# Multichannel Sampling of Pulse Streams at the Rate of Innovation

**Authors:** Kfir Gedalyahu, Ronen Tur, Yonina C. Eldar  
**arXiv:** [1004.5070](https://arxiv.org/abs/1004.5070)  
**Date:** 2010-04-28  
**Summary page:** https://www.summarizepaper.com/en/paper/1004.5070v1/

## Key points

- Authors Gedalyahu, Tur, and Eldar focus on minimal-rate sampling schemes for streams of delayed and weighted versions of a known pulse shape
- Signals fall under the finite rate of innovation (FRI) models, with sampling rate determined by the rate of innovation
- Previous studies on sampling pulse streams either failed to achieve the rate of innovation or were limited to dirac pulse shapes
- Proposed novel multichannel framework designed for pulse streams with arbitrary shapes operates at the rate of innovation
- Approach involves modulating input signal with selected waveforms followed by a bank of integrators
- Research demonstrates recovery of pulse stream from minimal-rate samples using standard tools derived from spectral estimation in a stable manner even at high rates of innovation
- Practical implementation issues such as reducing hardware complexity and ensuring immunity to failure in sampling channels are addressed
- Resulting scheme offers flexibility and improved noise robustness compared to previous approaches in this field

## Summary

In their work titled "Multichannel Sampling of Pulse Streams at the Rate of Innovation," authors Kfir Gedalyahu, Ronen Tur, and Yonina C. Eldar delve into the realm of minimal-rate sampling schemes for streams of delayed and weighted versions of a known pulse shape. These signals fall under the finite rate of innovation (FRI) models, where the minimal sampling rate is determined by the number of degrees of freedom per unit time, known as the rate of innovation. The authors highlight that previous studies on sampling pulse streams either failed to achieve the rate of innovation or were limited to pulse shapes consisting solely of diracs. To address these limitations, Gedalyahu, Tur, and Eldar propose a novel multichannel framework designed specifically for pulse streams with arbitrary shapes that operates at the rate of innovation. Their approach involves modulating the input signal with a carefully selected set of waveforms, followed by a bank of integrators. Through their research, they demonstrate that it is possible to recover the pulse stream from the proposed minimal-rate samples using standard tools derived from spectral estimation in a stable manner even at high rates of innovation. Additionally, practical implementation issues such as reducing hardware complexity and ensuring immunity to failure in sampling channels are addressed in their work. Overall, their resulting scheme offers flexibility and improved noise robustness compared to previous approaches in this field. By providing a comprehensive solution for sampling parametric signals with varying pulse shapes at optimal rates, Gedalyahu, Tur, and Eldar's work contributes significantly to advancing signal processing techniques in communication systems and beyond.