Physical layer security (PLS) has received a growing interest from the research community for its ability to safeguard data confidentiality without relying on key distribution or encryption/decryption. However, the evolution towards the 5G technology and beyond poses new security challenges that must be addressed in order to fulfill the unprecedented performance requirements of future wireless networks. Among the potential enabling technologies, RIS has attracted extensive attention due to its ability to proactively and intelligently reconfigure the wireless propagation environment to combat dynamic wireless channel impairments. Consequently, the RIS technology can be adopted to improve the information-theoretic security of both RF and OWC systems. This survey paper provides a comprehensive overview of the information-theoretic security of RIS-based RF and optical systems. The article first discusses the fundamental concepts of PLS and RIS technologies, followed by their combination in both RF and OWC systems. Subsequently, some optimization techniques are presented in the context of the underlying system model, followed by an assessment of the impact of RIS-assisted PLS through a comprehensive performance analysis. Given that the computational complexity of future communication systems that adopt RIS-assisted PLS is likely to increase rapidly as the number of interactions between the users and infrastructure grows, ML is seen as a promising approach to address this complexity issue while sustaining or improving the network performance. A discussion of recent research studies on RIS-assisted PLS-based systems embedded with ML is presented. Furthermore, some important open research challenges are proposed and discussed to provide insightful future research directions, with the aim of moving a step closer towards the development and implementation of the forthcoming 6G wireless technology.

翻译：物理层安全（PLS）因其无需依赖密钥分发或加密/解密即可保障数据机密性的能力，日益受到研究界的关注。然而，随着5G及未来技术的发展，必须解决新的安全挑战，以满足未来无线网络前所未有的性能要求。在众多潜在使能技术中，RIS因能主动且智能地重新配置无线传播环境以应对动态无线信道损伤而受到广泛关注。因此，RIS技术可被用于提升射频（RF）和光无线通信（OWC）系统的信息论安全性。本综述论文全面概述了基于RIS的RF与光学系统的信息论安全性。文章首先讨论了PLS和RIS技术的基本概念，随后探讨了二者在RF和OWC系统中的结合。接着，在基础系统模型背景下介绍了一些优化技术，并通过全面的性能分析评估了RIS辅助PLS的影响。考虑到采用RIS辅助PLS的未来通信系统因用户与基础设施交互次数增加而可能面临计算复杂度快速攀升的问题，机器学习（ML）被视为在维持或提升网络性能的同时解决该复杂度挑战的一种有前景方案。本文对嵌入ML的基于RIS辅助PLS系统的最新研究进行了讨论。此外，还提出并探讨了若干重要的开放研究挑战，以提供具有前瞻性的未来研究方向，旨在向即将到来的6G无线技术的开发与实施迈进一步。