Joint Task Offloading, Resource Allocation, and Trajectory Design for Multi-UAV Cooperative Edge Computing with Task Priority

Hao Hao,Changqiao Xu,Wei Zhang,Shujie Yang,Gabriel-Miro Muntean
DOI: https://doi.org/10.1109/tmc.2024.3350078
IF: 6.075
2024-01-01
IEEE Transactions on Mobile Computing
Abstract:Mobile edge computing (MEC) has emerged as a solution to address the demands of computation-intensive network services by providing computational capabilities at the network edge, thus reducing service delays. Due to the flexible deployment, wide coverage and reliable wireless communication, unmanned aerial vehicles (UAVs) have been employed to assist MEC. This paper investigates the task offloading problem in a UAV-assisted MEC system with collaboration of multiple UAVs, highlighting task priorities and binary offloading mode. We defined the system gain based on energy consumption and task delay. The joint optimization of UAVs' trajectory design, binary offloading decision, computation resources allocation, and communication resources management is formulated as a mixed integer programming problem with the goal of maximizing the long-term average system gain. Considering the discrete-continuous hybrid action space of this problem, we propose a novel deep reinforcement learning (DRL) algorithm based on the latent space to solve it. The evaluation results demonstrate that our proposed algorithm outperforms three state-of-the-art alternative solutions in terms of task delay and system gain.
computer science, information systems,telecommunications
What problem does this paper attempt to address?