Resource Assignment Based on Deep Neural Networks for MUDiv/OFDMA

In recent communication systems, access techniques are important to achieve massive connection systems. Orthogonal frequency division multiple access (OFDMA) is a kind of access techniques, and the subcarrier assignment of each user with the multiuser diversity (MUDiv) gain is enhanced for the channel capacity. Before now, the subcarrier assignment has been achieved by the convolutional neural networks (CNN). However, it is not indicated except the learning result. Therefore, in this paper, we propose the subcarrier resource assignment based on the deep neural networks (DNN) for MUDiv/OFDMA systems.