Otimização Conjunta e Escalável para Open RAN: Associação UE/O-RU e Posicionamento de VNFs

O-RAN has emerged as a promising architecture for scalable mobile networks, however, its disaggregation into VNFs poses significant challenges in orchestrating these distributed units. This work proposes a unified framework for the joint optimization of UE/O-RU association and O-DU and O-CU placement, considering QoS requirements of multiple network slices. The problem is initially formulated as an ILP model and subsequently addressed using a deep reinforcement learning-based approach. The results indicate gains of up to 20\% in user admission ratio, improved fairness among slices, and more efficient resource utilization, demonstrating robustness and practical applicability in large-scale scenarios.