AI赋能6G网络安全：架构与关键技术

摘要：针对6G网络架构中的需求与挑战，提出了“主动免疫-孪生互驱-弹性自治-分布协同”的智慧内生安全架构。该架构通过预训练威胁表征模型实现攻击前预判，利用数字孪生构建虚实结合防御体系，借助联邦学习与区块链技术建立跨域协同机制，形成“感知-决策-验证-优化”的安全闭环。介绍了5项6G网络安全中的关键技术：分布式机器学习、AI大模型、轻量级认证授权与访问控制、数字孪生、无线物理层安全技术，为6G网络的高效可信运行提供理论支撑与技术路径分析。

关键词：6G网络安全；人工智能；分布式机器学习；大模型；轻量级认证授权与访问控制；数字孪生；无线物理层安全

Abstract: In response to the requirements and challenges of the 6G network architecture, an intelligent endogenous security architecture of “proactive immunity, twin-driven, resilient autonomy, and distributed collaboration” is proposed. This architecture enables proactive threat prediction through pre-trained threat representation models, constructs a cyber-physical integrated defense system using digital twins, and leverages federated learning and blockchain technologies to facilitate cross-domain collaboration, forming a security closed loop of “perception, decision, verification, and optimization.” Five key technologies for 6G network security are introduced: distributed machine learning, AI large models, lightweight authentication and access control, digital twins, and wireless physical layer security technologies. These technologies provide theoretical support and technical pathways for the efficient and trustworthy operation of 6G networks.

Keywords: 6G cyber-security; artificial intelligence; distributed machine learning; large model; lightweight authentication authorization and access control; digital twin; wireless physical layer security