DRL-based xApps for Dynamic RAN and MEC Resource Allocation and Slicing in O-RAN

The evolution of wireless communication technologies, transitioning from the established fifth-generation (5G) to beyond 5G and ultimately the sixth-generation (6G), underscores the necessity for a paradigmatic shift in architecture. This transformation is essential to effectively accommodate the forthcoming surge in multi-connectivity and on-demand services.Machine Learning (ML) and more specifically Deep Reinforcement Learning (DRL) are a promising approach to solving the aforementioned challenges. In this paper, we propose an approach for dynamic resource allocation and the admission and management of slicing at both Radio Access Network (RAN) and Multi-Access Edge Computing (MEC) levels leveraging Deep Q-Network (DQN). This methodology ensures effective resource allocation while upholding the Quality of Services (QoS). The proposed solution is validated through simulation results, demonstrating its effectiveness in improving network efficiency and performance in future 5G and 6G networks.