MicroOpt: Model-driven Slice Resource Optimization in 5G and Beyond Networks

TL;DR

MicroOpt is a novel neural network-based framework that optimizes resource allocation in 5G slices, improving efficiency and QoS adherence over existing methods using real-world traffic data.

Contribution

It introduces a differentiable neural network model combined with gradient descent and Lagrangian decomposition for effective resource optimization in 5G slices.

Findings

Achieves up to 21.9% resource efficiency improvement

Effective across various QoS thresholds and traffic dynamics

Outperforms state-of-the-art approaches in real-world tests

Abstract

A pivotal attribute of 5G networks is their capability to cater to diverse application requirements. This is achieved by creating logically isolated virtual networks, or slices, with distinct service level agreements (SLAs) tailored to specific use cases. However, efficiently allocating resources to maintain slice SLA is challenging due to varying traffic and QoS requirements. Traditional peak traffic-based resource allocation leads to over-provisioning, as actual traffic rarely peaks. Additionally, the complex relationship between resource allocation and QoS in end-to-end slices spanning different network segments makes conventional optimization techniques impractical. Existing approaches in this domain use network models or simulations and various optimization methods but struggle with optimality, tractability, and generalizability across different slice types. In this paper, we propose MicroOpt, a novel framework that leverages a differentiable neural network-based slice model with gradient descent for resource optimization and Lagrangian decomposition for QoS constraint satisfaction. We evaluate MicroOpt against two state-of-the-art approaches using an open-source 5G testbed with real-world traffic traces. Our results demonstrate up to 21.9% improvement in resource allocation compared to these approaches across various scenarios, including different QoS thresholds and dynamic slice traffic.