EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

DSOQR: Deep Reinforcement Learning for Online QoS Routing in SDN-Based Networks

Lianming Zhang, Yong Lu, Dian Zhang, Haoran Cheng, Pingping Dong

Security and Communication Networks · 29.11.2022 · 2022 · 1-11

DOI: 10.1155/2022/4457645 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Security and Communication Networks, erschienen am 29.11.2022. Verfasst von Lianming Zhang und 4 weiteren Autor:innen. Die Arbeit wurde bisher 19-mal zitiert (Crossref) und nennt 22 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Natural Science Foundation of China, Natural Science Foundation of Hunan Province.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

With the rapid development of mobile communication technology, there are an increasing number of new network applications and services, and the existing best-effort routing algorithms cannot meet the quality-of-service (QoS) requirements of these applications and services. QoS-routing optimization solutions based on a software-defined network (SDN) are often targeted at specific network scenarios and difficult to adapt to changing business requirements. The current routing algorithms based on machine learning (ML) methods can generally only handle discrete, low-dimensional action spaces and use offline network data for training, which is not effective for dynamic network environments. In this study, we propose DSOQR, which is an online QoS-routing framework based on deep reinforcement learning (DRL) and SDN. DSOQR collects network status information in real time through the software-defined paradigm and carries out on-policy learning. Under this framework, we further propose SA3CR, which is a QoS-routing algorithm based on SDN and asynchronous advantage actor-critic (A3C). The SA3CR algorithm can dynamically switch routing paths that meet the conditions according to the current network status and the needs of different service types to ensure the QoS of target traffic. Experimental results show that DSOQR is effective and that the SA3CR algorithm has better performance in terms of delay, throughput, and packet loss rate.

Förderung

  • National Natural Science Foundation of China
  • Natural Science Foundation of Hunan Province

Zitieren

APA

Zhang, L., Lu, Y., Zhang, D., Cheng, H., & Dong, P. (2022). DSOQR: Deep Reinforcement Learning for Online QoS Routing in SDN-Based Networks. Security and Communication Networks, 2022, 1-11. https://doi.org/10.1155/2022/4457645

Vancouver

Zhang L, Lu Y, Zhang D, Cheng H, Dong P. DSOQR: Deep Reinforcement Learning for Online QoS Routing in SDN-Based Networks. Security and Communication Networks. 2022;2022:1-11. doi:10.1155/2022/4457645

BibTeX
@article{Zhang2022,
  author = {Zhang, Lianming and Lu, Yong and Zhang, Dian and Cheng, Haoran and Dong, Pingping},
  title = {DSOQR: Deep Reinforcement Learning for Online QoS Routing in SDN-Based Networks},
  journal = {Security and Communication Networks},
  year = {2022},
  volume = {2022},
  pages = {1--11},
  doi = {10.1155/2022/4457645}
}
Kostenlos starten →