A unified AI-driven framework for quantum-secured 6G THz networks with intelligent reflecting surfaces and federated edge learning

Abstract The main contribution of this manuscript is an innovative framework for integrating Artificial Intelligence (AI) in 6G wireless systems. With increased complexity, including bursty traffic, network complexity, and dynamic variability, there is a need for intelligence. This study develops and validates an AI-driven approach that enhances network performance through quantum communication decoding, beamforming, and decentralized edge processing. Kalman filtering predictive models are used to estimate variable channel conditions in a Terahertz (THz) network to support beamforming to optimize beamforming. Artificial Intelligence exploits smart reflective surfaces (IRS) strengthening signals and improving their coverage. Also, strong security of Quantum Key Distribution (QKD) protocols due to AI enhanced error correction technology, and rapid, yet privacy information conducting at edge nodes due to decentralised processing through federated learning are examples of enhanced capabilities. Extensive ns-3 simulations across 100 independent runs validate the framework’s effectiveness and prove the system in practical 6G deployment scenarios including THz links, IRS component and edge nodes. The simulation results demonstrate that the proposed framework achieves superior performance compared to conventional approaches, with statistical validation across multiple deployment scenarios. The system decreases latency by 30%, and adds 25% to spectral efficiency. In bursty traffic, the energy efficiency is increased by 20% and packets delivery ratio (PDR) is boosted by 15%. The AI algorithms work effectively to regulate the channel estimation, beamforming, and resource allocation, and, as a result, showed an improvement in the order of magnitudes over previous studies. These results support the fact that AI demonstrates significant potential for transformative impact to a 6G network. The framework has been efficient in addressing problems of channel estimation, beamforming and distributed processing and novel calculations in quantum communication security protocols. Such findings can be used as the foundation of the further inclusion of AI-based technologies in 6G systems, which will help to deploy robust, resilient, and autonomous wireless networks to address the needs of a connective society.