Modulation classification based on the collaboration of
dual-channel CNN-LSTM and residual network

doi:10.19682/j.cnki.1005-8885.2022.2012

中国邮电高校学报(英文版) ›› 2022, Vol. 29 ›› Issue (1): 113-124.doi: 10.19682/j.cnki.1005-8885.2022.2012

• • 上一篇

Modulation classification based on the collaboration of dual-channel CNN-LSTM and residual network

Li Hui, Li Shanshan, Zou Borong, Chen Yannan

School of Physics and Electronic Information Engineering, Henan Polytechnic University, Jiaozuo 454003, China

收稿日期:2021-04-09 修回日期:2021-07-06 接受日期:2021-09-27 出版日期:2022-02-26 发布日期:2022-02-28
通讯作者: Corresponding author: Zou Borong E-mail:wdzbr296@hpu.edu.cn
基金资助:
This work was supported by the Project of Henan Science and Technology Research (222102210247).

Modulation classification based on the collaboration of dual-channel CNN-LSTM and residual network

Li Hui, Li Shanshan, Zou Borong, Chen Yannan

School of Physics and Electronic Information Engineering, Henan Polytechnic University, Jiaozuo 454003, China

Received:2021-04-09 Revised:2021-07-06 Accepted:2021-09-27 Online:2022-02-26 Published:2022-02-28
Contact: Corresponding author: Zou Borong E-mail:wdzbr296@hpu.edu.cn
Supported by:
This work was supported by the Project of Henan Science and Technology Research (222102210247).

摘要/Abstract

摘要： Deep learning has recently been progressively introduced into the field of modulation classification due to its wide application in image, vision, and other areas. Modulation classification is not only the priority of cognitive radio and spectrum sensing, but also the link during signal demodulation. Combining the advantages of convolutional neural network (CNN), long short-term memory (LSTM), and residual network (ResNet), a modulation classification method based on dual-channel CNN-LSTM and ResNet is proposed to automatically classify the modulation signal more accurately. Specifically, CNN and LSTM are initially used to form a dual-channel structure to effectively explore the spatial and temporal features of the original complex signal. It solves the problem of only focusing on temporal or spatial aspects, and increases the diversity of features. Secondly, the features extracted from CNN and LSTM are fused, making the extracted features richer and conducive to signal classification. In addition, a convolutional layer is added within the residual unit to deepen the network depth. As a result, more representative features are extracted, improving the classification performance. Finally, simulation results on the radio machine learning (RadioML) 2018.01A dataset signify that the network's classification performance is superior to many classifiers in the literature.

关键词: convolutional neural network, deep neural network, long short-term memory, modulation classification, residual network

Abstract: Deep learning has recently been progressively introduced into the field of modulation classification due to its wide application in image, vision, and other areas. Modulation classification is not only the priority of cognitive radio and spectrum sensing, but also the link during signal demodulation. Combining the advantages of convolutional neural network (CNN), long short-term memory (LSTM), and residual network (ResNet), a modulation classification method based on dual-channel CNN-LSTM and ResNet is proposed to automatically classify the modulation signal more accurately. Specifically, CNN and LSTM are initially used to form a dual-channel structure to effectively explore the spatial and temporal features of the original complex signal. It solves the problem of only focusing on temporal or spatial aspects, and increases the diversity of features. Secondly, the features extracted from CNN and LSTM are fused, making the extracted features richer and conducive to signal classification. In addition, a convolutional layer is added within the residual unit to deepen the network depth. As a result, more representative features are extracted, improving the classification performance. Finally, simulation results on the radio machine learning (RadioML) 2018.01A dataset signify that the network's classification performance is superior to many classifiers in the literature.

Key words: convolutional neural network, deep neural network, long short-term memory, modulation classification, residual network

中图分类号:

TN911.7

Li Hui, Li Shanshan, Zou Borong, Chen Yannan. Modulation classification based on the collaboration of dual-channel CNN-LSTM and residual network[J]. The Journal of China Universities of Posts and Telecommunications, 2022, 29(1): 113-124.

参考文献

References

[1] WANG Y, LIU M, YANG J, et al. Data-driven deep learning for automatic modulation recognition in cognitive radios. IEEE Transactions on Vehicular Technology, 2019, 68(4): 4074 - 4077.

[2] GHASEMZADEH P, BANERJEE S, HEMPEL M, et al. A novel deep learning and polar transformation framework for an adaptive automatic modulation classification. IEEE Transactions on Vehicular Technology, 2020, 69(11): 13243 - 13258.

[3] ZHENG J P, LU Y F. Likelihood-based automatic modulation classification in OFDM with index modulation. IEEE Transactions on Vehicular Technology, 2018, 67(9): 8192 - 8204.

[4] HE Z M, PENG Y, ZHAO Y, et al. Deep learning-based automatic modulation recognition algorithm in non-cooperative communication systems. Proceedings of the 11th International Conference on Wireless Communications and Signal Processing (WCSP'19), 2019, Oct 23 - 25, Xi'an, China. Piscataway, NJ, USA: IEEE, 2019: 1 - 6.

