Optimized task offloading for federated learning based on β-skeleton graph in edge computing

Mahdi Fallah () and Pedram Salehpour ()
Additional contact information
Mahdi Fallah: University of Tabriz
Pedram Salehpour: University of Tabriz

Telecommunication Systems: Modelling, Analysis, Design and Management, 2024, vol. 87, issue 3, No 16, 759-778

Abstract: Abstract Edge computing is gaining prominence as a solution for IoT data management and processing. Task offloading, which distributes the processing load across edge devices, is a key strategy to enhance the efficiency of edge computing. However, traditional methods often overlook the dynamic nature of the edge environment and the interactions between devices. While reinforcement learning-based task offloading shows promise, it can sometimes lead to an imbalance by favoring weaker servers. To address these issues, this paper presents a novel task offloading method for federated learning that leverages the β-skeleton graph in edge computing. This model takes into account spatial and temporal dynamics, optimizing task assignments based on both the processing and communication capabilities of the edge devices. The proposed method significantly outperforms five state-of-the-art methods, showcasing substantial improvements in both initial and long-term performance. Specifically, this method demonstrates a 63.46% improvement over the Binary-SPF-EC method in the initial rounds and achieves an average improvement of 76.518% after 400 rounds. Moreover, it excels in sub-rewards and total latency reduction, underscoring its effectiveness in optimizing edge computing communication and processing tasks. These results underscore the superiority of the proposed method, highlighting its potential to enhance the efficiency and scalability of edge computing systems. This approach, by effectively addressing the dynamic nature of the edge environment and optimizing task offloading, contributes to the development of more robust and efficient edge computing frameworks. This work paves the way for future advancements in federated learning and edge computing integration, promising better management and utilization of IoT data.

Keywords: Deep reinforcement learning; Federated learning; Task offloading; Policy optimization (search for similar items in EconPapers)
Date: 2024
References: View complete reference list from CitEc
Citations:

Downloads: (external link)
http://link.springer.com/10.1007/s11235-024-01216-4 Abstract (text/html)
Access to the full text of the articles in this series is restricted.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:spr:telsys:v:87:y:2024:i:3:d:10.1007_s11235-024-01216-4

Ordering information: This journal article can be ordered from
http://www.springer.com/journal/11235

DOI: 10.1007/s11235-024-01216-4

Access Statistics for this article

Telecommunication Systems: Modelling, Analysis, Design and Management is currently edited by Muhammad Khan

More articles in Telecommunication Systems: Modelling, Analysis, Design and Management from Springer
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().