EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Adaptive Deep Reinforcement Learning: A Novel Framework for DDoS Detection on Resource-Constrained Edge Devices

Syaifuddin Saif, Widyawan Widyawan, Ridi Ferdiana

Engineering, Technology & Applied Science Research · 04.04.2026 · 16 (2) · 32962-32970

DOI: 10.48084/etasr.16594 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Engineering, Technology & Applied Science Research, online erschienen am 04.04.2026. Verfasst von Syaifuddin Saif und 2 weiteren Autor:innen. Die Arbeit nennt 20 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The rapid growth of the Internet of Things (IoT) has significantly increased exposure to Distributed Denial of Service (DDoS) attacks, particularly due to the limited resources and heterogeneous traffic characteristics of IoT devices. Conventional intrusion detection approaches, including supervised learning models, often fail to adapt to dynamic traffic patterns and zero-day attacks. This study proposes an adaptive DDoS detection framework based on Deep Reinforcement Learning (DRL) designed for deployment on resource-constrained edge devices. The framework is evaluated using the IoT-DH dataset, a real-world multi-protocol IoT traffic dataset collected via a honeypot. Experimental results show that the proposed DRL-based approach consistently outperforms static models, achieving an F1-score of up to 0.98–0.99 on in-distribution data while maintaining a low False Positive Rate (FPR). Cross-dataset evaluation on public benchmarks further demonstrates stable performance with F1-scores above 0.96 under distribution shifts. Implementation on a Raspberry Pi 3 confirms that the model operates with low inference latency and acceptable resource usage. These results indicate that adaptive DRL provides an effective and practical solution for real-time DDoS detection in edge-based IoT environments.

Zitieren

APA

Saif, S., Widyawan, W., & Ferdiana, R. (2026). Adaptive Deep Reinforcement Learning: A Novel Framework for DDoS Detection on Resource-Constrained Edge Devices. Engineering, Technology & Applied Science Research, 16(2), 32962-32970. https://doi.org/10.48084/etasr.16594

Vancouver

Saif S, Widyawan W, Ferdiana R. Adaptive Deep Reinforcement Learning: A Novel Framework for DDoS Detection on Resource-Constrained Edge Devices. Engineering, Technology & Applied Science Research. 2026;16(2):32962-32970. doi:10.48084/etasr.16594

BibTeX
@article{Saif2026,
  author = {Saif, Syaifuddin and Widyawan, Widyawan and Ferdiana, Ridi},
  title = {Adaptive Deep Reinforcement Learning: A Novel Framework for DDoS Detection on Resource-Constrained Edge Devices},
  journal = {Engineering, Technology & Applied Science Research},
  year = {2026},
  volume = {16},
  number = {2},
  pages = {32962--32970},
  doi = {10.48084/etasr.16594}
}
Kostenlos starten →