Due to the mobility and flexibility, using unmanned aerial vehicles (UAVs) for data collection from wireless sensors in the Internet of Things will be a promising solution. In this paper, we propose a fairness-aware trajectory design for UAV-aided data collection network. Particularly, we establish a Markov decision process (MDP) to represent the interaction between UAV and sensors, and a fairness-aware reward function is defined to evaluate the performance of UAV. Through the reinforcement learning algorithm deep deterministic policy gradient (DDPG), the optimal flight trajectory of UAV is obtained. Simulation results demonstrate that the proposed fairness-aware trajectory design method can significantly improve data collection ratio (DCR) and system fairness index compared with the common baseline method.
A Fairness-Aware UAV Trajectory Design with Reinforcement Learning
Lect. Notes Electrical Eng.
International Conference On Signal And Information Processing, Networking And Computers ; 2021 December 27, 2021 - December 29, 2021
Signal and Information Processing, Networking and Computers ; Kapitel : 132 ; 1026-1031
13.10.2022
6 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch