An AI-Enabled Federated Edge–Cloud Communication Framework for Dynamic Task Offloading and Network Anomaly Detection in IoT System
IEEE Open Journal of the Communications Society · 2026 · 7 · 4376-4388
DOI: 10.1109/ojcoms.2026.3686051 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Open Journal of the Communications Society, 2026. Verfasst von Sirisha Rani Kolli und 1 weiteren Autor:innen. Die Arbeit nennt 34 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
The rapid explosion of Internet of Things (IoT) devices is primarily reshaping modern communication networks. However, they introduce unique challenges in resource management, latency control, and secure data exchange across large-scale edge–cloud infrastructures. Conventional task offloading mechanisms fail to adapt effectively to highly dynamic network conditions and suffer significant communication overhead. This paper presents a task-aware federated learning framework for efficient computation offloading and anomaly detection in Internet of Things based mobile edge computing (MEC) environments. It integrates a hybrid deep learning architecture with adaptive federated learning to jointly optimize task execution decisions and network security. In contrast to the conventional approaches which treat offloading and intrusion detection as isolated entities, the proposed framework provides coordinated decision-making across edge and cloud resources and also reduces communication overhead. The model is evaluated to validate its performance and results demonstrate that the proposed method achieves up to 38–42% reduction in latency, 22–27% improvement in throughput, and nearly 30% reduction in federated learning communication overhead compared to state-of-the-art approaches. Moreover, the anomaly detection module achieves an accuracy of 99.61%, with very low false positive and false negative rates. This indicates that the model is superior in identifying malicious activity in IoT networks. These results highlight the efficiency of combining task-aware federated learning with intelligent offloading strategies for scalable, efficient, and secure edge computing systems.
Zitieren
APA
Kolli, S. R., & Rahman, M. Z. U. (2026). An AI-Enabled Federated Edge–Cloud Communication Framework for Dynamic Task Offloading and Network Anomaly Detection in IoT System. IEEE Open Journal of the Communications Society, 7, 4376-4388. https://doi.org/10.1109/ojcoms.2026.3686051
Vancouver
Kolli SR, Rahman MZU. An AI-Enabled Federated Edge–Cloud Communication Framework for Dynamic Task Offloading and Network Anomaly Detection in IoT System. IEEE Open Journal of the Communications Society. 2026;7:4376-4388. doi:10.1109/ojcoms.2026.3686051
BibTeX
@article{Kolli2026,
author = {Kolli, Sirisha Rani and Rahman, Md. Zia Ur},
title = {An AI-Enabled Federated Edge–Cloud Communication Framework for Dynamic Task Offloading and Network Anomaly Detection in IoT System},
journal = {IEEE Open Journal of the Communications Society},
year = {2026},
volume = {7},
pages = {4376--4388},
doi = {10.1109/ojcoms.2026.3686051}
}
EUVIMED