Abstract

Bayesian compressive sensing (BCS) helps address ill-posed signal recovery problems using the Bayesian estimation framework. Gibbs sampling is a technique used in Bayesian estimation that iteratively draws samples from conditional posterior distributions. However, Gibbs sampling is inherently sequential and existing parallel implementations focus on reducing the communication between computing units at the cost of increase in recovery error. In this work, we propose a two-stage parallel coefficient update scheme for wavelet-based Bayesian compressive sensing, where the first stage approximates the real distributions of the wavelet coefficients and the second stage computes the final estimate of the coefficients. While in the first stage the parallel computing units share information with each other, in the second stage, the parallel units work independently. We propose a new coefficient update scheme that updates coefficients in both stages based on data generated a few rounds ago. Such a scheme helps in relaxing the timing constraints for communication in the first stage and computations in the second stage. We design the corresponding parallel architecture and synthesize it in 7 nm technology node. We show that in a system with 8 computing units, our method helps reduce the execution time by 17.4× compared to a sequential implementation without any increase in the signal recovery error.

Original languageEnglish (US)
Title of host publicationProceedings of the IEEE Workshop on Signal Processing Systems, SiPS 2018
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages140-145
Number of pages6
ISBN (Electronic)9781538663189
DOIs
StatePublished - Dec 31 2018
Event2018 IEEE Workshop on Signal Processing Systems, SiPS 2018 - Cape Town, South Africa
Duration: Oct 21 2018Oct 24 2018

Publication series

NameIEEE Workshop on Signal Processing Systems, SiPS: Design and Implementation
Volume2018-October
ISSN (Print)1520-6130

Conference

Conference2018 IEEE Workshop on Signal Processing Systems, SiPS 2018
CountrySouth Africa
CityCape Town
Period10/21/1810/24/18

Fingerprint

Compressive Sensing
Gibbs Sampling
Wavelets
Sampling
Error Recovery
Unit
Parallel architectures
Update
Communication
Bayesian Estimation
Coefficient
Parallel processing systems
Recovery
Computing
Parallel Architectures
Wavelet Coefficients
Parallel Computing
Conditional Distribution
Parallel Implementation
Posterior distribution

Keywords

  • Bayesian compressive sensing
  • Gibbs sampling
  • parallel implementation

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Signal Processing
  • Applied Mathematics
  • Hardware and Architecture

Cite this

Zhou, J., & Chakrabarti, C. (2018). Parallel Wavelet-based Bayesian Compressive Sensing based on Gibbs Sampling. In Proceedings of the IEEE Workshop on Signal Processing Systems, SiPS 2018 (pp. 140-145). [8598375] (IEEE Workshop on Signal Processing Systems, SiPS: Design and Implementation; Vol. 2018-October). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/SiPS.2018.8598375

Parallel Wavelet-based Bayesian Compressive Sensing based on Gibbs Sampling. / Zhou, Jian; Chakrabarti, Chaitali.

Proceedings of the IEEE Workshop on Signal Processing Systems, SiPS 2018. Institute of Electrical and Electronics Engineers Inc., 2018. p. 140-145 8598375 (IEEE Workshop on Signal Processing Systems, SiPS: Design and Implementation; Vol. 2018-October).

