Sigma delta quantization for compressed sensing

C. Sinan Gunturk, Mark Lammers, Alex Powell, Rayan Saab, Özgür Yilmaz

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Recent results make it clear that the compressed sensing paradigm can be used effectively for dimension reduction. On the other hand, the literature on quantization of compressed sensing measurements is relatively sparse, and mainly focuses on pulse-code-modulation (PCM) type schemes where each measurement is quantized independently using a uniform quantizer, say, of step size δ. The robust recovery result of Candès et ale and Donoho guarantees that in this case, under certain generic conditions on the measurement matrix such as the restricted isometry property, ℓ1 recovery yields an approximation of the original sparse signal with an accuracy of O (δ). In this paper, we propose sigma-delta quantization as a more effective alternative to PCM in the compressed sensing setting. We show that if we use an rth order sigma-delta scheme to quantize m compressed sensing measurements of a k-sparse signal in RN, the reconstruction accuracy can be improved by a factor of (m/k)(r-1/2)α for any 0 < α < 1 if m ≳γ k(log N)1/(1-α) (with high probability on the measurement matrix). This is achieved by employing an alternative recovery method via γth-order Sobolev dual frames.

Original languageEnglish (US)
Title of host publication2010 44th Annual Conference on Information Sciences and Systems, CISS 2010
DOIs
StatePublished - 2010
Event44th Annual Conference on Information Sciences and Systems, CISS 2010 - Princeton, NJ, United States
Duration: Mar 17 2010Mar 19 2010

Other

Other44th Annual Conference on Information Sciences and Systems, CISS 2010
CountryUnited States
CityPrinceton, NJ
Period3/17/103/19/10

Fingerprint

Quantization (signal)
Compressed sensing
Pulse code modulation
Recovery
Quantization

ASJC Scopus subject areas

  • Information Systems
  • Information Systems and Management

Cite this

Gunturk, C. S., Lammers, M., Powell, A., Saab, R., & Yilmaz, Ö. (2010). Sigma delta quantization for compressed sensing. In 2010 44th Annual Conference on Information Sciences and Systems, CISS 2010 [5464825] https://doi.org/10.1109/CISS.2010.5464825

Sigma delta quantization for compressed sensing. / Gunturk, C. Sinan; Lammers, Mark; Powell, Alex; Saab, Rayan; Yilmaz, Özgür.

2010 44th Annual Conference on Information Sciences and Systems, CISS 2010. 2010. 5464825.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Gunturk, CS, Lammers, M, Powell, A, Saab, R & Yilmaz, Ö 2010, Sigma delta quantization for compressed sensing. in 2010 44th Annual Conference on Information Sciences and Systems, CISS 2010., 5464825, 44th Annual Conference on Information Sciences and Systems, CISS 2010, Princeton, NJ, United States, 3/17/10. https://doi.org/10.1109/CISS.2010.5464825
Gunturk CS, Lammers M, Powell A, Saab R, Yilmaz Ö. Sigma delta quantization for compressed sensing. In 2010 44th Annual Conference on Information Sciences and Systems, CISS 2010. 2010. 5464825 https://doi.org/10.1109/CISS.2010.5464825
Gunturk, C. Sinan ; Lammers, Mark ; Powell, Alex ; Saab, Rayan ; Yilmaz, Özgür. / Sigma delta quantization for compressed sensing. 2010 44th Annual Conference on Information Sciences and Systems, CISS 2010. 2010.
@inproceedings{8db52662f48645908d2552dddea0595e,
title = "Sigma delta quantization for compressed sensing",
abstract = "Recent results make it clear that the compressed sensing paradigm can be used effectively for dimension reduction. On the other hand, the literature on quantization of compressed sensing measurements is relatively sparse, and mainly focuses on pulse-code-modulation (PCM) type schemes where each measurement is quantized independently using a uniform quantizer, say, of step size δ. The robust recovery result of Cand{\`e}s et ale and Donoho guarantees that in this case, under certain generic conditions on the measurement matrix such as the restricted isometry property, ℓ1 recovery yields an approximation of the original sparse signal with an accuracy of O (δ). In this paper, we propose sigma-delta quantization as a more effective alternative to PCM in the compressed sensing setting. We show that if we use an rth order sigma-delta scheme to quantize m compressed sensing measurements of a k-sparse signal in RN, the reconstruction accuracy can be improved by a factor of (m/k)(r-1/2)α for any 0 < α < 1 if m ≳γ k(log N)1/(1-α) (with high probability on the measurement matrix). This is achieved by employing an alternative recovery method via γth-order Sobolev dual frames.",
author = "Gunturk, {C. Sinan} and Mark Lammers and Alex Powell and Rayan Saab and {\"O}zg{\"u}r Yilmaz",
year = "2010",
doi = "10.1109/CISS.2010.5464825",
language = "English (US)",
isbn = "9781424474172",
booktitle = "2010 44th Annual Conference on Information Sciences and Systems, CISS 2010",

}

TY - GEN

T1 - Sigma delta quantization for compressed sensing

AU - Gunturk, C. Sinan

AU - Lammers, Mark

AU - Powell, Alex

AU - Saab, Rayan

AU - Yilmaz, Özgür

PY - 2010

Y1 - 2010

N2 - Recent results make it clear that the compressed sensing paradigm can be used effectively for dimension reduction. On the other hand, the literature on quantization of compressed sensing measurements is relatively sparse, and mainly focuses on pulse-code-modulation (PCM) type schemes where each measurement is quantized independently using a uniform quantizer, say, of step size δ. The robust recovery result of Candès et ale and Donoho guarantees that in this case, under certain generic conditions on the measurement matrix such as the restricted isometry property, ℓ1 recovery yields an approximation of the original sparse signal with an accuracy of O (δ). In this paper, we propose sigma-delta quantization as a more effective alternative to PCM in the compressed sensing setting. We show that if we use an rth order sigma-delta scheme to quantize m compressed sensing measurements of a k-sparse signal in RN, the reconstruction accuracy can be improved by a factor of (m/k)(r-1/2)α for any 0 < α < 1 if m ≳γ k(log N)1/(1-α) (with high probability on the measurement matrix). This is achieved by employing an alternative recovery method via γth-order Sobolev dual frames.

AB - Recent results make it clear that the compressed sensing paradigm can be used effectively for dimension reduction. On the other hand, the literature on quantization of compressed sensing measurements is relatively sparse, and mainly focuses on pulse-code-modulation (PCM) type schemes where each measurement is quantized independently using a uniform quantizer, say, of step size δ. The robust recovery result of Candès et ale and Donoho guarantees that in this case, under certain generic conditions on the measurement matrix such as the restricted isometry property, ℓ1 recovery yields an approximation of the original sparse signal with an accuracy of O (δ). In this paper, we propose sigma-delta quantization as a more effective alternative to PCM in the compressed sensing setting. We show that if we use an rth order sigma-delta scheme to quantize m compressed sensing measurements of a k-sparse signal in RN, the reconstruction accuracy can be improved by a factor of (m/k)(r-1/2)α for any 0 < α < 1 if m ≳γ k(log N)1/(1-α) (with high probability on the measurement matrix). This is achieved by employing an alternative recovery method via γth-order Sobolev dual frames.

UR - http://www.scopus.com/inward/record.url?scp=77953706903&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77953706903&partnerID=8YFLogxK

U2 - 10.1109/CISS.2010.5464825

DO - 10.1109/CISS.2010.5464825

M3 - Conference contribution

SN - 9781424474172

BT - 2010 44th Annual Conference on Information Sciences and Systems, CISS 2010

ER -