TY - GEN
T1 - Deep Reinforcement Learning-Based Secure Transmission for UAV-Mounted RIS Aided ISAC Systems
AU - Sun, Gangcan
AU - Wang, Kaihao
AU - Zhu, Zhengyu
AU - Chu, Zheng
AU - Li, Zheng
N1 - Publisher Copyright:
© ICST Institute for Computer Sciences, Social Informatics and Telecommunications Engineering 2026.
PY - 2026
Y1 - 2026
N2 - In this paper, a reconfigurable intelligent surface mounted on an unmanned aerial vehicle (UAV-Mounted RIS) assisted integrated sensing and communication (ISAC) secure transmission system is investigated, in which the sensing target (ST) is also regarded as an illegal eavesdropper. Specifically, the airborne RIS is utilized as the relay platform, While introducing solid and dependable line-of-sight (LoS ) links. Our objective is to optimize the aggregate secure communication rate for legitimate users, ensuring compliance with the minimum perceived echo signal-to-noise-ratio requirement, while concurrently addressing the challenges posed by multiple eavesdroppers. To achieve this purpose, the active beamforming and the deployment of UAV in 3D-space together with the phase shift matrix are jointly optimized. The proposed optimization problem is nonconvex on account of the complex coupling between multiple variables in the channel state information (CSI). To address this intractable challenge, we propose a deep reinforcement learning (DRL) framework based on the soft-actor-critic (SAC) algorithm. Simulation results verify the feasibility and efficacy of our proposed scheme.
AB - In this paper, a reconfigurable intelligent surface mounted on an unmanned aerial vehicle (UAV-Mounted RIS) assisted integrated sensing and communication (ISAC) secure transmission system is investigated, in which the sensing target (ST) is also regarded as an illegal eavesdropper. Specifically, the airborne RIS is utilized as the relay platform, While introducing solid and dependable line-of-sight (LoS ) links. Our objective is to optimize the aggregate secure communication rate for legitimate users, ensuring compliance with the minimum perceived echo signal-to-noise-ratio requirement, while concurrently addressing the challenges posed by multiple eavesdroppers. To achieve this purpose, the active beamforming and the deployment of UAV in 3D-space together with the phase shift matrix are jointly optimized. The proposed optimization problem is nonconvex on account of the complex coupling between multiple variables in the channel state information (CSI). To address this intractable challenge, we propose a deep reinforcement learning (DRL) framework based on the soft-actor-critic (SAC) algorithm. Simulation results verify the feasibility and efficacy of our proposed scheme.
KW - Reconfigurable intelligent surface
KW - deep reinforcement learning
KW - deployment optimization
KW - integrated sensing and communication
KW - unmanned aerial vehicle
UR - https://www.scopus.com/pages/publications/105021333654
U2 - 10.1007/978-3-032-03131-0_10
DO - 10.1007/978-3-032-03131-0_10
M3 - Conference contribution
AN - SCOPUS:105021333654
SN - 9783032031303
T3 - Lecture Notes of the Institute for Computer Sciences, Social-Informatics and Telecommunications Engineering, LNICST
SP - 120
EP - 133
BT - Communications and Networking - 19th International Conference, ChinaCom 2024, Proceedings
A2 - Ning, Zhaolong
A2 - Wang, Xiaojie
A2 - Guo, Song
PB - Springer Science and Business Media Deutschland GmbH
T2 - 19th International Conference on Communications and Networking in China, ChinaCom 2024
Y2 - 2 November 2024 through 3 November 2024
ER -