On Enhancing Network Throughput using Reinforcement Learning in Sliced Testbeds

TL;DR

This paper presents the eMBB-Agent, a reinforcement learning-based approach using Deep Q-Networks to optimize network slicing throughput, demonstrating improved SLA adherence through experimental analysis of various parameters.

Contribution

It introduces the eMBB-Agent, a novel RL-based method for enhancing network slicing throughput in testbeds, integrating application analysis and discrete action proposals.

Findings

Higher throughput achieved with optimized DQN layers and learning rates.

Channel error rate significantly affects model convergence and performance.

Embedding RL in network slicing improves SLA compliance and adaptability.

Abstract

Novel applications demand high throughput, low latency, and high reliability connectivity and still pose significant challenges to slicing orchestration architectures. The literature explores network slicing techniques that employ canonical methods, artificial intelligence, and combinatorial optimization to address errors and ensure throughput for network slice data plane. This paper introduces the Enhanced Mobile Broadband (eMBB)-Agent as a new approach that uses Reinforcement Learning (RL) in a vertical application to enhance network slicing throughput to fit Service-Level Agreements (SLAs). The eMBB-Agent analyzes application transmission variables and proposes actions within a discrete space to adjust the reception window using a Deep Q-Network (DQN). This paper also presents experimental results that examine the impact of factors such as the channel error rate, DQN model layers, and learning rate on model convergence and achieved throughput, providing insights on embedding intelligence in network slicing.