EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Action-factorized Rainbow deep Q-network with token Transformer for computation offloading in edge computing-enabled Internet of Ships

Shengtian Zhang iD, Haolin Yang, Hyeonseok Kim, Incheol Shin iD, Robert M X Wu

PLOS One · 11.05.2026 · 21 (5) · e0348376

DOI: 10.1371/journal.pone.0348376 ↗

Kurz gefasst

Fachartikel in der Zeitschrift PLOS One, online erschienen am 11.05.2026. Verfasst von Shengtian Zhang und 4 weiteren Autor:innen. Beteiligte Einrichtung: Department of Artificial Intelligence Convergence, Pukyong National University, Busan, South Korea. Hauptthemen laut MEDLINE: Ships, Internet. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 46 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: the Ministry of Trade, Industry and Energy, Korea.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Edge computing (EC) in the Internet of Ships (IoS) reduces the latency and energy burdens of cloud-centric architectures, but fully realizing its benefits requires effective computation offloading strategies. Designing such strategies in dynamic maritime environments remains challenging due to the high-dimensional, combinatorial decision space, strict system constraints, and rapidly varying maritime wireless channels. This study proposes the action-factorized Rainbow deep Q-network (DQN) with token Transformer, a deep reinforcement learning (DRL) algorithm for discovering effective computation offloading strategies in EC-enabled IoS (EC-IoS). The core innovation of the algorithm lies in a novel action factorization mechanism coupled with our custom token Transformer-based state and action encoders, which effectively handle the complex decision space. Built upon Rainbow DQN and further accelerated with a parallel training architecture, the algorithm improves learning efficiency and stability. Experimental results illustrate that the computation offloading strategies learned by our algorithm significantly outperform multiple baselines on the weighted latency–energy objective. More importantly, these strategies achieve a zero rate of invalid actions, satisfy all system constraints, and ensure practical feasibility. Overall, the study demonstrates that the algorithm provides a robust method for computation offloading, effectively balancing latency and energy consumption in EC-IoS, thereby supporting maritime digitalization and automation.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

ShipsInternetAlgorithmsReinforcement Machine LearningWireless Technology

Autor:innen und Institutionen

  • Department of Artificial Intelligence Convergence, Pukyong National University, Busan, South Korea.
  • School of Computer and Network Engineering, Shanxi Datong University, Datong, China.
  • Faculty of Engineering and Information Technology, University of Technology Sydney, Sydney, Australia.

ORCID-Profile: Shengtian Zhang, Incheol Shin

Förderung

  • the Ministry of Trade, Industry and Energy, Korea

Zitieren

APA

Zhang, S., Yang, H., Kim, H., Shin, I., & Wu, R. M. X. (2026). Action-factorized Rainbow deep Q-network with token Transformer for computation offloading in edge computing-enabled Internet of Ships. PLOS One, 21(5), e0348376. https://doi.org/10.1371/journal.pone.0348376

Vancouver

Zhang S, Yang H, Kim H, Shin I, Wu RMX. Action-factorized Rainbow deep Q-network with token Transformer for computation offloading in edge computing-enabled Internet of Ships. PLOS One. 2026;21(5):e0348376. doi:10.1371/journal.pone.0348376

BibTeX
@article{Zhang2026,
  author = {Zhang, Shengtian and Yang, Haolin and Kim, Hyeonseok and Shin, Incheol and Wu, Robert M X},
  title = {Action-factorized Rainbow deep Q-network with token Transformer for computation offloading in edge computing-enabled Internet of Ships},
  journal = {PLOS One},
  year = {2026},
  volume = {21},
  number = {5},
  pages = {e0348376},
  doi = {10.1371/journal.pone.0348376}
}
Kostenlos starten →