RT-DETR-Based Wideband Signal Detection and Modulation Classification

Minghao Cao; Peng Chu; Pengfei Ma; Bo Fang

doi:10.54097/1b1a7k36

Authors

Minghao Cao
Peng Chu
Pengfei Ma
Bo Fang

DOI:

https://doi.org/10.54097/1b1a7k36

Keywords:

Feature Extraction Networks, Signal Modulation Recognition, Deep Learning

Abstract

In To address the problems of high computational complexity, low accuracy, and the cumbersome manual feature extraction process in traditional machine learning methods for communication signal modulation recognition, this study proposes a deep learning-based end-to-end recognition model. Built upon the Transformer architecture using the RT-DETR framework, the model directly identifies modulation types from sampled communication signals. It features high recognition accuracy, strong generalization ability, robustness to noise, and a streamlined processing pipeline. Extensive experiments validate the model’s effectiveness, demonstrating its superior performance in automatic feature extraction and modulation classification compared to traditional approaches.

Downloads

Download data is not yet available.

References

[1] Zell O, Pålsson J, Hernandez-Diaz K, et al. Image-Based Fire Detection in Industrial Environments with YOLOv4[J]. arXiv preprint arXiv:2212.04786, 2022.

[2] Dobre, O. A., et al. (2007). Survey of automatic modulation classification techniques: Classical approaches and new trends. IEEE Communications Surveys & Tutorials.

[3] O'Shea, T. J., & Hoydis, J. (2017). An introduction to deep learning for the physical layer. IEEE Transactions on Cognitive Communications and Networking.

[4] Kulin, M., et al. (2018). End-to-end learning from spectrum data IEEE Access.

[5] Zhang, F., et al. (2022). Deep learning based automatic modulation recognition: Models, datasets, and challenges.

[6] Li, Y., et al. (2022). EfficientFormerV2: Efficient and Accurate Lightweight Vision Transformers.

[7] Zhang, Y., et al. (2024). HistoFormer: Histogram Transformer for Lightweight Image Recognition. ECCV 2024.

[8] Kingma, D. P., & Ba, J. (2014). Adam: A Method for Stochastic Optimization.

[9] Amini, A., Shirani-Mehr, H., & Karbasi, A. Digital Modulation Classification: A Deep Learning Approach. IEEE Transactions on Communications, (2017) 65(11), 4658-4668.

[10] Hochreiter, S., & Schmidhuber, J. Long Short-Term Memory. Neural Computation, (1997) 9(8), 1735-1780.

[11] LeCun, Y., Bottou, L., Orr, G. B., & Müller, K.-R. (2012). Efficient BackProp. In G. Montavon, G. B. Orr, & K.-R. Müller (Eds.), Neural Networks: Tricks of the Trade (pp. 9-48). Springer.

[12] Shorten, C., & Khoshgoftaar, T. M. A survey on Image Data Augmentation for Deep Learning. Journal of Big Data, 6(60), (2019). 1-48.

[13] Prechelt, L. (1998). Early Stopping - But When? In G. B. Orr & K.-R. Müller (Eds.), Neural Networks: Tricks of the Trade (pp. 55-69). Springer.

[14] Dosovitskiy, A., Beyer, L., Kolesnikov, A., Weissenborn, D., et al. (2020). An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale. Proceedings of the International Conference on Learning Representations (ICLR).