Neural Network-based Quantization for Network Automation

TL;DR

This paper introduces a neural network-based implementation of the Bounding Sphere Quantization algorithm, significantly improving training speed for network management tasks like anomaly detection in mobile networks.

Contribution

It presents a novel neural network approach to BSQ, enabling faster training while maintaining effective quantization for network automation applications.

Findings

Neural network implementation of BSQ achieves faster training times.

The approach maintains quantization quality suitable for network management.

Applicable to anomaly detection in mobile networks.

Abstract

Deep Learning methods have been adopted in mobile networks, especially for network management automation where they provide means for advanced machine cognition. Deep learning methods utilize cutting-edge hardware and software tools, allowing complex cognitive algorithms to be developed. In a recent paper, we introduced the Bounding Sphere Quantization (BSQ) algorithm, a modification of the k-Means algorithm, that was shown to create better quantizations for certain network management use-cases, such as anomaly detection. However, BSQ required a significantly longer time to train than k-Means, a challenge which can be overcome with a neural network-based implementation. In this paper, we present such an implementation of BSQ that utilizes state-of-the-art deep learning tools to achieve a competitive training speed.