Transformer-Based Modulation Recognition with Mirroring Data Augmentation and Multi-Scale Convolutional Feature Fusion

LI Guojun; CHEN Shi’ao; WANG Jie; ZHANG Zheming; ZHENG Jianzhong

doi:10.12263/DZXB.20250566

您当前的位置：

首页 >

文章列表页 >

Transformer-Based Modulation Recognition with Mirroring Data Augmentation and Multi-Scale Convolutional Feature Fusion

PAPERS | 更新时间：2026-06-04

- Transformer-Based Modulation Recognition with Mirroring Data Augmentation and Multi-Scale Convolutional Feature Fusion
- ACTA ELECTRONICA SINICA Vol. 54, Issue 1, Pages: 329-339(2026)
- 作者机构：
  
  1.重庆邮电大学通信与信息工程学院，重庆 400065
  2.重庆邮电大学超视距可信信息传输研究所，重庆 400065
- 作者简介：
- 基金信息：
  
  National Natural Science Foundation of China(U22A2006);Natural Science Foundation of Chongqing, China(CSTB2023NSCQ-LZX0082);Basic Strengthening Project of the Science and Technology Committee(2023-ZQ-080);Key Basic Research Project of the Science and Technology Committee(2022-173-SD-1)
- DOI：10.12263/DZXB.20250566
  CLC： TN911.3;TP18
- Received：28 June 2025，
  
  Accepted：19 January 2026，
  
  Published：25 January 2026
- 稿件说明：
移动端阅览
李国军, 陈世奥, 王杰, 等. 基于镜像数据增强与多尺度特征卷积融合的Transformer调制识别技术[J]. 电子学报, 2026, 54(01): 329-339.

LI Guojun, CHEN Shi’ao, WANG Jie, et al. Transformer-Based Modulation Recognition with Mirroring Data Augmentation and Multi-Scale Convolutional Feature Fusion[J]. Acta Electronica Sinica, 2026, 54(01): 329-339.
李国军, 陈世奥, 王杰, 等. 基于镜像数据增强与多尺度特征卷积融合的Transformer调制识别技术[J]. 电子学报, 2026, 54(01): 329-339. DOI：10.12263/DZXB.20250566

LI Guojun, CHEN Shi’ao, WANG Jie, et al. Transformer-Based Modulation Recognition with Mirroring Data Augmentation and Multi-Scale Convolutional Feature Fusion[J]. Acta Electronica Sinica, 2026, 54(01): 329-339. DOI：10.12263/DZXB.20250566

摘要

调制识别是通信对抗的关键技术，现有的基于深度学习的调制识别研究大多在仿真数据集或开源数据集上进行，导致训练所得的模型在实际应用时面临着不能适应具体场景的巨大挑战。本文首先提出了一种针对调制信号的镜像数据增强方法，以矢量信号源发射且用接收机接收得到的信号作为原始数据，通过滤波、不同的速率抽取、相移、频移、加噪操作，实现信号增强。这样增强生成的数据集能够适应实际场景中不同的符号速率、多普勒频移、接收机载波偏移、信噪比（Signal-to-Noise Ratios，SNRs）、接收机特征等各种要素的影响，与实际信号相近。接着设计一个信号模态变换模块，对采集的同向正交信号样本进行模态变换，为后续的多模态处理提供数据基础。随后设计了多模态多尺度卷积融合加压缩激励去噪机制的Transformer调制识别模型。之后，用RadioML2018.10a的数据对提出设计模型进行训练，信噪比12 dB以上时测试集准确率达到98.3%，但此时训练的模型用于实际场景测试时的结果仅10.4%。最后使用增强的数据集对提出的模型进行训练，信噪比16 dB以上平均准确率为90.1%；使用训练所得的模型进行在线落地测试，信噪比12 dB时识别率为91.9%。

