부산대학교 도서관

Automatic modulation classification (AMC) is one of the important tasks in cognitive radios and spectrum surveillance. This paper proposes a deep learning (DL)-based AMC with prediction combination method for orthogonal frequency division multiplexing (OFDM) systems. We first predict the subcarrier modulation scheme of the transmitted OFDM signal with different DL models and then combine those predictions using the proposed prediction combination method to make the final decision. Through computer simulations, we show that the classification performance can be improved by the proposed prediction combination method in terms of classification accuracy.