[5] DOBRE O A, ABDI A, BAR-NESS Y, et al. Survey of automatic modulation classification techniques: Classical approaches and new trends. IET Communications, 2007, 1(2): 137 - 156.

[6] LEE J H, KIM K Y, SHIN Y. Feature image-based automatic modulation classification method using CNN algorithm. Proceedings of the 2019 International Conference on Artificial Intelligence in Information and Communication (ICAIIC'19), 2019, Feb 11 - 13, Okinawa, Japan. Piscataway, NJ, USA: IEEE, 2019: 1 - 4.

[7] YANG C, HE Z M, PENG Y, et al. Deep learning aided method for automatic modulation recognition. IEEE Access, 2019, 7: 109063 - 109068.

[8] KUMAR Y, SHEORAN M, JAJOO G, et al. Automatic modulation classification based on constellation density using deep learning. IEEE Communications Letters, 2020, 24(6): 1275 - 1278.

[9] SWAMI A, SADLER B M. Hierarchical digital modulation classification using cumulants. IEEE Transactions on Communications, 2000, 48(3): 416 - 429.

[10] YAN X, FENG G Y, WU H C, et al. Innovative robust modulation classification using graph-based cyclic-spectrum analysis. IEEE Communications Letters, 2017, 21(1): 16 - 19.

[11] HASSANPOUR S, PEZESHK A M, BEHNIA F. Automatic digital modulation recognition based on novel features and support vector machine. Proceedings of the 12th International Conference on Signal-Image Technology and Internet-Based Systems (SITIS'16), 2016, Nov 28 - Dec 1, Naples, Italy. Piscataway, NJ, USA: IEEE, 2016: 172 - 177.

[12] MOHAMMADI M, MOUSAVI S H, EFFATI S. Generalized variant support vector machine. IEEE Transactions on Systems, Man, and Cybernetics: Systems, 2021, 51(5): 2798 - 2809.

[13] PATIL S, KULKARNI U. Accuracy prediction for distributed decision tree using machine learning approach. Proceedings of the 3rd International Conference on Trends in Electronics and Informatics (ICOEI'19), 2019, Apr 23 - 25, Tirunelveli, India. Piscataway, NJ, USA: IEEE, 2019: 1365 - 1371.

[14] GAVANKAR S S, SAWARKAR S D. Eager decision tree. Proceedings of the 2nd International Conference for Convergence in Technology (I2CT'17), 2017, Apr 7 - 9, Mumbai, India. Piscataway, NJ, USA: IEEE, 2017: 837 - 840.

[15] YANG F Q, YANG L, WANG D, et al. Method of modulation recognition based on combination algorithm of K-means clustering and grading training SVM. China Communications, 2018, 15(12): 55 - 63.

[16] ASLAM M W, ZHU Z C, NANDI A K. Automatic modulation classification using combination of genetic programming and KNN. IEEE Transactions on Wireless Communications, 2012, 11(8): 2742 - 2750.

[17] O'SHEA T J, CORGAN J, CLANCY T C. Convolutional radio modulation recognition networks. Engineering Applications of Neural Networks: Proceedings of the 17th International Conference on Engineering Applications of Neural Networks (EANN'16), 2016, Sept 2 - 5, Aberdeen, UK. CCIS 629. Cham, Switzerland: Springer, 2016: 213 - 226.

[18] RAMJEE S, JU S T, YANG D Y, et al. Fast deep learning for automatic modulation classification. ArXiv: 1901. 05850, 2019.

[19] ZHANG Z F, LUO H, WANG C, et al. Automatic modulation classification using CNN-LSTM based dual-stream structure. IEEE Transactions on Vehicular Technology, 2020, 69(11): 13521 - 13531.

[20] SAINATH T N, VINYALS O, SENIOR A, et al. Convolutional, long short-term memory, fully connected deep neural networks. Proceedings of the 2015 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP'15), 2015, Apr 19 - 24, South Brisbane, Australia. Piscataway, NJ, USA: IEEE, 2015: 4580 - 4584.

[21] HE K M, ZHANG X Y, REN S Q, et al. Deep residual learning for image recognition. Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR'16), 2016, Jun 27 - 30, Las Vegas, NV, USA. Piscataway, NJ, USA: IEEE, 2016: 770 - 778.

[22] CHEN H T, GUO L, DONG C, et al. Automatic modulation classification using multi-scale convolutional neural network. Proceedings of the IEEE 31st Annual International Symposium on Personal, Indoor and Mobile Radio Communications (PIMRC'20), 2020, Aug 31 - Sept 3, London, UK. Piscataway, NJ, USA: IEEE, 2020: 1 - 6.

[23] WANG W B, TONG M, YU M. Blood glucose prediction with VMD and LSTM optimized by improved particle swarm optimization. IEEE Access, 2020, 8: 217908 - 217916.

[24] YU Y L, LIU F X. Effective neural network training with a new weighting mechanism-based optimization algorithm. IEEE Access, 2019, 7: 72403 - 72410.

[25] O'SHEA T J, ROY T, CLANCY T C. Over-the-air deep learning based radio signal classification. IEEE Journal of Selected Topics in Signal Processing, 2018, 12(1): 168 - 179.