Abstract

Modulation recognition is a key technology in communication countermeasures. Most of the existing modulation recognition studies based on deep learning are conducted on simulated datasets or open-source datasets. As a result

the models obtained through training face huge challenges of being unable to adapt to specific scenarios in practical applications. First

this paper proposes a mirror data augmentation method for modulated signals. The signals transmitted by a vector signal source and received by a receiver are used as the original data. Signal augmentation is achieved through operations such as filtering

different rate sampling

phase shift

frequency shift

and noise addition. The augmented dataset generated in this way can adapt to the influence of various factors in real-world scenarios

such as different symbol rates

Doppler frequency shifts

receiver carrier offsets

signal-to-noise ratios (SNRs)

and receiver characteristics

and is similar to real signals. Next

a signal modality transformation module is designed to perform modality transformation on IQ sampling data samples

providing a data basis for subsequent multi-modality processing. Then

a Transformer-based modulation recognition model with multi-modality multi-scale convolution fusion and SE denoising mechanism is designed. After that

the proposed model is trained using the RadioML2018.10a dataset. When the SNR is above 12 dB

the accuracy of the test set reaches 98.3%. However

when the trained model is used for testing in real-world scenarios

the result is only 10.4%. Finally

the proposed model is trained using the augmented dataset. When the SNR is above 16 dB

the average accuracy is 90.1%. The trained model is used for online practical testing

and the recognition rate reaches 91.9% when the SNR is 12 dB.

关键词

Keywords

references

Dobre O A , Abdi A , Bar-Ness Y , et al . Survey of automatic modulation classification techniques: Classical approaches and new trends [J ] . IET Communications , 2007 , 1 ( 2 ): 137 - 156 . DOI: 10.1049/iet-com:20050176 http://dx.doi.org/10.1049/iet-com:20050176

Zhu Zhechen , Nandi A K . Blind digital modulation classification using minimum distance centroid estimator and non-parametric likelihood function [J ] . IEEE Transactions on Wireless Communications , 2014 , 13 ( 8 ): 4483 - 4494 . DOI: 10.1109/twc.2014.2320724 http://dx.doi.org/10.1109/twc.2014.2320724

Zheng Jianping , Lv Yanfang . Likelihood-based automatic modulation classification in OFDM with index modulation [J ] . IEEE Transactions on Vehicular Technology , 2018 , 67 ( 9 ): 8192 - 8204 . DOI: 10.1109/tvt.2018.2839735 http://dx.doi.org/10.1109/tvt.2018.2839735

Chang D C , Shih P K . Cumulants-based modulation classification technique in multipath fading channels [J ] . IET Communications , 2015 , 9 ( 6 ): 828 - 835 . DOI: 10.1049/iet-com.2014.0773 http://dx.doi.org/10.1049/iet-com.2014.0773

Wang Sese , Sun Zhuo , Liu Siyuan , et al . Modulation classification of linear digital signals based on compressive sensing using high-order moments [C ] // 2014 European Modelling Symposium . Piscataway : IEEE , 2014 : 145 - 150 . DOI: 10.1109/ems.2014.25 http://dx.doi.org/10.1109/ems.2014.25

Qu Zhiyu , Hou Chenfan , Hou Changbo , et al . Radar signal intra-pulse modulation recognition based on convolutional neural network and deep Q-learning network [J ] . IEEE Access , 2020 , 8 : 49125 - 49136 . DOI: 10.1109/access.2020.2980363 http://dx.doi.org/10.1109/access.2020.2980363

李钦 , 刘伟 , 牛朝阳 , 等 . 低信噪比下基于分裂EfficientNet网络的雷达信号调制方式识别 [J ] . 电子学报 , 2023 , 51 ( 3 ): 675 - 686 .

