Deep Learning Based OFDM Channel Estimation Using Frequency-Time Division and Attention Mechanism
Abstract
In this paper, we propose a frequency-time division network (FreqTimeNet) to improve the performance of deep learning (DL) based OFDM channel estimation. This FreqTimeNet is designed based on the orthogonality between the frequency domain and the time domain. In FreqTimeNet, the input is processed by parallel frequency blocks and parallel time blocks sequentially. By introducing the attention mechanism using the SNR information, an attention based FreqTimeNet (AttenFreqTimeNet) is proposed. Using 3rd Generation Partnership Project (3GPP) channel models, the mean square error (MSE) performance of FreqTimeNet and AttenFreqTimeNet under different scenarios is evaluated. A method for constructing mixed training data is proposed, which could address the generalization problem in DL. It is observed that AttenFreqTimeNet outperforms FreqTimeNet, and FreqTimeNet outperforms other DL networks with reasonable complexity.
Cite
@article{arxiv.2107.07161,
title = {Deep Learning Based OFDM Channel Estimation Using Frequency-Time Division and Attention Mechanism},
author = {Ang Yang and Peng Sun and Tamrakar Rakesh and Bule Sun and Fei Qin},
journal= {arXiv preprint arXiv:2107.07161},
year = {2021}
}
Comments
2021 IEEE Globecom Workshops (GC Wkshps): Workshop on Towards Native-AI Wireless Networks