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

Review Computer Science, Interdisciplinary Applications

Reinforcement learning for combinatorial optimization: A survey

Nina Mazyavkina et al.

Summary: The study explores the recent advancements in applying reinforcement learning frameworks to hard combinatorial problems, showcasing their potential in training agents to automate the search for heuristics. It also compares RL methods with traditional algorithms, demonstrating that RL models could be a promising direction for solving combinatorial problems.

COMPUTERS & OPERATIONS RESEARCH (2021)

Add to Collection

Article Computer Science, Information Systems

Scaling UPF Instances in 5G/6G Core With Deep Reinforcement Learning

Hai T. Nguyen et al.

Summary: This paper investigates the application of Deep Reinforcement Learning (DRL) for scaling UPF instances in containers, proposing a threshold-based reward function and adapting the proximal policy optimization (PPO) algorithm. Additionally, a support vector machine (SVM) classifier is utilized to handle unwanted actions suggested by the agent due to stochastic policy. Extensive numerical results demonstrate that the proposed approach outperforms Kubernetes's built-in Horizontal Pod Autoscaler (HPA), with DRL saving 2.7-3.8% of average number of Pods and SVM achieving 0.7-4.5% saving compared to HPA.

IEEE ACCESS (2021)

Add to Collection

Article Computer Science, Information Systems

Network Slice Reconfiguration by Exploiting Deep Reinforcement Learning With Large Action Space

Fengsheng Wei et al.

IEEE TRANSACTIONS ON NETWORK AND SERVICE MANAGEMENT (2020)

Add to Collection

Proceedings Paper Computer Science, Hardware & Architecture

Delay-Aware NFV Resource Allocation with Deep Reinforcement Learning

Ningcheng Yuan et al.

NOMS 2020 - PROCEEDINGS OF THE 2020 IEEE/IFIP NETWORK OPERATIONS AND MANAGEMENT SYMPOSIUM 2020: MANAGEMENT IN THE AGE OF SOFTWARIZATION AND ARTIFICIAL INTELLIGENCE (2020)

Add to Collection

Proceedings Paper Computer Science, Hardware & Architecture