Joint 3-D Trajectory and Resource Optimization in Licensed and Unlicensed Spectrum for D2D-Underlaid UAV Communication Systems

In recent years, unmanned aerial vehicle (UAV) assisted wireless communication for Internet of Things (IoT) applications has garnered growing attention. Device to Device (D2D)-underlaid UAV communications (D2UAV) system has also become a promising method to provide high-quality services for the terrestrial users with low transmission delay and high communication demands in IoT. Due to the scarcity of licensed spectrum, the unlicensed spectrum is also expanded to the D2UAV system, which necessarily leads to the interference to incumbent users that mainly refers to WiFi systems. Considering the maneuverability of UAVs and the characteristics of the short distance and low transmit power for D2D communications in the D2UAV system, we propose a new unlicensed spectrum sharing (USS) method called direct access (DA) to improve the unlicensed spectrum utilization efficiency and fairness in the article, where the unlicensed spectrum can be directly used without the need for channel detection and without the limit of channel occupancy time but the interference to WiFi systems is strictly controlled, and then investigate the joint optimization problem of 3-D trajectory and resource allocation integrating the DA USS method for the D2UAV system over licensed and unlicensed spectrum, aiming to maximize the total throughput of cellular users (CUs) via jointly optimizing 3-D flight trajectory and resource allocation under various constraints. To solve this problem efficiently, the original problem is decomposed into three optimization subproblems of 3-D trajectory, power control, and bandwidth allocation, and then they are solved alternatively in an iterative manner. Theoretical analysis and results show that the proposed DA USS method with the proposed interference protection method can implement the complete spatial reuse of unlicensed spectrum. A large number of simulation results demonstrate the effectiveness and feasibility of the proposed algorithm integrating the DA USS method.