曲阜师范大学学报（自然科学版）2024，Vol.50Issue(4)：74-82,9.DOI:10.3969/j.issn.1001-5337.2024.4.074
多无人机辅助边缘计算场景下基于Q-learning的任务卸载优化
Task offloading optimization based on Q-learning in multi-UAV-assisted edge computing system
摘要
Abstract
Multiple unmanned aerial vehicles(UAVs)and the original edge server form a multi-UAV-assisted edge computing system to provide communication and computing resources for mobile users.The optimization problem is modeled as a total system utility maximization problem under resource competition and offloading decision constraints,where the total system utility is composed of three factors:satisfaction of user(SoU),task latency,and system energy consumption.Since the optimization model is a nonconvex problem with NP-hard property,the reinforcement learning method is used to solve the problem and ob-tain the optimal task offloading decision profile that maximize the total system utility.Simulation experi-mental results show that compared to the greedy sequential tuning offloading scheme and the random se-lection offloading scheme,the total system utility of the Q-learning scheme proposed in this paper is im-proved by more than 15％and 43％,respectively.关键词
多无人机辅助边缘计算系统/任务卸载/Q-learning算法Key words
multi-UAV-assisted edge computing system/task offloading/Q-learning algorithm分类
电子信息工程引用本文复制引用
张露,王康,燕晶,张博文,王茂励..多无人机辅助边缘计算场景下基于Q-learning的任务卸载优化[J].曲阜师范大学学报（自然科学版）,2024,50(4):74-82,9.基金项目
山东省自然科学基金(ZR202111260301) （ZR202111260301）
山东省农业重大应用技术创新项目(SD2019NJ007). （SD2019NJ007）
