Joint UAV 3D Trajectory Design and Resource Scheduling for Space-Air-Ground Integrated Power IoRT: A Deep Reinforcement Learning Approach

• Published in: IEEE transactions on network science and engineering Vol. 11; no. 3; pp. 2632 - 2646
• Main Authors: Liu, Jiayan, Zhao, Xiongwen, Qin, Peng, Du, Fei, Chen, Zhiyu, Zhou, Hongxi, Li, Junqin
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 01.05.2024; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; association control; Autonomous aerial vehicles; Communication; Deep learning; Design optimization; Energy consumption; multi-agent deep deterministic policy gradient (MADDPG); Multiagent systems; power optimization; Remote sensors; Resource scheduling; Satellites; Sensors; Space-air-ground integrated power Internet of Remote Things (SAG-PIoRT); Three-dimensional displays; Throughput; Tradeoffs; Trajectory; UAV three-dimensional (3D) trajectory design; Unmanned aerial vehicles
• ISSN: 2327-4697; 2334-329X
• DOI: 10.1109/TNSE.2023.3346445

The terrain-independent space-air-ground integrated power Internet of Remote Things (SAG-PIoRT) is able to bring efficient communication services with seamless coverage for sensors in remote areas without information and communication network coverage. In this paper, considering the fairness among sensors and the tradeoff between throughput and energy consumption in the SAG-PIoRT, we follow the uplink transmission communication issue through jointly optimizing UAV three-dimensional (3D) trajectory design, UAV-sensors association, and sensors transmission power, aiming to maximize the long-term network capacity and simultaneously minimize the sensors total energy consumption. We employ a deep reinforcement learning framework of multi-agent deep deterministic policy gradient (MADDPG) and propose 3D Trajectory design and Resource scheduling based on MADDPG in the SAG-PIoRT (TR-MADDPG) algorithm to obtain the optimal solution. Simulations reveal that TR-MADDPG can achieve the tradeoff between throughput and energy consumption. Compared to other benchmark algorithms, TR-MADDPG performs better in fairness index, total energy consumption, and fair throughput-energy consumption.