Radio Access Network Slicing

Radio Access Network (RAN) slicing dynamically partitions a shared physical network into multiple virtual networks, each tailored to specific application needs (e.g., high-bandwidth video streaming, low-latency industrial control). Current research heavily focuses on intelligent resource allocation within these slices, employing reinforcement learning (RL) algorithms, often enhanced by techniques like transfer learning and graph neural networks, to optimize resource management and ensure service-level agreements are met. This work aims to improve the efficiency and adaptability of RAN slicing, particularly in open and virtualized architectures like O-RAN, leading to more flexible and cost-effective network deployments for diverse applications. The ultimate goal is to enable seamless and efficient support for a wide range of services with varying quality-of-service requirements.