EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Cross-Domain Collaborative Federated Intelligence for Wireless Computing Power Networks

Yuan Feng, Jiayao Yuan, Jiajia Liu, Yunlong Lu, Hao Wu

Journal of Intelligent Computing and Networking · 25.03.2026 · 2 (1) · 22-34

DOI: 10.64509/jicn.21.75 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Intelligent Computing and Networking, online erschienen am 25.03.2026. Verfasst von Yuan Feng und 4 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref). Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

With the rapid development of wireless and intelligent computing technologies, emerging applications like autonomous driving and AI agents demand ubiquitous, low-latency, and efficient computing resources. Wireless computing power networks (WCPNs), which interconnect geographically distributed and heterogeneous computing nodes through wireless communication infrastructures, are regarded as a key paradigm for supporting such computing-intensive and latency-sensitive services. However, the current critical challenge is how to achieve efficient, low-latency joint optimisation across domains and multiple tasks by coordinating highly heterogeneous distributed computing and communication resources while protecting each node's data privacy. To address the challenge of privacy-preserving collaboration among heterogeneous computing nodes with non-IID data distributions in WCPNs, this paper introduces a federated intelligence–driven learning framework. By leveraging federated learning with a FedOpt-based aggregation mechanism, collaborative model training can be achieved without sharing raw data, while improving accuracy and convergence performance under data and node heterogeneity. To further address the cross-domain multi-task resource allocation problem under stringent real-time requirements, a joint optimisation model is formulated that incorporates wireless communication conditions, heterogeneous computing capabilities, task service demands, and system energy consumption. By minimising the total execution delay of multiple tasks subject to resource, energy, and scheduling constraints, a genetic algorithm–based solution is employed to derive near-optimal task orchestration and resource allocation strategies. Simulation results demonstrate that the proposed framework achieves lower task execution delay and higher energy efficiency than baseline schemes, validating its effectiveness in WCPNs.

Zitieren

APA

Feng, Y., Yuan, J., Liu, J., Lu, Y., & Wu, H. (2026). Cross-Domain Collaborative Federated Intelligence for Wireless Computing Power Networks. Journal of Intelligent Computing and Networking, 2(1), 22-34. https://doi.org/10.64509/jicn.21.75

Vancouver

Feng Y, Yuan J, Liu J, Lu Y, Wu H. Cross-Domain Collaborative Federated Intelligence for Wireless Computing Power Networks. Journal of Intelligent Computing and Networking. 2026;2(1):22-34. doi:10.64509/jicn.21.75

BibTeX
@article{Feng2026,
  author = {Feng, Yuan and Yuan, Jiayao and Liu, Jiajia and Lu, Yunlong and Wu, Hao},
  title = {Cross-Domain Collaborative Federated Intelligence for Wireless Computing Power Networks},
  journal = {Journal of Intelligent Computing and Networking},
  year = {2026},
  volume = {2},
  number = {1},
  pages = {22--34},
  doi = {10.64509/jicn.21.75}
}
Kostenlos starten →