Towards Intelligent Wireless Networks: The Synergy of Generative AI and Digital Twins

This paper proposes a generative AI (GenAI)-enabled digital twin (DT) framework for proactive and energy-aware wireless optimization in future 6G ecosystems. Most existing AI-assisted DT approaches remain fundamentally reactive, adjusting network parameters only after performance degradation occurs or restricting GenAI to isolated signal-level tasks such as channel estimation. This work adopts a proactive approach. Instead of responding to problems after they appear, the proposed framework continuously synchronizes channel states, mobility dynamics, traffic conditions, and energy information within a real-time DT environment, enabling the system to anticipate congestion, interference, and energy demand before they materialize. The result is a closed-loop proactive architecture that operates at the system level, jointly managing communication, mobility, and resource dynamics for autonomous wireless control. Evaluations on a UAV-assisted non-terrestrial network (NTN) scenario show approximately 69.2\% energy savings over reactive baselines while maintaining reliable quality-of-service (QoS) under dense and mobility-intensive conditions. Beyond this specific scenario, the framework offers a scalable foundation for broader AI-native 6G applications, including aerial platforms, autonomous systems, extended reality (XR), industrial automation, and space-air-ground-sea (SAGS) integrated infrastructures.