Journal on Communications ›› 2021, Vol. 42 ›› Issue (8): 61-69.doi: 10.11959/j.issn.1000-436x.2021128

• Papers • Previous Articles     Next Articles

Deep learning for compressed sensing based sparse channel estimation in FDD massive MIMO systems

Yuan HUANG1, Yigang HE1,2, Yuting WU1, Tongtong CHENG1, Yongbo SUI1, Shuguang NING1   

  1. 1 The School of Electrical Engineering and Automation, Hefei University of Technology, Hefei 230009, China
    2 The School of Electrical Engineering, Wuhan University, Wuhan 430072, China
  • Revised:2021-05-08 Online:2021-08-25 Published:2021-08-01
  • Supported by:
    The National Key Research and Development Program of China(2016YFF0102200);The National Natural Science Foundation of China(51577046);The National Natural Science Foundation of China(51637004)

Abstract:

For FDD massive multi-input multi-output (MIMO) downlink system, a novel deep learning method for compressed sensing based sparse channel estimation was proposed, which was called convolutional compressed sensing network (ConCSNet).In the ConCSNet, the convolutional neural network was utilized to solve the inverse transformation process from measurement vector y to signal h and solve the underdetermined optimization problem through data-driven method without sparsity.Simulation results show that the algorithm can recover the channel state information in massive MIMO Systems with unknown sparsity more quickly and accurately.

Key words: wireless communication, FDD massive MIMO system, sparse channel estimation, deep learning

CLC Number: 

No Suggested Reading articles found!