Trustworthy And Efficient Deep Reinforcement Learning-Driven Physical-Layer For Secure 6G Networks

Alex Pierron, Joaquin Garcia-Alfaro, Jose Rubio Hernan, Michel Barbeau, Luca De Cicco

Keywords: Physical~AI, Reinforcement Learning, 6G, Cybersecurity, Physical-Layer Security, Reconfigurable Intelligent Surfaces

Abstract: This short paper presents our research on Deep Reinforcement Learning (DRL) control of reconfigurable intelligent surfaces (RIS) for the physical security of future 6G networks. We first mention our results on integrating fairness into a multi-user DRL-RIS controller, then situate our ongoing work on DRL backdoors based on foundational literature that has yet to be widely applied to physical systems. Finally, we highlight the relevance of our "CTRL\_RIS" simulation environment as a reproducible open-source resource for linking current results and future avenues of research on fairness, security, and robustness in DRL-RIS controllers for 6G networks.

Email Sharing: We authorize the sharing of all author emails with Program Chairs.

Data Release: We authorize the release of our submission and author names to the public in the event of acceptance.

Submission Number: 8