Stochastic behavior of the nonnegative least mean fourth algorithm for stationary Gaussian inputs and slow learning

Jingen Ni; Jian Yang; Jie Chen; Cédric Richard; José Carlos M. Bermudez

doi:10.1016/j.sigpro.2016.03.020

Stochastic behavior of the nonnegative least mean fourth algorithm for stationary Gaussian inputs and slow learning

Jingen Ni, Jian Yang, Jie Chen, Cédric Richard, José Carlos M. Bermudez

School of Marine Science and Technology

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

7 Scopus citations

Abstract

Some system identification problems impose nonnegativity constraints on the parameters to be estimated due to inherent physical characteristics of the unknown system. The nonnegative least-mean-square (NNLMS) algorithm and its variants allow one to address this problem in an online manner. A nonnegative least mean fourth (NNLMF) algorithm has been recently proposed to improve the performance of these algorithms in cases where the measurement noise is not Gaussian. This paper provides a first theoretical analysis of the stochastic behavior of the NNLMF algorithm for stationary Gaussian inputs and slow learning. Simulation results illustrate the accuracy of the proposed analysis.

Original language	English
Pages (from-to)	18-27
Number of pages	10
Journal	Signal Processing
Volume	128
DOIs	https://doi.org/10.1016/j.sigpro.2016.03.020
State	Published - Nov 2016

Keywords

Adaptive filter
Least mean fourth (LMF)
Mean weight behavior
Nonnegativity constraint
Second-order moment
System identification

Access to Document

10.1016/j.sigpro.2016.03.020

Cite this

@article{0b738a25b875474e9cb4ceb707b2190c,

title = "Stochastic behavior of the nonnegative least mean fourth algorithm for stationary Gaussian inputs and slow learning",

abstract = "Some system identification problems impose nonnegativity constraints on the parameters to be estimated due to inherent physical characteristics of the unknown system. The nonnegative least-mean-square (NNLMS) algorithm and its variants allow one to address this problem in an online manner. A nonnegative least mean fourth (NNLMF) algorithm has been recently proposed to improve the performance of these algorithms in cases where the measurement noise is not Gaussian. This paper provides a first theoretical analysis of the stochastic behavior of the NNLMF algorithm for stationary Gaussian inputs and slow learning. Simulation results illustrate the accuracy of the proposed analysis.",

keywords = "Adaptive filter, Least mean fourth (LMF), Mean weight behavior, Nonnegativity constraint, Second-order moment, System identification",

author = "Jingen Ni and Jian Yang and Jie Chen and C{\'e}dric Richard and Bermudez, {Jos{\'e} Carlos M.}",

year = "2016",

month = nov,

doi = "10.1016/j.sigpro.2016.03.020",

language = "英语",

volume = "128",

pages = "18--27",

journal = "Signal Processing",

issn = "0165-1684",

publisher = "Elsevier B.V.",

}

TY - JOUR

T1 - Stochastic behavior of the nonnegative least mean fourth algorithm for stationary Gaussian inputs and slow learning

AU - Ni, Jingen

AU - Yang, Jian

AU - Chen, Jie

AU - Richard, Cédric

AU - Bermudez, José Carlos M.

PY - 2016/11

Y1 - 2016/11

N2 - Some system identification problems impose nonnegativity constraints on the parameters to be estimated due to inherent physical characteristics of the unknown system. The nonnegative least-mean-square (NNLMS) algorithm and its variants allow one to address this problem in an online manner. A nonnegative least mean fourth (NNLMF) algorithm has been recently proposed to improve the performance of these algorithms in cases where the measurement noise is not Gaussian. This paper provides a first theoretical analysis of the stochastic behavior of the NNLMF algorithm for stationary Gaussian inputs and slow learning. Simulation results illustrate the accuracy of the proposed analysis.

AB - Some system identification problems impose nonnegativity constraints on the parameters to be estimated due to inherent physical characteristics of the unknown system. The nonnegative least-mean-square (NNLMS) algorithm and its variants allow one to address this problem in an online manner. A nonnegative least mean fourth (NNLMF) algorithm has been recently proposed to improve the performance of these algorithms in cases where the measurement noise is not Gaussian. This paper provides a first theoretical analysis of the stochastic behavior of the NNLMF algorithm for stationary Gaussian inputs and slow learning. Simulation results illustrate the accuracy of the proposed analysis.

KW - Adaptive filter

KW - Least mean fourth (LMF)

KW - Mean weight behavior

KW - Nonnegativity constraint

KW - Second-order moment

KW - System identification

UR - http://www.scopus.com/inward/record.url?scp=84962859617&partnerID=8YFLogxK

U2 - 10.1016/j.sigpro.2016.03.020

DO - 10.1016/j.sigpro.2016.03.020

M3 - 文章

AN - SCOPUS:84962859617

SN - 0165-1684

VL - 128

SP - 18

EP - 27

JO - Signal Processing

JF - Signal Processing

ER -

Stochastic behavior of the nonnegative least mean fourth algorithm for stationary Gaussian inputs and slow learning

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this