and downlink transmissions, along with inter-cell co-channel interference, significantly challenges the benefits of full-duplex transmission in future multi-UAV assisted Air-Ground Integrated OFDMA Networks. Effective resource allocation is crucial for achieving high system performance in these complex full-duplex environments. This paper investigates the joint optimization of subcarrier scheduling and power assignment, a task complicated by nonconvex Quality of Service (QoS) constraints, the nonconvex nature of the objective function, and the combinatorial intricacies of subcarrier scheduling. To overcome these difficulties, we first propose a Time-Sharing Greedy Rounding algorithm (TS-GR) based on the alternating optimization (AO) method. To further enhance the solution quality, we also propose an $l_p$-norm regularization-based algorithm (LP-NR). Extensive simulation results and theoretical analyses confirm the convergence and efficiency of our proposed methods in UAV-assisted full-duplex OFDMA networks. The simulations highlight that while TS-GR can achieve higher rates under relaxed QoS requirements, LP-NR offers robust performance by consistently satisfying both uplink and downlink QoS requirements. Our findings demonstrate that the gains of multi-cell full-duplex wireless networks over their half-duplex counterparts are significant under optimal conditions but can be constrained by high self-interference and noise levels.