Li Qin , Liu Wei , Niu Chaoyang , et al . Radar signal modulation recognition based on split EfficientNet under low signal-to-noise ratio [J ] . Acta Electronica Sinica , 2023 , 51 ( 3 ): 675 - 686 . (in Chinese)

Chen Yantao , Dong Binhong , Liu Cuiting , et al . Abandon locality: Frame-wise embedding aided transformer for automatic modulation recognition [J ] . IEEE Communications Letters , 2023 , 27 ( 1 ): 327 - 331 . DOI: 10.1109/lcomm.2022.3213523 http://dx.doi.org/10.1109/lcomm.2022.3213523

Pan Yiyu , Gao Feng . Signal modulation recognition of satellite Internet based on dual input convolutional neural network [C ] // 2024 9th International Conference on Signal and Image Processing . Piscataway : IEEE , 2024 : 13 - 16 . DOI: 10.1109/icsip61881.2024.10671409 http://dx.doi.org/10.1109/icsip61881.2024.10671409

Gao Jiaxin , Cao Qinglong , Chen Yutian . MoE-AMC: Enhancing automatic modulation classification performance using mixture-of-experts [PP/OL ] . V1.arXiv ( 2023-12-04 )[ 2025-06-01 ] . https://doi.org/10.48550/arXiv.2312.02298 https://doi.org/10.48550/arXiv.2312.02298 .

Zhu Zhengyu , Zhou Ning , Wang Zixuan , et al . MSGNet: A multi-feature lightweight learning network for automatic modulation recognition [J ] . IEEE Communications Letters , 2024 , 28 ( 11 ): 2553 - 2557 . DOI: 10.1109/lcomm.2024.3414617 http://dx.doi.org/10.1109/lcomm.2024.3414617

Liu Fugang , Pan Jingyi , Zhou Ruolin . Contrastive learning-based multimodal fusion model for automatic modulation recognition [J ] . IEEE Communications Letters , 2024 , 28 ( 1 ): 78 - 82 . DOI: 10.1109/lcomm.2023.3336049 http://dx.doi.org/10.1109/lcomm.2023.3336049

Xing Jingyi , Xu Shengfa , Xu Chunpeng , et al . A lightweight modulation recognition network with depthwise separable convolution and SNR enhancement [C ] // 2025 7th International Conference on Electronics and Communication, Network and Computer Technology . Piscataway : IEEE , 2025 : 530 - 534 . DOI: 10.1109/ecnct66493.2025.11172616 http://dx.doi.org/10.1109/ecnct66493.2025.11172616

Lu Yitong , Hou Shujuan , Zhang Qin , et al . DAWFNN: An automatic modulation recognition method based on multi feature fusion [C ] // 2025 10th International Conference on Computer and Communication System . Piscataway : IEEE , 2025 : 480 - 485 . DOI: 10.1109/icccs65393.2025.11069634 http://dx.doi.org/10.1109/icccs65393.2025.11069634

Vaswani A , Shazeer N , Parmar N , et al . Attention is all you need [C ] // Proceedings of the 31st International Conference on Neural Information Processing Systems . New York : Curran Associates Inc. , 2017 : 6000 - 6010 .

Dao T T , Noh D I , Pham Q V , et al . VT-MCNet: High-accuracy automatic modulation classification model based on vision transformer [J ] . IEEE Communications Letters , 2024 , 28 ( 1 ): 98 - 102 . DOI: 10.1109/lcomm.2023.3336985 http://dx.doi.org/10.1109/lcomm.2023.3336985

Dosovitskiy A , Beyer L , Kolesnikov A , et al . An image is worth 16 × 16 words: Transformers for image recognition at scale [PP/OL ] . V2.arXiv ( 2021-06-03 )[ 2025-06-01 ] . https://doi.org/10.48550/arXiv.2010.11929 https://doi.org/10.48550/arXiv.2010.11929 .

