Exact NLMS algorithm with lp-norm constraint

Luis Weruaga; Shihab Jimaa

doi:10.1109/LSP.2014.2360889

Exact NLMS algorithm with l_p-norm constraint

Luis Weruaga
, Shihab Jimaa

Electrical Engineering

Research output: Contribution to journal › Article › peer-review

26 Scopus citations

Abstract

This letter presents the exact normalized least-mean-square (NLMS) algorithm for the l_p-norm-regularized square error, a popular choice for the identification of sparse systems corrupted by additive noise. The resulting exact l_p-NLMS algorithm manifests differences to the original one, such as an independent update for each weight, a new sparsity-promoting compensated update, and the guarantee of stable convergence for any configuration (regardless the choice of l_p norm and sparsity-tradeoff constant). Simulation results show that the exact l_p-NLMS is stable and it outperforms the original one, thus validating the optimality of the proposed methodology.

Original language	British English
Article number	6913546
Pages (from-to)	366-370
Number of pages	5
Journal	IEEE Signal Processing Letters
Volume	22
Issue number	3
DOIs	https://doi.org/10.1109/LSP.2014.2360889
State	Published - 1 Mar 2015

Keywords

l-norm constraint
Newton optimization
normalized least mean square (NLMS) algorithm
sparsity

Access to Document

10.1109/LSP.2014.2360889

Cite this

@article{e1f59cd44a6d418593b83739f0ca47bd,

title = "Exact NLMS algorithm with lp-norm constraint",

abstract = "This letter presents the exact normalized least-mean-square (NLMS) algorithm for the lp-norm-regularized square error, a popular choice for the identification of sparse systems corrupted by additive noise. The resulting exact lp-NLMS algorithm manifests differences to the original one, such as an independent update for each weight, a new sparsity-promoting compensated update, and the guarantee of stable convergence for any configuration (regardless the choice of lp norm and sparsity-tradeoff constant). Simulation results show that the exact lp-NLMS is stable and it outperforms the original one, thus validating the optimality of the proposed methodology.",

keywords = "l-norm constraint, Newton optimization, normalized least mean square (NLMS) algorithm, sparsity",

author = "Luis Weruaga and Shihab Jimaa",

note = "Publisher Copyright: {\textcopyright} 2014 IEEE.",

year = "2015",

month = mar,

day = "1",

doi = "10.1109/LSP.2014.2360889",

language = "British English",

volume = "22",

pages = "366--370",

journal = "IEEE Signal Processing Letters",

issn = "1070-9908",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

number = "3",

}

TY - JOUR

T1 - Exact NLMS algorithm with lp-norm constraint

AU - Weruaga, Luis

AU - Jimaa, Shihab

PY - 2015/3/1

Y1 - 2015/3/1

N2 - This letter presents the exact normalized least-mean-square (NLMS) algorithm for the lp-norm-regularized square error, a popular choice for the identification of sparse systems corrupted by additive noise. The resulting exact lp-NLMS algorithm manifests differences to the original one, such as an independent update for each weight, a new sparsity-promoting compensated update, and the guarantee of stable convergence for any configuration (regardless the choice of lp norm and sparsity-tradeoff constant). Simulation results show that the exact lp-NLMS is stable and it outperforms the original one, thus validating the optimality of the proposed methodology.

AB - This letter presents the exact normalized least-mean-square (NLMS) algorithm for the lp-norm-regularized square error, a popular choice for the identification of sparse systems corrupted by additive noise. The resulting exact lp-NLMS algorithm manifests differences to the original one, such as an independent update for each weight, a new sparsity-promoting compensated update, and the guarantee of stable convergence for any configuration (regardless the choice of lp norm and sparsity-tradeoff constant). Simulation results show that the exact lp-NLMS is stable and it outperforms the original one, thus validating the optimality of the proposed methodology.

KW - l-norm constraint

KW - Newton optimization

KW - normalized least mean square (NLMS) algorithm

KW - sparsity

UR - https://www.scopus.com/pages/publications/84907929433

U2 - 10.1109/LSP.2014.2360889

DO - 10.1109/LSP.2014.2360889

M3 - Article

AN - SCOPUS:84907929433

SN - 1070-9908

VL - 22

SP - 366

EP - 370

JO - IEEE Signal Processing Letters

JF - IEEE Signal Processing Letters

IS - 3

M1 - 6913546

ER -

Exact NLMS algorithm with l_p-norm constraint

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this