Multikernel adaptive filtering

研究成果: Article

88 引用 (Scopus)

抄録

This paper exemplifies that the use of multiple kernels leads to efficient adaptive filtering for nonlinear systems. Two types of multikernel adaptive filtering algorithms are proposed. One is a simple generalization of the kernel normalized least mean square (KNLMS) algorithm , adopting a coherence criterion for dictionary designing. The other is derived by applying the adaptive proximal forward-backward splitting method to a certain squared distance function plus a weighted block l 1 norm penalty, encouraging the sparsity of an adaptive filter at the block level for efficiency. The proposed multikernel approach enjoys a higher degree of freedom than those approaches which design a kernel as a convex combination of multiple kernels. Numerical examples show that the proposed approach achieves significant gains particularly for nonstationary data as well as insensitivity to the choice of some design-parameters.

元の言語English
記事番号6203609
ページ(範囲)4672-4682
ページ数11
ジャーナルIEEE Transactions on Signal Processing
60
発行部数9
DOI
出版物ステータスPublished - 2012
外部発表Yes

Fingerprint

Adaptive filtering
Adaptive filters
Glossaries
Nonlinear systems

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Signal Processing

これを引用

Multikernel adaptive filtering. / Yukawa, Masahiro.

:: IEEE Transactions on Signal Processing, 巻 60, 番号 9, 6203609, 2012, p. 4672-4682.

研究成果: Article

@article{1bd6172b91cc47dd9fe4167cc863a2ad,
title = "Multikernel adaptive filtering",
abstract = "This paper exemplifies that the use of multiple kernels leads to efficient adaptive filtering for nonlinear systems. Two types of multikernel adaptive filtering algorithms are proposed. One is a simple generalization of the kernel normalized least mean square (KNLMS) algorithm , adopting a coherence criterion for dictionary designing. The other is derived by applying the adaptive proximal forward-backward splitting method to a certain squared distance function plus a weighted block l 1 norm penalty, encouraging the sparsity of an adaptive filter at the block level for efficiency. The proposed multikernel approach enjoys a higher degree of freedom than those approaches which design a kernel as a convex combination of multiple kernels. Numerical examples show that the proposed approach achieves significant gains particularly for nonstationary data as well as insensitivity to the choice of some design-parameters.",
keywords = "Block soft-thresholding operator, kernel adaptive filtering, reproducing kernel Hilbert space",
author = "Masahiro Yukawa",
year = "2012",
doi = "10.1109/TSP.2012.2200889",
language = "English",
volume = "60",
pages = "4672--4682",
journal = "IEEE Transactions on Signal Processing",
issn = "1053-587X",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "9",

}

TY - JOUR

T1 - Multikernel adaptive filtering

AU - Yukawa, Masahiro

PY - 2012

Y1 - 2012

N2 - This paper exemplifies that the use of multiple kernels leads to efficient adaptive filtering for nonlinear systems. Two types of multikernel adaptive filtering algorithms are proposed. One is a simple generalization of the kernel normalized least mean square (KNLMS) algorithm , adopting a coherence criterion for dictionary designing. The other is derived by applying the adaptive proximal forward-backward splitting method to a certain squared distance function plus a weighted block l 1 norm penalty, encouraging the sparsity of an adaptive filter at the block level for efficiency. The proposed multikernel approach enjoys a higher degree of freedom than those approaches which design a kernel as a convex combination of multiple kernels. Numerical examples show that the proposed approach achieves significant gains particularly for nonstationary data as well as insensitivity to the choice of some design-parameters.

AB - This paper exemplifies that the use of multiple kernels leads to efficient adaptive filtering for nonlinear systems. Two types of multikernel adaptive filtering algorithms are proposed. One is a simple generalization of the kernel normalized least mean square (KNLMS) algorithm , adopting a coherence criterion for dictionary designing. The other is derived by applying the adaptive proximal forward-backward splitting method to a certain squared distance function plus a weighted block l 1 norm penalty, encouraging the sparsity of an adaptive filter at the block level for efficiency. The proposed multikernel approach enjoys a higher degree of freedom than those approaches which design a kernel as a convex combination of multiple kernels. Numerical examples show that the proposed approach achieves significant gains particularly for nonstationary data as well as insensitivity to the choice of some design-parameters.

KW - Block soft-thresholding operator

KW - kernel adaptive filtering

KW - reproducing kernel Hilbert space

UR - http://www.scopus.com/inward/record.url?scp=84865211015&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84865211015&partnerID=8YFLogxK

U2 - 10.1109/TSP.2012.2200889

DO - 10.1109/TSP.2012.2200889

M3 - Article

AN - SCOPUS:84865211015

VL - 60

SP - 4672

EP - 4682

JO - IEEE Transactions on Signal Processing

JF - IEEE Transactions on Signal Processing

SN - 1053-587X

IS - 9

M1 - 6203609

ER -