Deep Learning Enabled Algorithm for Automatic Modulation Classification
Brahmjit Singh, Shalu, Chandra Prakash · 2024
In this paper, we present a joint convolutional neural network and gated recurrent unit based deep leaning model for recognition of modulation schemes in blind RF environment. The raw I/Q samples from RadioML2016.10b dataset are used as the input data. The proposed model offers significantly improved accuracy of the modulation schemes at the low SNR region ranging from -2dB value to 4dB. Through numerical results, it is shown that the proposed model yields recognition accuracy upto 99% at 0dB SNR. The average accuracy achieved is 93.3%, which is a significant improvement in the recognition performance.