Integrating Optimization and Reinforcement Learning for Optimal Design in 6G Wireless Networks

  • Nguyen, Thi My Tuyen; 
  • Nguyen, The Vi; 
  • Oh, Junsuk; 
  • Cho, Sungrae
Citations

SCOPUS

0

초록

The proliferation of demanding wireless applications and rapidly growing network density, poses significant challenges in designing optimal control policies under dynamic and uncertain environments. While conventional optimization methods offer mathematical rigor, they often suffer from high computational complexity and poor adaptability to channel dynamics. Deep reinforcement learning (DRL) provides a model-free alternative with enhanced robustness but struggles with slow convergence and substantial training overhead. To overcome these limitations, recent research has focused on hybrid frameworks that integrate model-based optimization with model-free DRL, leveraging the complementary strengths of both approaches. This paper surveys key developments in such optimization-driven DRL methods, highlighting their ability to improve learning efficiency, stability, and scalability in complex wireless networks.

키워드

Deep reinforcement learning (DRL); optimization-driven DRL; Wireless networks
제목
Integrating Optimization and Reinforcement Learning for Optimal Design in 6G Wireless Networks
저자
Nguyen, Thi My Tuyen; Nguyen, The Vi; Oh, Junsuk; Cho, Sungrae
DOI
10.1109/ICTC66702.2025.11389079
발행일
2025
유형
Conference Paper
저널명
International Conference on ICT Convergence
페이지
684 ~ 686