The expectation and sparse maximization algorithm

Steffen Barembruch, Anna Scaglione, Eric Moulines

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

In recent years, many sparse estimation methods, also known as compressed sensing, have been developed. However, most of these methods presume that the measurement matrix is completely known. We develop a new blind maximum likelihood method-the expectation-sparse-maximization (ESpaM) algorithm-for models where the measurement matrix is the product of one unknown and one known matrix. This method is a variant of the expectation-maximization algorithm to deal with the resulting problem that the maximization step is no longer unique. The ESpaM algorithm is justified theoretically. We present as well numerical results for two concrete examples of blind channel identification in digital communications, a doubly-selective channel model and linear time invariant sparse channel model.

Original languageEnglish (US)
Pages (from-to)317-329
Number of pages13
JournalJournal of Communications and Networks
Volume12
Issue number4
StatePublished - Aug 2010
Externally publishedYes

Fingerprint

Compressed sensing
Maximum likelihood
Identification (control systems)
Communication

Keywords

  • Compressive sensing (CS)
  • Deconvolution
  • Multipath channels
  • Smoothing methods

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Information Systems

Cite this

The expectation and sparse maximization algorithm. / Barembruch, Steffen; Scaglione, Anna; Moulines, Eric.

In: Journal of Communications and Networks, Vol. 12, No. 4, 08.2010, p. 317-329.

Research output: Contribution to journalArticle

Barembruch, S, Scaglione, A & Moulines, E 2010, 'The expectation and sparse maximization algorithm', Journal of Communications and Networks, vol. 12, no. 4, pp. 317-329.
Barembruch, Steffen ; Scaglione, Anna ; Moulines, Eric. / The expectation and sparse maximization algorithm. In: Journal of Communications and Networks. 2010 ; Vol. 12, No. 4. pp. 317-329.
@article{243172d5d27247d2a6c978ce9dfc6f84,
title = "The expectation and sparse maximization algorithm",
abstract = "In recent years, many sparse estimation methods, also known as compressed sensing, have been developed. However, most of these methods presume that the measurement matrix is completely known. We develop a new blind maximum likelihood method-the expectation-sparse-maximization (ESpaM) algorithm-for models where the measurement matrix is the product of one unknown and one known matrix. This method is a variant of the expectation-maximization algorithm to deal with the resulting problem that the maximization step is no longer unique. The ESpaM algorithm is justified theoretically. We present as well numerical results for two concrete examples of blind channel identification in digital communications, a doubly-selective channel model and linear time invariant sparse channel model.",
keywords = "Compressive sensing (CS), Deconvolution, Multipath channels, Smoothing methods",
author = "Steffen Barembruch and Anna Scaglione and Eric Moulines",
year = "2010",
month = "8",
language = "English (US)",
volume = "12",
pages = "317--329",
journal = "Journal of Communications and Networks",
issn = "1229-2370",
publisher = "Korean Institute of Communication Sciences",
number = "4",

}

TY - JOUR

T1 - The expectation and sparse maximization algorithm

AU - Barembruch, Steffen

AU - Scaglione, Anna

AU - Moulines, Eric

PY - 2010/8

Y1 - 2010/8

N2 - In recent years, many sparse estimation methods, also known as compressed sensing, have been developed. However, most of these methods presume that the measurement matrix is completely known. We develop a new blind maximum likelihood method-the expectation-sparse-maximization (ESpaM) algorithm-for models where the measurement matrix is the product of one unknown and one known matrix. This method is a variant of the expectation-maximization algorithm to deal with the resulting problem that the maximization step is no longer unique. The ESpaM algorithm is justified theoretically. We present as well numerical results for two concrete examples of blind channel identification in digital communications, a doubly-selective channel model and linear time invariant sparse channel model.

AB - In recent years, many sparse estimation methods, also known as compressed sensing, have been developed. However, most of these methods presume that the measurement matrix is completely known. We develop a new blind maximum likelihood method-the expectation-sparse-maximization (ESpaM) algorithm-for models where the measurement matrix is the product of one unknown and one known matrix. This method is a variant of the expectation-maximization algorithm to deal with the resulting problem that the maximization step is no longer unique. The ESpaM algorithm is justified theoretically. We present as well numerical results for two concrete examples of blind channel identification in digital communications, a doubly-selective channel model and linear time invariant sparse channel model.

KW - Compressive sensing (CS)

KW - Deconvolution

KW - Multipath channels

KW - Smoothing methods

UR - http://www.scopus.com/inward/record.url?scp=77956588222&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77956588222&partnerID=8YFLogxK

M3 - Article

VL - 12

SP - 317

EP - 329

JO - Journal of Communications and Networks

JF - Journal of Communications and Networks

SN - 1229-2370

IS - 4

ER -