Welcome to Diffusion Reinforcement Learning

Diffusion Reinforcement Learning (DRL) is an advanced approach that integrates the principles of diffusion processes with reinforcement learning algorithms. This hybrid method leverages the strength of diffusion models, which are adept at generating high-quality, continuous data representations, with the strategic decision-making capabilities of reinforcement learning. DRL is particularly effective in environments where the agent’s actions can be seen as influencing or navigating through a continuous state space, allowing for more nuanced and precise control strategies. By combining these two methodologies, DRL enhances the agent’s ability to learn optimal policies in complex, dynamic environments, making it a powerful tool for tasks that require sophisticated, long-term planning and execution.

Representative Publications

Enhancing Physical Layer Communication Security through Generative AI with Mixture of Experts
Changyuan Zhao, Hongyang Du, Dusit Niyato, Jiawen Kang, Zehui Xiong, Dong In Kim, Xuemin (Sherman) Shen, Khaled B. Letaief
Under Review

The paper explores mixture of experts-enabled generative AI, focusing on physical layer communication security.
Paper

Defining Problem from Solutions: Inverse Reinforcement Learning (IRL) and Its Applications for Next-Generation Networking
Yinqiu Liu, Ruichen Zhang, Hongyang Du, Dusit Niyato, Jiawen Kang, Zehui Xiong, and Dong In Kim
Under Review

This paper introduces Inverse Reinforcement Learning and its applications in networking.
Paper | Code | YouTube | Bilibili

Beyond Deep Reinforcement Learning: A Tutorial on Generative Diffusion Models in Network Optimization
Hongyang Du, Ruichen Zhang, Yinqiu Liu, Jiacheng Wang, Yijing Lin, Zonghang Li, Dusit Niyato, Jiawen Kang, Zehui Xiong, Shuguang Cui, Bo Ai, Haibo Zhou, Dong In Kim
IEEE COMST

The first tutorial paper to explain various applications of diffusion models and coding examples.
Paper | Code | YouTube | Bilibili

User-Centric Interactive AI for Distributed Diffusion Model-based AI-Generated Content
Hongyang Du, Ruichen Zhang, Dusit Niyato, Jiawen Kang, Zehui Xiong, Shuguang Cui, Xuemin Shen, Dong In Kim
Under Review

This paper introduces Reinforcement Learning With Large Language Models Interaction (RLLI), a user-centric Interactive AI technique, to optimize AIGC services.
Paper | Code | YouTube | Bilibili

Diffusion-based Reinforcement Learning for Edge-enabled AI-Generated Content Services
Hongyang Du, Zonghang Li, Dusit Niyato, Jiawen Kang, Zehui Xiong, Huawei Huang, and Shiwen Mao
IEEE Transactions on Mobile Computing

This paper proposes a diffusion model-based AI-generated optimal decision (AGOD) algorithm, which can generate the optimal ASP selection decisions.
Paper | Code

Federated Learning-Empowered AI-Generated Content in Wireless Networks
Xumin Huang, Peichun Li, Hongyang Du, Jiawen Kang, Dusit Niyato, Dong In Kim, Yuan Wu
IEEE Network Magazine

This paper presents the federated learning design to support AIGC with privacy preservation.
Paper | Code | YouTube | Bilibili

Deep Generative Model and Its Applications in Efficient Wireless Network Management: A Tutorial and Case Study
Yinqiu Liu, Hongyang Du, Dusit Niyato, Jiawen Kang, Zehui Xiong, Dong In Kim, Abbas Jamalipour
IEEE Network Magazine

The paper gives specific examples of generative AI to optimize wireless networks.
Paper | Code | YouTube | Bilibili

AI-Generated 6G Internet Design: A Diffusion Model-based Learning Approach
Yudong Huang, Minrui Xu, Xinyuan Zhang, Dusit Niyato, Zehui Xiong, Shuo Wang, Tao Huang
IEEE Network Magazine

The paper introduces a concept of AI-Generated Network.
Paper | Code | YouTube | Bilibili

back