Research Article
Automatic Modulation Classification Using Quantum-Empowered Training
@INPROCEEDINGS{10.1007/978-3-032-44094-5_16, author={To Truong An and Simon L. Cotton and Hans-J\'{y}rgen Zepernick and Trung Q. Duong}, title={Automatic Modulation Classification Using Quantum-Empowered Training}, proceedings={Industrial Networks and Intelligent Systems. 12th EAI International Conference, INISCOM 2026, Da Nang, Vietnam, February 26--27, 2026, Proceedings}, proceedings_a={INISCOM}, year={2026}, month={11}, keywords={Automatic modulation classification Quantum machine learning Training efficiency}, doi={10.1007/978-3-032-44094-5_16} }- To Truong An
Simon L. Cotton
Hans-Jürgen Zepernick
Trung Q. Duong
Year: 2026
Automatic Modulation Classification Using Quantum-Empowered Training
INISCOM
Springer
DOI: 10.1007/978-3-032-44094-5_16
Abstract
Automatic modulation classification (AMC) is a technique designed to identify the modulation of unknown wireless signals. However, existing AMC systems are resource-intensive since they rely on deep learning (DL) models. Training an AMC model is particularly demanding because a large number of parameters need to be updated repeatedly. In this work, we introduce quantum-empowered training (QET), a framework that enhances the training efficiency of AMC systems. The QET framework is a hybrid quantum classical neural network which aims to generate the classical parameters for the AMC model during training. This indirect training strategy significantly reduces the number of trainable parameters and the overall memory requirements compared to traditional training methods for DL models. The experimental results demonstrate that the QET enables training an AMC model with 70% fewer trainable parameters than classical training methods, while maintaining a competitive classification accuracy.