Knowledge-Driven Deep Reinforcement Learning for Wireless Networks

  • Nguyen, Thi My Tuyen; 
  • Nguyen, The Vi; 
  • Do, Tung Son; 
  • Won, Dongwook; 
  • Cho, Sungrae
Citations

WEB OF SCIENCE

0
Citations

SCOPUS

0

초록

Deep reinforcement learning (DRL) is a promising technique for solving numerous optimization problems in wireless networks, owing to its high adaptability to dynamic and uncertain wireless environments. However, the lack of interpretability, inherent in the black-box nature of the deep neural networks (DNNs) within DRL, limits its deployment in real-world applications. In this paper, we focus on integrating domain knowledge into DRL to enhance robustness and learning efficiency. Specifically, we review a teacher-student framework, where a teacher agent utilizes knowledge from classical model-based methods to guide a student DRL agent toward improved decision-making.

키워드

Deep reinforcement learning (DRL); knowledge-driven DRL; Wireless networks
제목
Knowledge-Driven Deep Reinforcement Learning for Wireless Networks
저자
Nguyen, Thi My Tuyen; Nguyen, The Vi; Do, Tung Son; Won, Dongwook; Cho, Sungrae
DOI
10.1109/ICOIN68469.2026.11480536
발행일
2026
유형
Proceedings Paper
저널명
International Conference on Information Networking
페이지
426 ~ 428