Deep Learning-Based Modulation Identification for OFDM Systems
Geonho Song, Mingyu Jang, Dongweon Yoon · 2023
This paper deals with a deep learning (DL)-based automatic modulation classification (AMC) method for orthogonal frequency division multiplexing (OFDM) systems. Among DL methods for AMC, convolution neural network (CNN) has been widely studied to classify the modulation scheme used in the OFDM systems. Although conventional CNN has performed well in previous studies, its classification performance can be degraded when an additional modulation scheme is considered. In this paper, we investigate the CNN-based AMC for the OFDM systems to improve the classification performance by using a deeper CNN model with a residual connection. Through computer simulations, we show that the proposed model outperforms the conventional CNN model for various ranges of training signal-to-noise ratios in terms of classification accuracy.