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

Zhou, J & Chakrabarti, C 2018, Parallel Wavelet-based Bayesian Compressive Sensing based on Gibbs Sampling. in Proceedings of the IEEE Workshop on Signal Processing Systems, SiPS 2018., 8598375, IEEE Workshop on Signal Processing Systems, SiPS: Design and Implementation, vol. 2018-October, Institute of Electrical and Electronics Engineers Inc., pp. 140-145, 2018 IEEE Workshop on Signal Processing Systems, SiPS 2018, Cape Town, South Africa, 10/21/18. https://doi.org/10.1109/SiPS.2018.8598375
Zhou J, Chakrabarti C. Parallel Wavelet-based Bayesian Compressive Sensing based on Gibbs Sampling. In Proceedings of the IEEE Workshop on Signal Processing Systems, SiPS 2018. Institute of Electrical and Electronics Engineers Inc. 2018. p. 140-145. 8598375. (IEEE Workshop on Signal Processing Systems, SiPS: Design and Implementation). https://doi.org/10.1109/SiPS.2018.8598375
Zhou, Jian ; Chakrabarti, Chaitali. / Parallel Wavelet-based Bayesian Compressive Sensing based on Gibbs Sampling. Proceedings of the IEEE Workshop on Signal Processing Systems, SiPS 2018. Institute of Electrical and Electronics Engineers Inc., 2018. pp. 140-145 (IEEE Workshop on Signal Processing Systems, SiPS: Design and Implementation).
@inproceedings{8a552c636f3a4d31ac381c697a6bafa2,
title = "Parallel Wavelet-based Bayesian Compressive Sensing based on Gibbs Sampling",
abstract = "Bayesian compressive sensing (BCS) helps address ill-posed signal recovery problems using the Bayesian estimation framework. Gibbs sampling is a technique used in Bayesian estimation that iteratively draws samples from conditional posterior distributions. However, Gibbs sampling is inherently sequential and existing parallel implementations focus on reducing the communication between computing units at the cost of increase in recovery error. In this work, we propose a two-stage parallel coefficient update scheme for wavelet-based Bayesian compressive sensing, where the first stage approximates the real distributions of the wavelet coefficients and the second stage computes the final estimate of the coefficients. While in the first stage the parallel computing units share information with each other, in the second stage, the parallel units work independently. We propose a new coefficient update scheme that updates coefficients in both stages based on data generated a few rounds ago. Such a scheme helps in relaxing the timing constraints for communication in the first stage and computations in the second stage. We design the corresponding parallel architecture and synthesize it in 7 nm technology node. We show that in a system with 8 computing units, our method helps reduce the execution time by 17.4× compared to a sequential implementation without any increase in the signal recovery error.",
keywords = "Bayesian compressive sensing, Gibbs sampling, parallel implementation",
author = "Jian Zhou and Chaitali Chakrabarti",
year = "2018",
month = "12",
day = "31",
doi = "10.1109/SiPS.2018.8598375",
language = "English (US)",
series = "IEEE Workshop on Signal Processing Systems, SiPS: Design and Implementation",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
pages = "140--145",
booktitle = "Proceedings of the IEEE Workshop on Signal Processing Systems, SiPS 2018",

}

TY - GEN

T1 - Parallel Wavelet-based Bayesian Compressive Sensing based on Gibbs Sampling

AU - Zhou, Jian

AU - Chakrabarti, Chaitali

PY - 2018/12/31

Y1 - 2018/12/31

N2 - Bayesian compressive sensing (BCS) helps address ill-posed signal recovery problems using the Bayesian estimation framework. Gibbs sampling is a technique used in Bayesian estimation that iteratively draws samples from conditional posterior distributions. However, Gibbs sampling is inherently sequential and existing parallel implementations focus on reducing the communication between computing units at the cost of increase in recovery error. In this work, we propose a two-stage parallel coefficient update scheme for wavelet-based Bayesian compressive sensing, where the first stage approximates the real distributions of the wavelet coefficients and the second stage computes the final estimate of the coefficients. While in the first stage the parallel computing units share information with each other, in the second stage, the parallel units work independently. We propose a new coefficient update scheme that updates coefficients in both stages based on data generated a few rounds ago. Such a scheme helps in relaxing the timing constraints for communication in the first stage and computations in the second stage. We design the corresponding parallel architecture and synthesize it in 7 nm technology node. We show that in a system with 8 computing units, our method helps reduce the execution time by 17.4× compared to a sequential implementation without any increase in the signal recovery error.

AB - Bayesian compressive sensing (BCS) helps address ill-posed signal recovery problems using the Bayesian estimation framework. Gibbs sampling is a technique used in Bayesian estimation that iteratively draws samples from conditional posterior distributions. However, Gibbs sampling is inherently sequential and existing parallel implementations focus on reducing the communication between computing units at the cost of increase in recovery error. In this work, we propose a two-stage parallel coefficient update scheme for wavelet-based Bayesian compressive sensing, where the first stage approximates the real distributions of the wavelet coefficients and the second stage computes the final estimate of the coefficients. While in the first stage the parallel computing units share information with each other, in the second stage, the parallel units work independently. We propose a new coefficient update scheme that updates coefficients in both stages based on data generated a few rounds ago. Such a scheme helps in relaxing the timing constraints for communication in the first stage and computations in the second stage. We design the corresponding parallel architecture and synthesize it in 7 nm technology node. We show that in a system with 8 computing units, our method helps reduce the execution time by 17.4× compared to a sequential implementation without any increase in the signal recovery error.

KW - Bayesian compressive sensing

KW - Gibbs sampling

KW - parallel implementation

UR - http://www.scopus.com/inward/record.url?scp=85061357131&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85061357131&partnerID=8YFLogxK

U2 - 10.1109/SiPS.2018.8598375

DO - 10.1109/SiPS.2018.8598375

M3 - Conference contribution

AN - SCOPUS:85061357131

T3 - IEEE Workshop on Signal Processing Systems, SiPS: Design and Implementation

SP - 140

EP - 145

BT - Proceedings of the IEEE Workshop on Signal Processing Systems, SiPS 2018

PB - Institute of Electrical and Electronics Engineers Inc.

ER -