Fast Opportunistic Spectrum Sensing and Throughput Maximization for Cognitive Unmanned Aerial Vehicle Communications

Unmanned aerial vehicle (UAV) assisted industrial Internet of Things (IIoT) mainly relies on unlicensed frequency bands for communication, and its development is facing the challenge of a shortage of spectrum resources. By integrating cognitive radio (CR) technology into the UAV-assisted IIoT, unlicensed UAV users can dynamically and opportunistically access the spectrum of the primary users to transmit data. Due to the limited access time, an inevitable trade-off exists between the sensing time and achievable throughput. In UAV scenarios, spectrum sensing mainly focuses on energy detection, which has poor real-time performance and cannot meet the low latency requirements of IIoT. Therefore, we propose a distributed fast spectrum sensing method to effectively balance sensing time and throughput. In the considering cognitive UAV network (CUAVN), monitoring UAVs (MUAVs) serve as sensing nodes, responsible for collecting industrial data and opportunistically utilizing the idle spectrum resources of the terrestrial network to transmit the data to the receiving UAV (RUAV). We define a distributed sensing sphere, where the central MUAV performs grouped sequential detection, leveraging the spatial similarity of sensing results among MUAVs within the same sensing sphere to shorten the sensing time. Based on the proposed distributed fast spectrum sensing method, the formula for sensing time is derived and used to calculate the throughput. Our goal is to maximize the throughput of CUAVN by jointly optimizing the false alarm probability, MUAVs' transmit power, and RUAV's location. Simulation results show that compared with the classical sequential detection-based and energy detection-based spectrum sensing methods, the average sensing time of the proposed method is reduced by 78.53% and 88.96%, and the throughput is increased by 19.4% and 34.3%, respectively.