Optimizing Automatic Modulation Classification through Gaussian-Regularized Hybrid CNN-LSTM Architecture

Document Type : Research Paper

Authors

Electrical Engineering Department, Faculty of Engineering, Sohag University, Sohag, Egypt.

Abstract

This paper presents an innovative deep-learning model for Automatic Modulation Classification (AMC) in wireless communication systems. The proposed architecture integrates Convolutional Neural Network (CNN) and Long Short-Term Memory (LSTM) networks, augmented by a Gaussian noise layer to mitigate overfitting. The integration of both networks seeks to enhance classification accuracy and performance by leveraging the unique capabilities of CNNs and LSTMs in capturing spatial and temporal features, respectively. The model is expected to distinguish between eight digital and two analog modulation modes. Experimental evaluation on the RadioML2016.10b dataset demonstrates a peak recognition accuracy of 93.2% at 18 dB SNR. Comparative analyses validate the superior performance of the proposed architecture. The Gaussian noise layer contributes significantly to a 3% performance improvement at 18 dB SNR. The model achieves recognition accuracy exceeding 96% for most modulation modes, highlighting its robustness. Finally, computational complexity analysis underscores the efficiency of the proposed architecture, reinforcing its practical viability.

Keywords

Main Subjects

References

[1] J.C. Clement, N. Indira, P. Vijayakumar, R. Nandakumar, "Deep learning-based modulation  
     Classification for 5G and beyond wireless systems," Peer-- Netw. Appl., vol. 14, pp. 319–
     332, 2021.
[2] L. Hong, K.C. Ho, "Antenna array likelihood modulation classifier for BPSK and QPSK
      Signals," in MILCOM 2002 Proc., 2002, pp. 647–651 vol.1.
[3] B.F. Beidas, C.L. Weber, "Asynchronous classification of MFSK signals using the higher order
      Correlation domain," IEEE Trans. Commun., vol. 46, pp. 480–493, 1998.
[4] J.A. Sills, "Maximum-likelihood modulation classification for PSK/QAM," in MILCOM 1999
      IEEE Mil. Commun. Conf. Proc. Cat No99CH36341, 1999, pp. 217–220 vol.1.
[5] K. Kim, A. Polydoros, "Digital modulation classification: the BPSK versus QPSK case,"
      MILCOM 88 21st Century Mil. Commun. - Whats Possible Conf. Rec. Mil. Commun. Conf.,
      1988.
[6] Z. Sun, S. Wang, X. Chen, "Feature-Based Digital Modulation Recognition Using Compressive
      Sampling," Mob. Inf. Syst., vol. 2016, 2016, e9754162.
[7] M.V. Subbarao, P. Samundiswary, "Automatic Modulation Recognition in Cognitive Radio
      Receivers using Multi-Order Cumulants and Decision Trees," vol. 7, 2018, p. 10.
[8] A. Puengnim, T. Robert, N. Thomas, J. Vidal, "Hidden Markov Models for digital modulation
      Classification in unknown ISI channels," in 2007 15th Eur. Signal Process. Conf., 2007, pp1882–     1886.
[9] W. Dai, Y. Wang, J. Wang, "Joint power estimation and modulation classification using second-
      And higher statistics," in 2002 IEEE Wirel. Commun. Netw. Conf. Rec. WCNC 2002 Cat
      No02TH8609, 2002, pp. 155–158 vol.1.
[10] Khan et al., "3D convolutional neural networks based automatic modulation classification in
       The presence of channel noise," IET Communications, 2022. [Online]. Available: 
        https://ietresearch.onlinelibrary.wiley.com/doi/full/10.1049/cmu2.12269 [Accessed: January 2,
        2023].
[11] A. Shah, A. Miry, T. Salman, "Automatic modulation classification based deep learning with
       Mixed feature," Int. J. Electr. Comput. Eng., vol. 13, pp. 1647–1653, 2023.
[12] N. Wang, Y. Liu, L. Ma, Y. Yang, H. Wang, "Automatic Modulation Classification Based on
        CNN and Multiple Kernel Maximum Mean Discrepancy," Electronics, vol. 12, 2023, p. 66.
[13] F. Meng, P. Chen, L. Wu, X. Wang, "Automatic Modulation Classification: A Deep Learning
        Enabled Approach," IEEE Trans. Veh. Technol., vol. 67, pp. 10760–10772, 2018.
[14] M. Lotfy, M.H. Essai, H.A. Atallah, "Automatic Modulation Classification: Convolutional
        Deep Learning Neural Networks Approaches," SVU-Int. J. Eng. Sci. Appl., vol. 4, pp. 48–54,
        2023.
[15] T. Huynh-The, Q.-V. Pham, T.-V. Nguyen, T.T. Nguyen, D.B. da Costa, D.-S. Kim, "RanNet:
        Learning   Residual-Attention Structure in CNNs for Automatic Modulation Classification,"
        IEEE Wirel. Commun. Lett. vol. 11, pp. 1243–1247, 2022.
[16] T. O’Shea, J. Corgan, T. Clancy, "Convolutional Radio Modulation Recognition Networks," in
        2016. [17] T.J. O’Shea, T. Roy, T.C. Clancy, "Over-the-Air Deep Learning Based Radio Sign       Classification," IEEE J. Sel. Top. Signal Process. vol. 12, pp. 168–179, 2018.
[18] X. Hao, Y. Luo, Q. Ye, Q. He, G. Yang, C.-C. Chen, "Automatic Modulation Recognition
       Method Based on Hybrid Model of Convolutional Neural Networks and Gated Recurrent
       Units," Sens. Mater. vol. 33, pp. 4229–4243, 2021.
[19] Y. Wang, M. Liu, J. Yang, G. GUI, "Data-Driven Deep Learning for Automatic Modulation
        Recognition in Cognitive Radios," IEEE Trans. Veh. Technol., vol. 68, pp. 4074–4077, 2019.
[20] D. Inc, "RF Datasets for Machine Learning," DeepSig. [Online]. Available:
        https://www.deepsig.ai/datasets.
[21] "Google Colaboratory," [Online]. Available: https://colab.research.google.com/.
[22] M.M. Elsagheer, S.M. Ramzy, "A hybrid model for automatic modulation classification based
        On residual neural networks and long short-term memory," Alex. Eng. J., vol. 67, pp. 117–128,
        2023.
[23] J.N. Njoku, M.E. Morocho-Cayamcela, W. Lim, "CGDNet: Efficient Hybrid Deep Learning
        Model for Robust Automatic Modulation Recognition," IEEE Netw. Lett. vol. 3, pp. 47–51,
       2021.
[24] J. Xu, C. Luo, G. Parr, Y. Luo, "A Spatiotemporal Multi-Channel Learning Framework for
       Automatic Modulation Recognition," IEEE Wirel. Commun. Lett. pp. 1–1, 2020.
[25] Y. Lin, Y. Tu, Z. Dou, "An Improved Neural Network Pruning Technology for Automatic
        Modulation   Classification in Edge Devices," IEEE Trans. Veh. Technol., vol. 69, pp. 5703–  
       5706, 2020.

Files

Share

How to cite

Statistics