电讯技术2024，Vol.64Issue(1)：36-42,7.DOI:10.20079/j.issn.1001-893x.220819005
深度学习辅助的5G OFDM系统的信道估计
Deep Learning Assisted Channel Estimation for 5G OFDM Systems
摘要
Abstract
The traditional channel estimation algorithm is difficult to meet the requirement of high speed and low delay in 5G system.For this problem,the authors propose a channel estimation method based on image restoration technology by considering the time-frequency response of communication channel as a two-dimensional image.First,parameters are set to generate a channel data information data set of physical downlink shared channel(PDSCH)based on 5G new radio(NR)standard,and the generated channel matrix is treated as a two-dimensional image.Then,an image restoration network based on convolutional neural network is constructed,and residual connection is incorporated to improve the performance of the network.Finally,the trained network model is used for channel estimation.The simulation results show that the performance of the proposed channel estimation algorithm is significantly improved compared with those of the Least Square(LS),Practical Channel Estimation(PCE)and Image-based Super-resolution ChannelNet network.关键词
5G/正交频分复用(OFDM)/信道估计/深度学习/卷积神经网络Key words
5G/channel estimation/orthogonal frequency division multiplexing(OFDM)/deep learning/convolutional neural network分类
信息技术与安全科学引用本文复制引用
王义元,常俊,卢中奎,余福慧,魏家齐..深度学习辅助的5G OFDM系统的信道估计[J].电讯技术,2024,64(1):36-42,7.基金项目
国家自然科学基金资助项目(61562090) （61562090）
