Optimized Hybrid CLDNN Architecture with Enhanced Temporal-Spatial Feature Extraction for Robust Automatic Modulation Classification in Cognitive Radio Networks

Daryan Pratama Alifi, Universitas Pendidikan Indonesia, Hane Yorda Dinata, Universitas Pendidikan Indonesia, Galura Muhammad Suranegara, Universitas Pendidikan Indonesia, Ichwan Nul Ichsan, Universitas Pendidikan Indonesia
Received: 2026-01-30
Accepted: 2026-03-10
Published: 2026-04-01

Abstract


Automatic Modulation Classification (AMC) is a pivotal technology for efficient spectrum management in future cognitive radio networks. While Deep Learning has advanced the field, standard Convolutional Neural Networks (CNN) often struggle to capture long-term temporal dependencies in signals affected by fading. This study proposes an Optimized Hybrid CLDNN architecture that integrates a "Wide-Kernel" CNN (k=7) for enhanced spatial feature extraction and a "High-Capacity" LSTM (100 units) for robust temporal modeling. Experimental validation using the RadioML 2016.10a dataset demonstrates that the proposed optimizations yield significant performance gains. Specifically, the model achieves a classification accuracy of 84.5% at 0 dB SNR, outperforming standard baselines in the critical transition regime. Furthermore, it reaches a peak accuracy of 92.4% at high SNR (+18 dB). A notable finding is the reduction of inter-class confusion between 16-QAM and 64-QAM, where the misclassification rate is suppressed to approximately 15%, confirming the architecture's effectiveness in resolving hierarchical modulation ambiguities in dynamic wireless environments.


Keywords


Automatic Modulation Classification; Cognitive Radio; Deep Learning; Hybrid CLDNN; QAM Disambiguation.

Full Text:

PDF

References


Liao, K., Zhao, Y., Gu, J., Zhang, Y., & Zhong, Y. (2021). Sequential Convolutional Recurrent Neural Networks for Fast Automatic Modulation Classification. IEEE Access, 9, 27182–27188. https://doi.org/10.1109/ACCESS.2021.3053427

Kassri, N., En‐Nouaary, A., & Bah, S. (2024). Efficient Squeeze-and-Excitation-Enhanced Deep Learning Method for Automatic Modulation Classification. International Journal of Advanced Computer Science and Applications, 15(6). https://doi.org/10.14569/ijacsa.2024.0150655

Ramjee, S., Ju, S., Yang, D., Liu, X., El Gamal, A., & Eldar, Y. C. (2019). Fast Deep Learning for Automatic Modulation Classification. https://arxiv.org/abs/1901.05850v1

He, C., Dong, P., Zhou, F., & Wu, Q. (2024). Collaborative Automatic Modulation Classification via Deep Edge Inference for Hierarchical Cognitive Radio Networks. https://doi.org/10.48550/arxiv.2409.07946

Padhya, D., Acharya, K., Dahal, B. K., & Kshatri, D. B. (2025). CNN-LSTM Hybrid Architecture for Over-the-Air Automatic Modulation Classification Using SDR. https://arxiv.org/abs/2511.21040v2

Emam, A., Shalaby, M., Aboelazm, M. A., Bakr, H. E.-D. A., & Mansour, H. A. A. (2020). A Comparative Study between CNN, LSTM, and CLDNN Models in The Context of Radio Modulation Classification. International Conference on Electrical Engineering, 190–195. https://doi.org/10.1109/ICEENG45378.2020.9171706

Elsagheer, M. M., Abd Elsayed, K., & Ramzy, safwat. (2024). Optimizing Automatic Modulation Classification through Gaussian-Regularized Hybrid CNN-LSTM Architecture. JES. Journal of Engineering Sciences. https://doi.org/10.21608/jesaun.2024.271102.1315

Wang, N., Liu, Y., Ma, L., Yang, Y., & Wang, H. (2021). Multidimensional CNN-LSTM Network for Automatic Modulation Classification. Electronics, 10(14), 1649. https://doi.org/10.3390/ELECTRONICS10141649

Murshed, R. U., Ashraf, Z. B., Hridhon, A. H., Munasinghe, K. S., Jamalipour, A., & Hossain, Md. F. (2022). A CNN-LSTM-Based Fusion Separation Deep Neural Network for 6G Ultra-Massive MIMO Hybrid Beamforming. IEEE Access, 11, 38614–38630. https://doi.org/10.1109/ACCESS.2023.3266355

Liu, B., Zheng, Q., Wei, H., Zhao, J., Yu, H., Zhou, Y., Chao, F., & Ji, R. (2024). Deep hybrid transformer network for robust modulation classification in wireless communications. Knowledge Based Systems. https://doi.org/10.1016/j.knosys.2024.112191

Li, J. (2023). An improved CNN‐LSTM network for modulation identification relying on periodic features of signal. Iet Communications. https://doi.org/10.1049/cmu2.12682

Kumar, A., & Satija, U. (2023). Residual stack-aided hybrid CNN-LSTM-based automatic modulation classification for orthogonal time-frequency space system. https://ieeexplore.ieee.org/abstract/document/10298111/

Gupta, A., & Fernando, X. (2022). Automatic Modulation Classification for Cognitive Radio Systems using CNN with Probabilistic Attention Mechanism. IEEE Vehicular Technology Conference, 1–6. https://doi.org/10.1109/VTC2022-Spring54318.2022.9860557

Häring, L. (2022). Deep Learning-Assisted Automatic Modulation Classification in Adaptive FBMC/OQAM Systems. IEEE International Symposium on Personal, Indoor and Mobile Radio Communications, 994–999. https://doi.org/10.1109/PIMRC54779.2022.9977504

Mousa, N. A., & Sadkhan, S. B. (2021). Identification of digitally modulated signal used in cognitive radio network -A survey. International Conference on Information Technology. https://doi.org/10.1109/BICITS51482.2021.9509927

Xie, X., Yang, G., Jiang, M., Ye, Q., & Yang, C.-F. (2021). A Kind of Wireless Modulation Recognition Method Based on DenseNet and BLSTM. IEEE Access, 9, 125706–125713. https://doi.org/10.1109/ACCESS.2021.3111406

Kim, S.-H., Kim, J.-W., Doan, V.-S., & Kim, D.-S. (2020). Lightweight Deep Learning Model for Automatic Modulation Classification in Cognitive Radio Networks. IEEE Access, 8, 197532–197541. https://doi.org/10.1109/ACCESS.2020.3033989




DOI: https://doi.org/10.30596/jcositte.v7i1.28935

Refbacks

  • There are currently no refbacks.