[26] O'SHEA T J, WEST N. Radio machine learning dataset generation with GNU radio. Proceedings of the 6th GNU Radio Conference (GRCon'16), 2016, Sept 12 - 16, Boulder, CO, USA. 2016: 1 - 6.

[27] PENG S L, JIANG H Y, WANG H X, et al. Modulation classification based on signal constellation diagrams and deep learning. IEEE Transactions on Neural Networks and Learning Systems, 2019, 30(3): 718 - 727.

[28] ZENG Y, ZHANG M, HAN F, et al. Spectrum analysis and convolutional neural network for automatic modulation recognition. IEEE Wireless Communications Letters, 2019, 8(3): 929 - 932.

[29] ZHANG M, ZENG Y, HAN Z D, et al. Automatic modulation recognition using deep learning architectures. Proceedings of the IEEE 19th International Workshop on Signal Processing Advances in Wireless Communications (SPAWC'18), 2018, Jun 25 - 28, Kalamata, Greece. Piscataway, NJ, USA: IEEE, 2018: 1 - 5.

[30] TUNZE G B, HUYNH-THE T, LEE J M, et al. Sparsely connected CNN for efficient automatic modulation recognition. IEEE Transactions on Vehicular Technology, 2020, 69(12): 15557 - 15568.

[31] HUYNH-THE T, HUA C H, PHAM Q V, et al. MCNet: An efficient CNN architecture for robust automatic modulation classification. IEEE Communications Letters, 2020, 24(4): 811 - 815.

[32] KIM S H, KIM J W, DOAN V S, et al. Lightweight deep learning model for automatic modulation classification in cognitive radio networks. IEEE Access, 2020, 8: 197532 - 197541.

[33] WEST N E, O'SHEA T. Deep architectures for modulation recognition. Proceedings of the 2017 IEEE International Symposium on Dynamic Spectrum Access Networks (DySPAN'17), 2017, Mar 6 - 9, Baltimore, MD, USA. Piscataway, NJ, USA: IEEE, 2017: 1 - 6.

[34] MENG F, CHEN P, WU L N, et al. Automatic modulation classification: A deep learning enabled approach. IEEE Transactions on Vehicular Technology, 2018, 67(11): 10760 - 10772.

Modulation classification based on the collaboration of dual-channel CNN-LSTM and residual network

Modulation classification based on the collaboration of dual-channel CNN-LSTM and residual network

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 12

编辑推荐

Metrics

本文评价

[1]	Ma Shexiang, Mei Xiaobing. Side-lobe constrained beamforming under virtual expansion of L-shaped array[J]. 中国邮电高校学报(英文版), 2021, 28(2): 79-88.
[2]	Gao Miao, Yu Zhiguo, Li Qingqing, Wei Jinghe, Gu Xiaofeng. Improved ECG signal compression method for E-healthcare[J]. 中国邮电高校学报(英文版), 2019, 26(4): 62-69.
[3]	张凯尹长川 Xu Guixian 刘丹谱. Robust hybrid beamforming design for millimeter-wave massive MIMO systems[J]. 中国邮电高校学报(英文版), 2019, 26(3): 44-49.
[4]	Yang Jinsheng, Xiang Yang, Chen Weigang, Dong Yangyang. Computationally efficient 2-D DOA estimation for non-uniform two-L-shaped array[J]. 中国邮电高校学报(英文版), 2018, 25(6): 81-89.
[5]	Wang Qi, Wang Shigang, Jia Bowen, Du Hailong. Curvelet transform and contrast adaptive clip histogram equalization-based image defogging algorithm[J]. 中国邮电高校学报(英文版), 2018, 25(2): 96-104.
[6]	Li Ying, Wang Jian, Song Zhanjie. Multivariate spectral estimation based on THREE[J]. Acta Metallurgica Sinica(English letters), 2015, 22(4): 26-32.
[7]	王波赵彦平刘娟娟. Sparse recovery method for far-field and near-field sources localization using oblique projection[J]. Acta Metallurgica Sinica(English letters), 2013, 20(3): 90-96.
[8]	耿烜刘锋何晨. Outage analysis of ANC in the FDD two-way fading channel with channel estimation error[J]. Acta Metallurgica Sinica(English letters), 2013, 20(2): 12-18.
[9]	耿烜何晨孙作雷刘坤. Tomlinson-Harashima precoding minimizing total MSE with transmitter spatial correlation[J]. Acta Metallurgica Sinica(English letters), 2012, 19(6): 19-24.
[10]	赵发勇;酆广增. Joint spatial-frequency blind multiuser detection based on LCCMA[J]. Acta Metallurgica Sinica(English letters), 2007, 14(2): 19-22.
[11]	齐轶; 杜世培; 黄永贵; 齐洪钢. Pre-processing for video coding with rate-distortion optimization decision[J]. Acta Metallurgica Sinica(English letters), 2006, 13(2): 79-83.
[12]	王友国;吴乐南 . Noise-Assisted Signal Reception in Threshold Neuron[J]. Acta Metallurgica Sinica(English letters), 2006, 13(1): 79-83.