DCNN-Based Multipath Channel Prediction Model in Mobile Communication Environment

Abstract

Nowadays, the next generation mobile communication systems (Beyond 5G / 6G) have been actively investigated all over the world, and we have proposed the radio propagation loss prediction model with deep convolutional neural networks (DCNN) to improve the prediction accuracy. In our model, the features which needed for propagation loss prediction are automatically extracted from image maps. So, it is expected that our proposed model can also predict multipath channels. In this paper, we extend our proposed DCNN-based model to multipath channel prediction and clarify its performance.