Channel Estimation Based on Echo State Networks in Wireless MIMO Systems

Echo state networks (ESNs) provide architecture and supervised learning principle for recurrent neural networks (RNNs). In this paper, we apply ESN to channel estimation in wireless Multiple-Input and Multiple-Output (MIMO). There is the multipath propagation environment between a transmitter and re...

Full description

Saved in:
Bibliographic Details
Published in2015 Fifth International Conference on Instrumentation and Measurement, Computer, Communication and Control (IMCCC) pp. 1541 - 1546
Main Authors Yongbo Liao, Yanhu Wang, Wenchang Li
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.09.2015
Subjects
channel coefficient
Channel estimation
ESN
MIMO
NRMS
Receivers
Recurrent neural networks
reservoir
Reservoirs
Training
Wireless communication
Online AccessGet full text

Cover

More Information
Summary:Echo state networks (ESNs) provide architecture and supervised learning principle for recurrent neural networks (RNNs). In this paper, we apply ESN to channel estimation in wireless Multiple-Input and Multiple-Output (MIMO). There is the multipath propagation environment between a transmitter and receiver. Thus the received signal undergoes phase shift, attenuation and time delay. In order to mitigate these random effects and decoding the transmitted signal at the reservoir, we present ESNs for learning nonlinear systems and realizing accurate channel estimation. We also design a teaching scheme to train the output weights of ESNs. The potential for engineering application is illustrated by channel estimation. Numerical results show that accuracy is improved by the number of reservoir units.
DOI:10.1109/IMCCC.2015.327