Convolutional neural network and clustering-based codebook design method for massive MIMO systems

Convolutional neural network and clustering-based codebook design method for massive MIMO systems

Abstract In this paper, we propose a convolutional neural network (CNN) and clustering-based codebook design method. Specifically, we train two different CNNs, i.e., CNN1 and CNN2, to compress the channel state information (CSI) matrices into the channel vectors and recover the channel vectors back...

Full description

Bibliographic Details
Main Authors: Jing Xing, Die Hu
Format: Article
Language:English
Published: SpringerOpen 2022-06-01
Series:EURASIP Journal on Advances in Signal Processing
Subjects:
Convolutional neural network (CNN)
Codebook design
Clustering
Online Access:https://doi.org/10.1186/s13634-022-00879-y

Internet

https://doi.org/10.1186/s13634-022-00879-y

Similar Items