Time-domain channel estimation for wideband millimeter wave systems with hybrid architecture

Millimeter wave (mmWave) systems will likely employ large antennas at both the transmitter and receiver for directional beamforming. Hybrid analog/digital MIMO architectures have been proposed previously for leveraging both array gain and multiplexing gain, while reducing the power consumption in an...

Full description

Saved in:
Bibliographic Details
Main Authors: Venugopal, Kiran, Alkhateeb, Ahmed, Heath, Robert W., Gonzalez Prelcic, Nuria
Format: Conference Proceeding
Language:English
Subjects:
Channel estimation
compressed sensing
hybrid architecture
Millimeter wave communication
MIMO
Radio frequency
Receivers
Sparse matrices
Training
Transmitters
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Millimeter wave (mmWave) systems will likely employ large antennas at both the transmitter and receiver for directional beamforming. Hybrid analog/digital MIMO architectures have been proposed previously for leveraging both array gain and multiplexing gain, while reducing the power consumption in analog-to-digital converters. Channel knowledge is needed to design the hybrid precoders/combiners, which is difficult to obtain due to the large antenna arrays and the frequency selective nature of the channel. In this paper, we propose a sparse recovery based time-domain channel estimation technique for hybrid architecture based frequency selective mmWave systems. The proposed compressed sensing channel estimation algorithm is shown to provide good estimation error performance, while requiring small training overhead. The simulation results show that using multiple RF chains at the receiver and the transmitter further reduces the training overhead.
ISSN:2379-190X
DOI:10.1109/ICASSP.2017.7953407