Quantum Reinforcement Learning for Cognitive Radio-Assisted UAV Traffic Offloading

With the continuous development of communication industry, the demand for spectrum resources is gradually increasing. Simultaneously, unmanned aerial vehicles (UAVs) have been extensively studied as a means to assist with traffic offloading by establishing wireless links between base stations (BSs) and UAVs. However, the current spectrum resources are limited and far from optimal utilization. In response to this situation, cognitive radio (CR) technology and non-orthogonal multiple access (NOMA) technologies have been proposed to enhance the efficient utilization of spectrum resources. In this paper, we address the challenge of using CR technology to assist UAVs in traffic offloading within specific areas. At each instance, the UAV must consider the allocation of time for data collection, traffic transmission, and UAV charging. Given the dynamic nature of spectrum resources, UAVs must make adaptive decisions based on the current conditions. Finally, we model the aforementioned issues as a quantum reinforcement learning (QRL) system to expedite the training process. In particular, we design a quantum soft actor-critic algorithm employing a variable quantum circuit (VQC) and encode the experience data into a quantum summation tree while transforming it into a quantum priority experience replay buffer. Simulation results demonstrate the superior performance of the proposed framework.