FAIR-Q: Fairness and Adaptive Intelligent Resource Management with QoS Optimization in Dynamic 6G Radio Access Networks

The advent of 6G networks brings diverse services, such as immersive multimedia, augmented reality, and massive IoT, each with stringent requirements for Quality of Service (QoS) and fairness. These challenges expose the limitations of traditional scheduling algorithms, which struggle to dynamically adapt to evolving network conditions. To address this, we propose FAIR-Q, a novel Fairness and Adaptive Intelligent Resource Management framework with QoS optimization driven by Reinforcement Learning (RL) approach. FAIR-Q integrates a multi-objective reward function to optimize fairness, packet loss, delay, and rate constraints. The framework features two key controllers: a Parameterization Controller, which dynamically adjusts scheduling parameters to ensure fairness, and a Scheduling Rule Controller, which intelligently selects scheduling rules to adapt to real-time network conditions and align with QoS requirements. Simulation results demonstrate up to a 15% improvement in fairness and QoS satisfaction compared to static scheduling methods, underscoring the adaptability and scalability of FAIR-Q in dynamic 6G Radio Access Networks.