A Modulation Recognition Method Combining Wavelet Denoising Convolution and Sparse Transformer

ZHENG Qinghe; LIU Fanglin; YU Lisu; JIANG Weiwei; HUANG Chongwen; GUI Guan

doi:10.11999/JEIT241159

Volume 47 Issue 7

Jul. 2025

Turn off MathJax

Article Contents

Article Navigation > Journal of Electronics & Information Technology > 2025 > 47(7): 2361-2374

ZHENG Qinghe, LIU Fanglin, YU Lisu, JIANG Weiwei, HUANG Chongwen, GUI Guan. A Modulation Recognition Method Combining Wavelet Denoising Convolution and Sparse Transformer[J]. Journal of Electronics & Information Technology, 2025, 47(7): 2361-2374. doi: 10.11999/JEIT241159

Citation:

ZHENG Qinghe, LIU Fanglin, YU Lisu, JIANG Weiwei, HUANG Chongwen, GUI Guan. A Modulation Recognition Method Combining Wavelet Denoising Convolution and Sparse Transformer[J]. Journal of Electronics & Information Technology, 2025, 47(7): 2361-2374. doi: 10.11999/JEIT241159

Citation:

ZHENG Qinghe, LIU Fanglin, YU Lisu, JIANG Weiwei, HUANG Chongwen, GUI Guan. A Modulation Recognition Method Combining Wavelet Denoising Convolution and Sparse Transformer[J]. Journal of Electronics & Information Technology, 2025, 47(7): 2361-2374. doi: 10.11999/JEIT241159

PDF( 5646 KB)

A Modulation Recognition Method Combining Wavelet Denoising Convolution and Sparse Transformer

doi: 10.11999/JEIT241159 cstr: 32379.14.JEIT241159

1.
School of Intelligent Engineering, Shandong Management University, Jinan 250357, China
2.
School of Information Engineering, Nanchang University, Nanchang 330031, China
3.
School of Information and Communication Engineering, Beijing University of Posts and Telecommunications, Beijing 100876, China
4.
School of Information and Electronic Engineering, Zhejiang University, Hangzhou 310058, China
5.
School of Communication and Information Engineering, Nanjing University of Posts and Telecommunications, Nanjing 210023, China

Funds: The National Key R&D Program (2018YFF01014304), The National Natural Science Foundation of China (62401070), Shandong Provincial Key R&D Program (2024TSGC0055), Shandong Provincial Natural Science Foundation (ZR2019ZD01, ZR2023QF125), Shandong Provincial Youth Innovation Team Plan of Higher Education Institutions (2024KJH005)

