AUTOMATIC MODULATION RECOGNITION USING DEEP LEARNING TECHNIQUES
DOI:
https://doi.org/10.24867/31BE14JankovKeywords:
Automatic Modulation Recognition, Cognitive Radio, Convolutional Neural Networks, Long Short-Term Memory, Deep LearningAbstract
This paper analyzes the performance of two machine learning models in modulation scheme recognition: Convolutional Neural Network (CNN) and Long Short-Term Memory (LSTM) network. Using the synthetic RadioML2016.10a dataset, the models were trained and tested on 11 different modulation types within an SNR range from -20 dB to +20 dB. The results show that both models achieve high accuracy, with the LSTM network demonstrating slightly better performance in modulation classification at higher SNR values. This analysis contributes to the development of advanced modulation recognition techniques that can enhance the efficiency of cognitive radio in complex spectral environments.
References
[1] Y. Wang, M. Liu, J. Yang, G. Gui, “Data-Driven Deep Learning for Automatic Modulation Recognition in Cognitive Radios”, IEEE Transactions on Vehicular Technology, vol. 68, no. 4, pp. 4074-4077, April 2019. Dostupno na: https://ieeexplore.ieee.org/document/8645696.
[2] Z. Ke, H. Vikalo, “Real-Time Radio Technology and Modulation Classification via an LSTM AutoEncoder”, Novembar 2020, Dostupno na: https://arxiv.org/abs/2011.08295.
[3] D. P. Kingma, J. Ba, “Adam: A Method for Stochastic Optimization”, Januar 2017. Dostupno na: https://arxiv.org/abs/1412.6980
[4] J. L. Ziegler, R. T. Arn, W. Chambers, “Modulation recognition with GNU radio, keras, and HackRF”, Mart 2017. Dostupno na: https://ieeexplore.ieee.org/document/7920747.
