A DRL-Based Energy-Efficient Service Caching and Task Offloading Scheme for 6G MEC SAGINs
IEEE Transactions on Communications · 2025-12 · 73 (12) · 13967-13982
DOI: 10.1109/tcomm.2025.3604318 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Communications, 2025-12. Verfasst von Yi-Huai Hsu und 1 weiteren Autor:innen. Die Arbeit wurde bisher 14-mal zitiert (Crossref) und nennt 36 Literaturangaben. Gefördert von: National Science and Technology Council of Taiwan.
Abstract (Originalsprache)
Space-air-ground integrated networks (SAGINs) integrated with mobile edge computing (MEC) are considered a vital enabler for supporting ubiquitous computing for remote Internet of Things (IoT) devices, especially in areas lacking terrestrial infrastructure, in 6G networks. However, service caching and task offloading strategies significantly affect the amount of data that needs to be transmitted between MEC-enabled network nodes, further influencing the delay experienced by IoT devices and the energy efficiency of an SAGIN. We formulate service caching and task offloading problems as a mixed-integer nonlinear programming problem aiming to minimize the energy consumption of an SAGIN while meeting the delay constraint of each IoT device. We propose an energy-efficient two-event-scale service caching and task offloading scheme (ETST), which contains a large-event-scale-based genetic algorithm, that has a strong global search capability, for the service caching problem in a collaborative multi-uncrewed aerial vehicles environment and a small-event-scale-based deep reinforcement learning-based algorithm to manage stochastic task offloading request events arriving at the resource manager to achieve long-term optimization of network performance for the task offloading problem. The simulation results show that the proposed ETST can successfully lower the energy consumption of the SAGIN while satisfying the delay constraint of each IoT device.
Förderung
- National Science and Technology Council of Taiwan
Zitieren
APA
Hsu, Y., & Phan, T. T. T. (2025). A DRL-Based Energy-Efficient Service Caching and Task Offloading Scheme for 6G MEC SAGINs. IEEE Transactions on Communications, 73(12), 13967-13982. https://doi.org/10.1109/tcomm.2025.3604318
Vancouver
Hsu Y, Phan TTT. A DRL-Based Energy-Efficient Service Caching and Task Offloading Scheme for 6G MEC SAGINs. IEEE Transactions on Communications. 2025;73(12):13967-13982. doi:10.1109/tcomm.2025.3604318
BibTeX
@article{Hsu2025,
author = {Hsu, Yi-Huai and Phan, Thi Thanh Tuyen},
title = {A DRL-Based Energy-Efficient Service Caching and Task Offloading Scheme for 6G MEC SAGINs},
journal = {IEEE Transactions on Communications},
year = {2025},
volume = {73},
number = {12},
pages = {13967--13982},
doi = {10.1109/tcomm.2025.3604318}
}
EUVIMED