Received Date: 2024-12-31
Rev Recd Date: 2025-05-13

Available Online: 2025-05-26

Publish Date: 2025-07-22

Abstract

Abstract

Objective Automatic Modulation Classification (AMC) is a key process in signal detection and demodulation, enabling the automatic identification of various modulation schemes in non-cooperative communication scenarios. The high transmission rate and low latency requirements of 6G wireless communications necessitate AMC to accurately and promptly identify different modulation schemes to adapt to complex communication scenes. As integrated 6G communication, sensing, and computing technologies evolve, an increasing number of modulation schemes are being designed and adopted. In the face of a more complex electromagnetic environment, AMC encounters challenges such as low recognition accuracy, susceptibility to environmental interference, and poor robustness. Existing feature selection methods struggle to generalize well in practical applications. By contrast, integrating adaptive feature selection methods with deep learning can generate more reliable and precise classification constraints, even for complex signal features that generalize across time. This study proposes an AMC method that combines learnable wavelet denoising convolution with a sparse Transformer, providing technical support for the advancement of integrated communication, sensing, and computing. Methods In the proposed AMC method combining wavelet denoising convolution and sparse Transformer, the learnable wavelet denoising convolution is proposed to assist deep learning in extracting appropriate denoised signal representations. It incorporates adaptive time-frequency features into the functional strategy of the objective function, delivering accurate spatiotemporal information to the Transformer and mitigating the effects of signal parameter offsets. A Sparse Feedforward Neural Network (SFFN) is then designed to replace the attention mechanism in the Transformer, modeling element relationships based on a limited set of key elements in the signal domain. This approach effectively optimizes gradients during training to address the challenges posed by limited signal length and the disregard of ordered feature element correlations in traditional Transformers. SFFN enhances spatial sampling positions within sparse cells by introducing additional offsets, focusing on a small group of key sampling points around the reference point. Moreover, it learns these offsets from the target task without requiring additional supervision. Given that most AMC frameworks benefit from multi-scale or multidimensional features, the proposed SFFN can be extended to multi-scale features without the need for feature pyramid networks. Results and Discussions During the experimental phase, modulation recognition performance testing, ablation studies, and comparative analyses are conducted on two publicly available datasets (RadioML 2016.10a and RML22) to demonstrate the excellent performance of the proposed method. Experimental results show that the sparse Transformer achieves classification accuracies of 63.84% and 71.13% on RadioML 2016.10a and RML22, respectively (Fig. 4). The primary limitation in modulation recognition performance occurs in the classification between the following modulation schemes: {WBFM, AM-DSB, QPSK, 8PSK, 16QAM, 64QAM} (Fig. 5). By comparing the sparse Transformer to a range of deep learning models (including CGDNet, CLDNN, DenseNet, GRU, and ResNet on RadioML 2016.10a, and DAENet, ICNet, LSTM, MCDNN, and MCNet on RML22), experimental results confirm the superior performance of the proposed method (Table 1). The classification accuracy of all models at various Signal-to-Noise Ratios (SNRs) is shown to highlight the performance advantage of the proposed method (Fig. 6). To qualitatively explore the effect of wavelet denoising convolution on improving classification accuracy, the power spectrum in the wavelet domain, represented by the selected denoised signals in RadioML 2016.10a, is presented (Fig. 7). The different wavelet coefficients selected for denoising signal representation reveal the component specificity of different wavelet basis functions in feature extraction, which helps determine the applicable signal types or model architectures (Figure 8). The sampling points of interest in the power concentration area are visualized on the wavelet scale map to observe the corresponding relationships between the elements learned by SFFN (Fig. 9). Furthermore, the denoised signal representation constructed by wavelet denoising convolution under different hyperparameter settings undergoes an ablation study to evaluate its impact on model performance (Table 2). The model structure is ablated and tested to observe the roles played by different modules (Table 3). Finally, different numbers of training samples are set to demonstrate the robustness of the model, with less than a 1.5% decrease in classification accuracy when the training set ratio is adjusted from 90% to 40% (Figure 10). Conclusions This study addresses the challenges of limited time-domain length and the neglect of ordered feature element correlations in Transformer models for signal processing by proposing an effective AMC method that combines wavelet denoising convolution and sparse Transformer. The proposed method significantly enhances AMC accuracy in complex communication scenarios and demonstrates robustness to model structure and hyperparameter selection. First, the learnable wavelet denoising convolution is proposed to assist deep learning models in adaptively extracting appropriate denoised signal representations. Subsequently, the SFFN module is designed to replace the attention mechanism in the Transformer, effectively modeling spatiotemporal element relationships. Experimental results on two public datasets (RadioML 2016.10a and RML22) show that the proposed method significantly improves AMC accuracy. By comparing a range of deep learning models, the proposed method demonstrates state-of-the-art AMC accuracy and robustness across different communication scenarios. Additionally, visualization experiments and ablation studies further confirm the robustness of the proposed method.
- Modulation classification,
- Deep learning,
- Sparse Transformer,
- Wavelet denoising convolution

FullText(HTML)

References(24)

