liu.seSearch for publications in DiVA
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • oxford
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
DRACO: Decentralized Asynchronous Federated Learning Over Row-Stochastic Wireless Networks
Linköping University, Department of Computer and Information Science, Database and information techniques. Linköping University, Faculty of Science & Engineering.ORCID iD: 0000-0001-8838-0463
Univ Granada, Spain; EURECOM, France.
2025 (English)In: IEEE Open Journal of the Communications Society, E-ISSN 2644-125X, Vol. 6, p. 4818-4839Article in journal (Refereed) Published
Abstract [en]

Emerging technologies and use cases, such as smart Internet of Things (IoT), Internet of Agents, and Edge AI, have generated significant interest in training neural networks over fully decentralized, serverless networks. A major obstacle in this context is ensuring stable convergence without imposing stringent assumptions, such as identical data distributions across devices or synchronized updates. In this paper, we introduce DRACO, a novel framework for decentralized asynchronous Stochastic Gradient Descent (SGD) over row-stochastic gossip wireless networks. Our approach leverages continuous communication, allowing edge devices to perform local training and exchange model updates along a continuous timeline, thereby eliminating the need for synchronized timing. Additionally, our algorithm decouples communication and computation schedules, enabling complete autonomy for all users while effectively addressing straggler issues. Through a thorough convergence analysis, we show that DRACO achieves high performance in decentralized optimization while maintaining low variance across users even without predefined scheduling policies. Numerical experiments further validate the effectiveness of our approach, demonstrating that controlling the maximum number of received messages per client significantly reduces redundant communication costs while maintaining robust learning performance.

Place, publisher, year, edition, pages
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC , 2025. Vol. 6, p. 4818-4839
Keywords [en]
Training; Optimization; Stochastic processes; Directed graphs; Delays; Wireless networks; Schedules; Computational modeling; Synchronization; Internet of Things; Collaborative intelligence; continuous time models; decentralized learning; asynchronous learning; federated learning; row-stochastic matrices; networked intelligent systems; peer-to-peer networks
National Category
Communication Systems
Identifiers
URN: urn:nbn:se:liu:diva-215400DOI: 10.1109/OJCOMS.2025.3574098ISI: 001508097800005Scopus ID: 2-s2.0-105006825700OAI: oai:DiVA.org:liu-215400DiVA, id: diva2:1976526
Note

Funding Agencies|Horizon Europe JU SNS Project [101139068]

Available from: 2025-06-25 Created: 2025-06-25 Last updated: 2025-09-23

Open Access in DiVA

fulltext(2436 kB)56 downloads
File information
File name FULLTEXT01.pdfFile size 2436 kBChecksum SHA-512
3b20421b44a28beaa129b3a15ffadb37f13bb6b600cd5c7152c80c0f834e9aef1c7f312a72cb0ef68c83ca6ae2306f64b888c41a0e89193f8df65e07548e5543
Type fulltextMimetype application/pdf

Other links

Publisher's full textScopus

Search in DiVA

By author/editor
Jeong, Eunjeong
By organisation
Database and information techniquesFaculty of Science & Engineering
In the same journal
IEEE Open Journal of the Communications Society
Communication Systems

Search outside of DiVA

GoogleGoogle Scholar
Total: 56 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

doi
urn-nbn

Altmetric score

doi
urn-nbn
Total: 84 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • oxford
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf