Hybrid Offline-Online UAV Trajectory Design and Subchannel Allocation in UAV Relaying OFDMA Networks

In this article, we study an unmanned aerial vehicle (UAV) relaying orthogonal frequency-division multiple access (OFDMA) network with multiple user equipment (UE) pairs, each having explicitly given Quality-of-Service (QoS) requirements. Our goal is to maximize the system throughput while satisfying the QoS requirements for UE pairs, despite the air-to-ground (A2G) channel randomness. To this end, we develop a hybrid offline-online algorithm that designs the UAV's 3-D trajectory in an offline manner and then allocates subchannels during the UAV flight in an online manner. Under the proposed algorithm, the UAV's 3-D trajectory is elaborately designed with statistical channel state information (CSI), and subchannels are opportunistically allocated based on instantaneous CSI. In practice, the QoS requirements might not be guaranteed since perfect CSI cannot be obtained a priori. Nevertheless, the proposed algorithm significantly improves the QoS satisfaction of UE pairs to the fullest extent possible by leveraging available CSI. Through simulation, we validate the performance of the proposed algorithm in enhancing the QoS satisfaction of UE pairs and the system throughput.