基于强化学习的STAR-RIS辅助的通信抗干扰方法  

STAR-RIS-assisted anti-Jamming Communication Based on Smooth Q-learning of Similar Action

作　　者：叶子绿 许魁 夏晓晨 邓诚 魏琛 谢威 YE Zilv;XU Kui;XIA Xiaochen;DENG Cheng;WEI Chen;XIE Wei(College of Communication Engineering,Army Engineering University of PLA,Nanjing 210007,China;PLA 32579 Force,Guilin 541000,China)

机构地区：[1]陆军工程大学,江苏南京210007 [2]中国人民解放军32579部队,广西桂林541000

出　　处：《移动通信》2024年第6期69-74,共6页Mobile Communications

基　　金：国家自然科学基金“通信定位一体去蜂窝大规模MIMO智能传输方法研究”(62071485),“面向低空空域的无定形大规模MIMO一体化感知与通信方法研究”(62271503);江苏省基础研究计划“天地融合卫星移动通信组网理论与技术”(BK20192002);江苏省自然科学基金“基于无定形网络的低空空域融合感知与通信方法研究”(BK20231485)。

摘　　要：同时透射与反射可重构智能表面(STAR-RIS)技术作为6G的潜在关键技术之一,能够同时透射和反射入射信号,实现全空间覆盖。提出一种基于强化学习的STAR-RIS辅助通信抗干扰方法,利用相似动作空间平滑Q学习方法对基站发射功率、主动波束成形以及STAR-RIS的被动波束成形进行联合优化,显著增强无线传输系统的抗干扰性能。仿真结果验证提出的SQSA方法相比于Q学习算法具有更快的学习收敛速度、更好的抗干扰性能。As one of the potential key technologies for the sixth generation(6G)mobile networks,simultaneously transmitting and reflecting reconfigurable intelligent surface(STAR-RIS)can simultaneously transmit and reflect incoming signals,achieving full-space coverage.This paper proposes a reinforcement learning-based anti-jamming method for STAR-RIS-assisted communication,utilizing the smooth Q-learning of similar action(SQSA)to jointly optimize the transmit power of base stations,active beamforming,and the passive beamforming of the STAR-RIS,significantly enhancing the anti-jamming performance of wireless transmission systems.Simulation results validate that the proposed SQSA method outperforms the conventional Q-learning algorithms in terms of the learning convergence speed and anti-jamming performance.

关 键 词：通信抗干扰 可重构智能表面 强化学习 波束成形 

分 类 号：TN929.5[电子电信—通信与信息系统]