A Nonparametric VSS NLMS Algorithm

Benesty, Jacob; Rey, Hernán G.; Vega, Leonardo Rey; Tressens, S.

doi:10.1109/lsp.2006.876323

Cited by 336 publications

(186 citation statements)

References 13 publications

Supporting

Mentioning

174

Contrasting

Unclassified

Order By: Relevance

“…In all simulations, we show the normalized MSD, E w o − w(k) 2 / w o 2 which is evaluated by ensemble averaging over 20 independent trials. Also, we assume that the noise variance, σ 2 v , is known a priori [13]. For all simulations we consider α = 0.99, and C = 10 −5 .…”

Section: Simulation Resultsmentioning

confidence: 99%

The New Normalized Subband Adaptive Filter Algorithms with Variable Step-Size

Abadi

Shafiee

2014

Automatika

View full text Add to dashboard Cite

Original scientific paperThis paper presents a new variable step-size normalized subband adaptive filter (VSS-NSAF) algorithm. In the proposed VSS-NSAF, the step-size changes in order to have largest decrease in the mean square deviation (MSD) for sequential iterations. To reduce the computational complexity of VSS-NSAF, the variable step-size selective partial update normalized subband adaptive filter (VSS-SPU-NSAF) is proposed. In this algorithm the filter coefficients are partially updated in each subband at every iteration. Simulation results show the good performance of the proposed algorithms in convergence speed and steady-state MSD.

show abstract

Section: Simulation Resultsmentioning

confidence: 99%

The New Normalized Subband Adaptive Filter Algorithms with Variable Step-Size

Abadi

Shafiee

2014

Automatika

View full text Add to dashboard Cite

show abstract

“…It is a technique used by majority of the algorithms presented in this paper, namely by Harris, 7 Shan, 15 Karni, 16 Benveniste, 17 Evans, 18 Mathews, 19 Ang, 20 Benesty, 21 Wahab, 22 Hwang 23 and Wang. 24 In many solutions this technique is based on the orthogonality principle, which states that (under some assumptions) the necessary and sufficient condition for the mean-square error to attain its minimum value is that the error signal e(n) and the input signal u(n) are orthogonal.…”

Section: Introductionmentioning

confidence: 99%

“…19 Three parameters, but without an upper bound for the step size, are also required by Benesty's algorithm. 21 The remaining algorithms are parametrized by more than three values, with the maximum of eight in case of Zou's algorithm. 14 The goal of this paper is to review the VS-LMS techniques developed by different authors for different uses and to compare them in three typical applications: identification, line enhancement, and adaptive noise cancellation.…”

Section: Introductionmentioning

confidence: 99%

Review and Comparison of Variable Step-Size LMS Algorithms

Bismor¹,

Czyż²,

Ogonowski³

2016

IJAV

View full text Add to dashboard Cite

The inherent feature of the Least Mean Squares (LMS) algorithm is the step size, and it requires careful adjustment. Small step size, required for small excess mean square error, results in slow convergence. Large step size, needed for fast adaptation, may result in loss of stability. Therefore, many modifications of the LMS algorithm, where the step size changes during the adaptation process depending on some particular characteristics, were and are still being developed.The paper reviews seventeen of the best known variable step-size LMS (VS-LMS) algorithms to the degree of detail that allows to implement them. The performance of the algorithms is compared in three typical applications: parametric identification, line enhancement, and adaptive noise cancellation. The paper suggests also one general modification that can simplify the choice of the upper bound for the step size, which is a crucial parameter for many VS-LMS algorithms. u(n)y

show abstract

“…However, it is not desirable to rely only on DTD because the NEC has tight requirements for detecting double-talk. Thus, several adaptive filtering algorithms are developed focusing on the robustness to double-talk [10]- [14]. One of the method is using L 1 -norm minimization of error because L 1 -norm algorithms are especially robust against impulsive noise such as speech.…”

Section: Introductionmentioning

confidence: 99%

Proportionate Sign Subband Adpative Filtering Algorithm for Network Echo Cancellers

Seo¹,

Jung²,

Park³

2015

JACN

View full text Add to dashboard Cite

Abstract-This paper proposes a proportionate sign subband adaptive filtering (PSSAF) and an improved PSSAF (IPSSAF) algorithm for network echo cancellers that deal with impulsive interferences and sparse echo paths. Based on a sign subband adaptive filtering (SSAF) algorithm that is robust against impulsive interferences, the proposed algorithm minimizes L 1 -norm of the subband a posteriori error vector subject to a weighted constraint on the filter coefficients. A positive definite weighting matrix is used for the constraint. In this paper, we adopt a diagonal proportionate matrix for the constraint to achieve fast initial convergence rate when the impulse response is sparse. The components of it is roughly proportional to the absolute value of current estimate of the filter, so the resulting algorithm is called PSSAF algorithm. To achieve fast initial convergence rate even for rather non-sparse impulse responses, an improved proportionate matrix is also used for the constraint, and the resulting algorithm is named as IPSSAF algorithm. Experimental results show that the proposed algorithms are more robust against highly correlated input signals, impulsive interferences, and double-talk than the original normalized subband adaptive filtering algorithm and the class of proportionate subband adaptive filtering algorithms.Index Terms-Adaptive filters, proportionate, sign algorithm, subband adaptive filtering algorithm, weighted constraint, weighting matrix.

show abstract

A Nonparametric VSS NLMS Algorithm

Cited by 336 publications

References 13 publications

The New Normalized Subband Adaptive Filter Algorithms with Variable Step-Size

The New Normalized Subband Adaptive Filter Algorithms with Variable Step-Size

Review and Comparison of Variable Step-Size LMS Algorithms

Proportionate Sign Subband Adpative Filtering Algorithm for Network Echo Cancellers

Contact Info

Product

Resources

About