Deep Learning-Based Massive MIMO CSI Feedback

Jialing Li, Zihe Gao, Jinxi Qian, Qi Zhang, Xiangjun Xin, Ying Tao, Qinghua Tian, Feng Tian, Dong Chen, Yufei Shen, Guixing Cao

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

14 Citations (Scopus)

Abstract

Massive multi-input and multi-output technology is a key technology for future 5G wireless communication. The channel feedback problem of massive mimo becomes more and more challenging as the size of the mimo channel matrix becomes larger. A supervised deep learning-based encoder-decoder scheme was proposed to improve recinstruction quality recovery channel state information.Compared with the traditional compression-based sensing algorithm, Residual Attention-Net can still maintain good performance when compression is low.

Original languageEnglish
Title of host publication2019 18th International Conference on Optical Communications and Networks, ICOCN 2019
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781728127644
DOIs
Publication statusPublished - Aug 2019
Externally publishedYes
Event18th International Conference on Optical Communications and Networks, ICOCN 2019 - Huangshan, China
Duration: 5 Aug 20198 Aug 2019

Publication series

Name2019 18th International Conference on Optical Communications and Networks, ICOCN 2019

Conference

Conference18th International Conference on Optical Communications and Networks, ICOCN 2019
Country/TerritoryChina
CityHuangshan
Period5/08/198/08/19

Keywords

  • attention model
  • compressed sensing
  • deep learning
  • massive MIMO
  • residual network

Fingerprint

Dive into the research topics of 'Deep Learning-Based Massive MIMO CSI Feedback'. Together they form a unique fingerprint.

Cite this