References

[1]	郭业才, 姚文强. 基于信噪比分类网络的调制信号分类识别算法[J]. 电子与信息学报, 2022, 44(10): 3507–3515. doi: 10.11999/JEIT210825. GUO Yecai and YAO Wenqiang. Modulation signal classification and recognition algorithm based on signal to noise ratio classification network[J]. Journal of Electronics & Information Technology, 2022, 44(10): 3507–3515. doi: 10.11999/JEIT210825.
[2]	张正宇, 何睿斯, 杨汨, 等. 面向6G的无线信道语义特征及建模[J]. 电子学报, 2025, 53(1): 14–23. doi: 10.12263/DZXB.20240595. ZHANG Zhengyu, HE Ruisi, YANG Mi, et al. Semantic characteristics and modeling of wireless channels for 6G[J]. Acta Electronica Sinica, 2025, 53(1): 14–23. doi: 10.12263/DZXB.20240595.
[3]	张思成, 林云, 涂涯, 等. 基于轻量级深度神经网络的电磁信号调制识别技术[J]. 通信学报, 2020, 41(11): 12–21. doi: 10.11959/j.issn.1000-436x.2020237. ZHANG Sicheng, LIN Yun, TU Ya, et al. Electromagnetic signal modulation recognition technology based on lightweight deep neural network[J]. Journal on Communications, 2020, 41(11): 12–21. doi: 10.11959/j.issn.1000-436x.2020237.
[4]	孟磊, 曲卫, 马爽, 等. 基于LSTM的雷达脉冲重复间隔调制模式识别[J]. 现代雷达, 2021, 43(1): 50–57. doi: 10.16592/j.cnki.1004-7859.2021.01.008. MENG Lei, QU Wei, MA Shuang, et al. Radar PRI Modulation pattern recognition method based on LSTM[J]. Modern Radar, 2021, 43(1): 50–57. doi: 10.16592/j.cnki.1004-7859.2021.01.008.
[5]	KONG Weisi, JIAO Xun, XU Yuhua, et al. A transformer-based contrastive semi-supervised learning framework for automatic modulation recognition[J]. IEEE Transactions on Cognitive Communications and Networking, 2023, 9(4): 950–962. doi: 10.1109/TCCN.2023.3264908.
[6]	HU Mutian, MA Jitong, YANG Zhengyan, et al. Feature fusion convolution-aided transformer for automatic modulation recognition[J]. IEEE Communications Letters, 2023, 27(10): 2643–2647. doi: 10.1109/LCOMM.2023.3298941.
[7]	KONG Weisi, YANG Qinghai, JIAO Xun, et al. A transformer-based CTDNN structure for automatic modulation recognition[C]. The 7th International Conference on Computer and Communications, Chengdu, China, 2021: 159–163. doi: 10.1109/ICCC54389.2021.9674558.
[8]	ZHU Xizhou, SU Weijie, LU Lewei, et al. Deformable DETR: Deformable transformers for end-to-end object detection[C]. The 9th International Conference on Learning Representations, 2021.
[9]	ZHANG Qi and DENG Linfeng. An intelligent fault diagnosis method of rolling bearings based on short-time Fourier transform and convolutional neural network[J]. Journal of Failure Analysis and Prevention, 2023, 23(2): 795–811. doi: 10.1007/s11668-023-01616-9.
[10]	ZHOU Jie, MENG Ming, GAO Yunyuan, et al. Classification of motor imagery EEG using wavelet envelope analysis and LSTM networks[C]. Chinese Control and Decision Conference, Shenyang, China, 2018: 5600–5605. doi: 10.1109/CCDC.2018.8408108.
[11]	O'SHEA T J, ROY T, and 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.
[12]	SATHYANARAYANAN V, GERSTOFT P, and EL GAMAL A. RML22: Realistic dataset generation for wireless modulation classification[J]. IEEE Transactions on Wireless Communications, 2023, 22(11): 7663–7675. doi: 10.1109/TWC.2023.3254490.
[13]	崔凯, 崔天舒, 朱岩, 等. 基于多尺度时序特征的信号调制样式识别算法[J]. 信号处理, 2021, 37(8): 1507–1517. doi: 10.16798/j.issn.1003-0530.2021.08.018. CUI Kai, CUI Tianshu, ZHU Yan, et al. Signal modulation pattern recognition algorithm based on multiscale temporal features[J]. Journal of Signal Processing, 2021, 37(8): 1507–1517. doi: 10.16798/j.issn.1003-0530.2021.08.018.
[14]	DAI Jifeng, QI Haozhi, XIONG Yuwen, et al. Deformable convolutional networks[C]. IEEE International Conference on Computer Vision, Venice, Italy, 2017: 764–773. doi: 10.1109/ICCV.2017.89.
[15]	NJOKU J N, MOROCHO-CAYAMCELA M E, and LIM W. CGDNet: Efficient hybrid deep learning model for robust automatic modulation recognition[J]. IEEE Networking Letters, 2021, 3(2): 47–51. doi: 10.1109/LNET.2021.3057637.
[16]	WEST N E and O’SHEA T. Deep architectures for modulation recognition[C]. IEEE International Symposium on Dynamic Spectrum Access Networks, Baltimore, USA, 2017: 1–6. doi: 10.1109/DySPAN.2017.7920754.
[17]	LIU Xiaoyu, YANG Diyu, and EL GAMAL A. Deep neural network architectures for modulation classification[C]. The 51st Asilomar Conference on Signals, Systems, and Computers, Pacific Grove, USA, 2017: 915–919. doi: 10.1109/ACSSC.2017.8335483.
[18]	HONG Dehua, ZHANG Zilong, and XU Xiaodong. Automatic modulation classification using recurrent neural networks[C]. The 3rd IEEE International Conference on Computer and Communications, Chengdu, China, 2017: 695–700. doi: 10.1109/CompComm.2017.8322633.
[19]	LU Xiao, TAO Mengyuan, FU Xue, et al. Lightweight network design based on ResNet structure for modulation recognition[C]. IEEE 94th Vehicular Technology Conference, Norman, USA, 2021: 1–5. doi: 10.1109/VTC2021-Fall52928.2021.9625558.
[20]	KE Ziqi and VIKALO H. Real-time radio technology and modulation classification via an LSTM auto-encoder[J]. IEEE Transactions on Wireless Communications, 2022, 21(1): 370–382. doi: 10.1109/TWC.2021.3095855.
[21]	HERMAWAN A P, GINANJAR R R, KIM D S, et al. CNN-based automatic modulation classification for beyond 5G communications[J]. IEEE Communications Letters, 2020, 24(5): 1038–1041. doi: 10.1109/LCOMM.2020.2970922.
[22]	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.
[23]	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.
[24]	HUYNH-THE T, HUA C H, PHAM Q V, et al. MCNet: An efficient CNN architecture for robust automatic modulation classification[J]. IEEE Communications Letters, 2020, 24(4): 811–815. doi: 10.1109/LCOMM.2020.2968030.