A finite rate of innovation multichannel sampling hardware system for multi-pulse signals
Ning Fu, Liwen Sun, Guoxing Huang, Shuaile Du
Abstract
Multi-pulse signals are composed of finite pulse streams of arbitrary pulse shape. With the pulse shape known, a multichannel sampling scheme for multi-pulse signals can operate at the rate of innovation, which is much lower than the Nyquist rate. The sampling system is based on low-pass filters, oscillators and integrators. By now there is no hardware to practice the approach. In this paper, we design a hardware system and discover that the non-idealities of low-pass filters will lead to failing in signal reconstruction. We research how the low-pass filters affect the reconstruction and solve the problem by channel calibration. The experiments show that channel calibration compensates most of the errors induced by low-pass filters, and this approach can achieve better estimation of time-delays and amplitudes of multi-pulse signals with a known pulse shape.
BibTeX
@inproceedings{icassp2017_afiniterateofinn,
title = {A finite rate of innovation multichannel sampling hardware system for multi-pulse signals},
author = {Ning Fu and Liwen Sun and Guoxing Huang and Shuaile Du},
booktitle = {ICASSP 2017},
year = {2017}
}