电信科学 ›› 2021, Vol. 37 ›› Issue (1): 39-47.doi: 10.11959/j.issn.1000-0801.2021011

• 研究与开发 • 上一篇    下一篇

一种新型的基于深度学习的时变信道预测方法

张捷, 杨丽花, 王增浩, 呼博, 聂倩   

  1. 南京邮电大学江苏省无线通信重点实验室,江苏 南京 210003
  • 修回日期:2020-12-07 出版日期:2021-01-20 发布日期:2021-01-01
  • 作者简介:张捷(1996- ),女,南京邮电大学硕士生,主要研究方向为宽带移动通信。
    杨丽花(1984- ),女,南京邮电大学副教授,主要研究方向为移动无线通信、通信信号处理、多载波通信系统等。
    王增浩(1994- ),男,南京邮电大学硕士生,主要研究方向为移动通信。
    呼博(1996- ),男,南京邮电大学硕士生,主要研究方向为宽带移动通信。
    聂倩(1997- ),女,南京邮电大学硕士生,主要研究方向为移动通信。
  • 基金资助:
    江苏省科技厅自然科学基金资助项目(BK20191378);江苏省高等学校自然科学研究面上项目(18KJB510034);第11批中国博士后科学基金特别资助项目(2018T110530);国家自然科学基金资助项目(61771255)

A novel deep learning based time-varying channel prediction method

Jie ZHANG, Lihua YANG, Zenghao WANG, Bo HU, Qian NIE   

  1. Jiangsu Key Laboratory of Wireless Communication, Nanjing University of Posts and Telecommunications, Nanjing 210003, China
  • Revised:2020-12-07 Online:2021-01-20 Published:2021-01-01
  • Supported by:
    The National Science Foundation Program of Jiangsu Province(BK20191378);The National Science Research Project of Jiangsu Higher Education Institutions(18KJB510034);The National Natural Science Foundation of China(61771255)

摘要:

针对高速移动正交频分复用系统,提出了一种新型的基于深度学习的时变信道预测方法。为了避免网络参数随机初始化造成的影响,本文方法首先基于数据与导频信息获取较理想的信道估计,利用其对 BP神经网络进行预训练处理,以获取理想的网络初始参数;然后,基于预训练获取网络初始值,利用基于导频获取的信道估计对BP神经网络进行再次训练,以获取最终的信道预测网络模型;最后,本文方法基于该预测网络模型通过线上预测实现了时变信道的单时刻与多时刻预测。仿真结果表明,本文方法可以显著地提高时变信道预测精度,且具有较低的计算复杂度。

关键词: 高速移动, 正交频分复用, 深度学习, BP神经网络, 时变信道预测, 预训练

Abstract:

For high-speed mobile orthogonal frequency division multiplexing (OFDM) systems, a novel time-varying channel prediction method based on deep learning was proposed.To avoid the influence caused by random initialization of network parameters, the proposed method firstly obtains an ideal channel estimation based on data and pilot, and then pre-trains the back propagation (BP) neural network based on the channel estimation to obtain the ideal network initial parameters.Then, based on the initial network value obtained by pre-training, the proposed method uses the channel estimation based on pilot to train the BP neural network again.Finally, the proposed method realizes the single-time-and multi-time prediction of time-varying channels through on-line prediction.Simulation results show that the proposed method can significantly improve the prediction accuracy of time-varying channels and has a low computational complexity.

Key words: high-speed mobility, OFDM, deep learning, BP neural network, time-varying channel prediction, pre-training

中图分类号: 

No Suggested Reading articles found!