李广丽 , 叶艺源 , 吴光庭 , 等 . 联合多视角Transformer编码与在线融合互学习的乳腺癌病理图像分类模型 [J ] . 电子学报 , 2024 , 52 ( 7 ): 2369 - 2381 . DOI: 10.12263/DZXB.20230305 http://dx.doi.org/10.12263/DZXB.20230305

Li Guangli , Ye Yiyuan , Wu Guangting , et al . Breast cancer pathological image classification model via combining multi-view transformer coding and online fusion mutual learning [J ] . Acta Electronica Sinica , 2024 , 52 ( 7 ): 2369 - 2381 . (in Chinese) . DOI: 10.12263/DZXB.20230305 http://dx.doi.org/10.12263/DZXB.20230305

Zeng Rui , Lu Zhilin , Zhang Xudong , et al . Convolutional neural network assisted transformer for automatic modulation recognition under large CFOs and SROs [J ] . IEEE Signal Processing Letters , 2024 , 31 : 741 - 745 . DOI: 10.1109/lsp.2024.3372770 http://dx.doi.org/10.1109/lsp.2024.3372770

O’Shea T J , Roy T , Clancy T C . Over-the-air deep learning based radio signal classification [J ] . IEEE Journal of Selected Topics in Signal Processing , 2018 , 12 ( 1 ): 168 - 179 . DOI: 10.1109/jstsp.2018.2797022 http://dx.doi.org/10.1109/jstsp.2018.2797022

Tekbıyık K , Ekti A R , Görçin A , et al . Robust and fast automatic modulation classification with CNN under multipath fading channels [C ] // 2020 IEEE 91st Vehicular Technology Conference . Piscataway : IEEE , 2020 : 1 - 6 . DOI: 10.1109/vtc2020-spring48590.2020.9128408 http://dx.doi.org/10.1109/vtc2020-spring48590.2020.9128408

Wei Tao , Li Zan , Bi Dexin , et al . Adaptive multi-dimensional shrinkage block for automatic modulation recognition [J ] . IEEE Communications Letters , 2023 , 27 ( 11 ): 2968 - 2972 . DOI: 10.1109/lcomm.2023.3314623 http://dx.doi.org/10.1109/lcomm.2023.3314623

Tunze G B , Huynh-The T , Lee J M , et al . Sparsely connected CNN for efficient automatic modulation recognition [J ] . IEEE Transactions on Vehicular Technology , 2020 , 69 ( 12 ): 15557 - 15568 . DOI: 10.1109/tvt.2020.3042638 http://dx.doi.org/10.1109/tvt.2020.3042638

Xu Jialang , Luo Chunbo , Parr G , et al . A spatiotemporal multi-channel learning framework for automatic modulation recognition [J ] . IEEE Wireless Communications Letters , 2020 , 9 ( 10 ): 1629 - 1632 . DOI: 10.1109/lwc.2020.2999453 http://dx.doi.org/10.1109/lwc.2020.2999453

Rajendran S , Meert W , Giustiniano D , et al . Deep learning models for wireless signal classification with distributed low-cost spectrum sensors [J ] . IEEE Transactions on Cognitive Communications and Networking , 2018 , 4 ( 3 ): 433 - 445 . DOI: 10.1109/tccn.2018.2835460 http://dx.doi.org/10.1109/tccn.2018.2835460

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Variable Horizon Multi-Directional Scanning Method for Time Series Anomaly Detection

Object Detection Based on EIMYOLO for High-Resolution Remote Sensing Images

Method on Alzheimer’s Disease Classification Utilizing Deep Learning and Hippocampus Heterogeneous Feature Fusion

Radar Signal Modulation Recognition Based on Split EfficientNet Under Low Signal-to-Noise Ratio

A Priority-Weighted and Packet Arrival Time Based Scheduling Algorithm for MP-QUIC

Related Author

HUANG Yu-zhe

GUAN Yong-yuan

WEI Song-jie

CAO Feng

ZENG Ke-wen

LI De-yu

LUO Xi-zhao

TAO Chong-ben

Related Institution

School of Computer Science and Engineering, School of Cyber Science and Engineering, Nanjing University of Science and Technology

School of Information and Technology, Shanxi University

School of Computer Science and Technology, Soochow University

School of Electronics and Information Engineering, Suzhou University of Science and Technology

School of Microelectronics and Communication Engineering， Chongqing University

⁰