Joint Workload Migration and Resource Scheduling for Efficient Demand Response in Sustainable Edge Computing
IEEE Transactions on Mobile Computing · 2026-12 · 25 (12) · 23733-23747
DOI: 10.1109/tmc.2026.3731529 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Mobile Computing, 2026-12. Verfasst von Yongchao Zhang und 2 weiteren Autor:innen. Die Arbeit nennt 39 Literaturangaben. Gefördert von: Horizon Europe, UKRI.
Abstract (Originalsprache)
The rapid expansion of edge computing infrastructure has significantly increased its energy consumption and placed substantial pressure on power grids. To develop sustainable edge computing, integrating efficient demand response mechanisms has emerged as a promising solution that encourages edge nodes to dynamically optimize their energy consumption based on grid conditions, enhancing energy efficiency and grid stability while maintaining service quality. Existing demand response methods in edge computing often overlook the diverse service quality requirements of users and rely on centralized algorithms with poor scalability. To address these limitations, we propose a novel distributed workload migration and resource scheduling approach (WMRS) for efficient demand response in sustainable edge computing. WMRS leverages multi-agent model-free reinforcement learning to optimize workload migration and resource scheduling without prior knowledge of system dynamics. To enhance coordination among edge nodes, we introduce a reward sharing mechanism, allowing each node to dynamically share local rewards with its neighbors. Furthermore, we develop a novel constrained optimization approach to effectively manage reward sharing while enabling each node to learn optimal workload migration and resource scheduling policies. Extensive experiments on real-world datasets demonstrate that WMRS achieves rapid convergence and consistently outperforms two state-of-the-art algorithms and a greedy baseline across various setups.
Förderung
- Horizon Europe
- UKRI
Zitieren
APA
Zhang, Y., Hu, J., & Min, G. (2026). Joint Workload Migration and Resource Scheduling for Efficient Demand Response in Sustainable Edge Computing. IEEE Transactions on Mobile Computing, 25(12), 23733-23747. https://doi.org/10.1109/tmc.2026.3731529
Vancouver
Zhang Y, Hu J, Min G. Joint Workload Migration and Resource Scheduling for Efficient Demand Response in Sustainable Edge Computing. IEEE Transactions on Mobile Computing. 2026;25(12):23733-23747. doi:10.1109/tmc.2026.3731529
BibTeX
@article{Zhang2026,
author = {Zhang, Yongchao and Hu, Jia and Min, Geyong},
title = {Joint Workload Migration and Resource Scheduling for Efficient Demand Response in Sustainable Edge Computing},
journal = {IEEE Transactions on Mobile Computing},
year = {2026},
volume = {25},
number = {12},
pages = {23733--23747},
doi = {10.1109/tmc.2026.3731529}
}
EUVIMED