A Study of Automatic Modulation Classification for Rayleigh Fading Channel
Naman Batra, Brahmjit Singh · 2021 IEEE Bombay Section Signature Conference (IBSSC) · 2021
The automatic modulation recognition is an important research problem with potential applications in the defense and aerospace applications. In this paper, we present a study of automatic modulation classification approach. The scheme presented is simple to implement with low complexity. It is based on fourth-order cumulants. We aim to identify four types of digital modulation techniques. These include BPSK, PAM-4, 8-PSK, QAM-4. A data set of Rayleigh fading channel along with the confusion matrix is generated of size 4x4 and the overall accuracy of the identification problem is computed. The accuracy is measured for varied settings of SNR ranging from -10dB to 20dB. The data on the signal and noise is also generated characterizing the real part and imaginary part for the Rayleigh channel modeling. The generated dataset may be used for training of deep learning model for automatic modulation detection.