Abstract:Machine learning techniques, particularly deep reinforcement learning (DRL), have been proposed as powerful tools for optimizing slice admission control (SAC) in cloud radio access networks (C‐RANs) within the context of 5G networks. DRL‐based approaches enable SAC systems to learn from historical data and dynamically adapt to changing network conditions, addressing the dynamic and heterogeneous challenges posed by C‐RANs. The research presents a comprehensive classification of DRL methods exclusively tailored for C‐RAN architecture, specifically delving into the use of DRL and methods for controlling slice acceptance. This unique analysis sheds light on the intricacies of adapting these approaches to the challenges posed by C‐RANs, offering valuable insights and guidance for researchers and practitioners navigating the landscape of SAC optimization in the evolving realm of 5G networks. While DRL‐based SAC methods offer potential benefits such as improved network performance and efficiency, deploying these systems in real‐world scenarios presents several challenges and trade‐offs that need to be carefully considered. Further research and development are required to address these challenges and ensure the successful deployment of DRL‐based SAC systems in wireless networks, particularly in the context of C‐RANs. Summary The emergence of 5G networks has increased the demand for network resources, making efficient resource management crucial. Slice admission control (SAC) is a process that governs the creation and allocation of virtualized network environments, known as "network slices," which can be tailored to meet specific user requirements. However, traditional SAC methods face dynamic and heterogeneous challenges in wireless networks, especially in cloud radio access networks (C‐RANs). To address this issue, machine learning (ML) techniques, particularly deep reinforcement learning (DRL), have been proposed as powerful tools for optimizing SAC. DRL‐based approaches enable SAC systems to learn from previous interactions with the network environment and dynamically adapt to changing network conditions. This review article comprehensively explains the current state‐of‐the‐art DRL‐based SAC, focusing on C‐RANs. The article identifies key challenges and future research directions and highlights the potential benefits of using DRL for SAC, including improved network performance and efficiency. However, deploying these systems in real‐world scenarios presents several challenges and trade‐offs that need to be carefully considered. Further research and development are required to address these challenges and ensure the successful deployment of DRL‐based SAC systems in wireless networks.

Neural Architecture Search-enabled Deep Reinforcement Learning for Slice Deployment in a Converged Optical-Wireless Access Network

Toward Scalable and Efficient Hierarchical Deep Reinforcement Learning for 5G RAN Slicing

Towards efficient RAN slicing: A deep hierarchical reinforcement learning approach

RAN Slice Strategy Based on Deep Reinforcement Learning for Smart Grid

Deep Reinforcement Learning for Resource Management in Network Slicing

OnSlicing: Online End-to-End Network Slicing with Reinforcement Learning

Constrained Reinforcement Learning for Resource Allocation in Network Slicing

Digital Twin-Enhanced Deep Reinforcement Learning for Resource Management in Networks Slicing

Attention-based Open RAN Slice Management using Deep Reinforcement Learning

Safe and Accelerated Deep Reinforcement Learning-based O-RAN Slicing: A Hybrid Transfer Learning Approach

Evolutionary Deep Reinforcement Learning for Dynamic Slice Management in O-RAN

Safe and Fast Reinforcement Learning for Network Slicing Resource Allocation

Dynamic SDN-based Radio Access Network Slicing with Deep Reinforcement Learning for URLLC and eMBB Services

Ultra-Dense LEO Satellite Access Network Slicing: A Deep Reinforcement Learning Approach

Using Deep Reinforcement Learning for 5G RAN Slicing Resource Allocation in New Power Load Management System

Deep Reinforcement Learning Approaches to Network Slice Scaling and Placement: A Survey

Deep Reinforcement Learning for End-to-End Network Slicing: Challenges and Solutions

Communication and Computation O-RAN Resource Slicing for URLLC Services Using Deep Reinforcement Learning

Demo: Deep Reinforcement Learning for Resource Management in Cellular Network Slicing.

Network Slicing via Transfer Learning aided Distributed Deep Reinforcement Learning

Slice admission control in 5G cloud radio access network using deep reinforcement learning: A survey