PRAXIS: Proximal Reinforcement-aided Adaptive eXpert Intelligent System for Efficient Traffic Management in URLLC

Proximal Reinforcement-aided Adaptive eXpert Intelligent System (PRAXIS) is an innovative decentralized framework leveraging reinforcement learning to optimize traffic management in 5G or even 6G Ultra-Reliable Low-Latency Communications (URLLC) scenarios, ensuring efficiency and reliability under stringent network demands. By leveraging multiple autonomous agents to optimize traffic flow and mitigate congestion dynamically, PRAXIS ensures low latency and high reliability. The framework integrates adaptive traffic scheduling and scalable resource allocation, enabling real-time decision-making based on network conditions. Experimental results demonstrate PRAXIS’s ability to outperform conventional methods by significantly enhancing network efficiency, reducing latency, and improving overall reliability, offering a scalable solution tailored to the demands of 5G and 6G networks.