Download pdf
A Lightweight Modulation Recognition Network Based on MobileNetV4
- 1 School of Electronic Science and Engineering, Nanjing University, Nanjing, China, 210023
* Author to whom correspondence should be addressed.
Abstract
Deep learning and neural networks have become widely applied in signal modulation recognition, offering numerous advantages in performance. Traditional approaches to modulation recognition using neural networks typically focus on improving model accuracy, which often results in increased model size and computational complexity. This makes deployment on mobile devices challenging. Therefore, this study aims to reduce the model size while ensuring recognition accuracy and proposes a lightweight neural network architecture based on MobileNetV4. The network incorporates an inverted bottleneck structure, which helps reduce the model’s running time through depthwise separable convolution and residual concatenation. The performance of the model was evaluated on the public dataset RadioML 2018.01A dataset across multiple signal-to-noise ratios and compared with convolutional neural networks and residual networks. The results show that the proposed network reduces the running time to approximately 1/2 to 1/3 of the original models while maintaining comparable or even slightly improved recognition accuracy.
Keywords
Signal modulation identification, deep learning, model lightweighting, inverted bottleneck structure
[1]. Zhao, Y., Xu, Y. T., Jiang, H., Luo, Y. J., & Wang, Z. W. (2015). Recognition of digital modulation signals based on high-order cumulants. In 2015 International Conference on Wireless Communications and Signal Processing (WCSP) (pp. 1–5). Nanjing, China.
[2]. Wei, W., & Mendel, J. M. (2000). Maximum-likelihood classification for digital amplitude-phase modulations. IEEE Transactions on Communications, 48(2), 189–193.
[3]. O’Shea, T. J., & West, N. (2016). Radio machine learning dataset generation with gnu radio. In Proceedings of the GNU Radio Conference (pp. 69–74).
[4]. O’Shea, T. J., Roy, T., & Clancy, T. C. (2018). Over-the-air deep learning based radio signal classification. IEEE Journal on Selected Topics in Signal Processing, 12(1), 168–179.
[5]. Kumar, Y., Sheoran, M., Jajoo, G. (2020). Automatic modulation classification based on constellation density using deep learning. IEEE Communications Letters, 24(6), 1275–1278.
[6]. Chang, S., Huang, S., Zhang, R. Y. (2022). Multitask learning-based deep neural network for automatic modulation classification. IEEE Internet of Things Journal, 9(3), 2192–2206.
[7]. Hakimi, S., & Hodtani, G. A. (2017). Optimized distributed automatic modulation classification in wireless sensor networks using information theoretic measures. IEEE Sensors Journal, 17(10), 3079–3091.
[8]. Lu, M., Peng, T., Yue, G.(2021). Dual-channel hybrid neural network for modulation recognition. IEEE Access, 9, 76260–76269.
[9]. Qi, P., Zhou, X., Zheng, S. (2021). Automatic modulation classification based on deep residual networks with multimodal information. IEEE Transactions on Cognitive Communications and Networking, 7(1), 21–33.
[10]. Zhang, F., Luo, C., Xu, J., Luo, Y., & Zheng, F.-C. (2022). Deep learning based automatic modulation recognition: Models, datasets, and challenges. Digital Signal Processing, 129, 103650. https://doi.org/10.1016/j.dsp.2022.103650
Cite this article
Wu,Y. (2025). A Lightweight Modulation Recognition Network Based on MobileNetV4. Applied and Computational Engineering,146,43-50.
Data availability
The datasets used and/or analyzed during the current study will be available from the authors upon reasonable request.
Disclaimer/Publisher's Note
The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of EWA Publishing and/or the editor(s). EWA Publishing and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
About volume
Volume title: Proceedings of SEML 2025 Symposium: Machine Learning Theory and Applications
© 2024 by the author(s). Licensee EWA Publishing, Oxford, UK. This article is an open access article distributed under the terms and
conditions of the Creative Commons Attribution (CC BY) license. Authors who
publish this series agree to the following terms:
1. Authors retain copyright and grant the series right of first publication with the work simultaneously licensed under a Creative Commons
Attribution License that allows others to share the work with an acknowledgment of the work's authorship and initial publication in this
series.
2. Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the series's published
version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgment of its initial
publication in this series.
3. Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and
during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See
Open access policy for details).