Analysis of the TDLMS algorithm operating in a nonstationary environment

Kuhn, Eduardo Vinicius; Kolodziej, Javier E.; Seara, Rui

doi:10.1016/j.dsp.2015.05.013

Cited by 10 publications

(3 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, it has a slow convergence in the case of highly correlated input signals [2,3]. This is due to the fact that the autocorrelation matrix of the input signal has a large eigenvalue spread.…”

Section: Introductionmentioning

confidence: 93%

A new pre-whitening transform domain LMS algorithm and its application to speech denoising

Chergui

Bouguezel

2017

Signal Processing

View full text Add to dashboard Cite

Section: Introductionmentioning

confidence: 93%

A new pre-whitening transform domain LMS algorithm and its application to speech denoising

Chergui

Bouguezel

2017

Signal Processing

View full text Add to dashboard Cite

“…Its transformed version X k N is also of length N and obtained using the N×N orthogonal transform matrix T N as Figure 1. Block diagram of the TDLMS adaptive filter [13,18] The sample y k at the output of the filter is obtained as…”

Section: Conventional Tdlms Algorithmmentioning

confidence: 99%

A New Post-whitening Transform Domain LMS Algorithm

Chergui¹,

Bouguezel²

2019

View full text Add to dashboard Cite

This paper proposes a new post-whitening transform domain LMS (POW-TDLMS) algorithm for system identification purposes, where the post whitened and original transformed signals are used during the adaptation and filtering phases, respectively. The main idea behind the proposed algorithm is to introduce a first order adaptive post-whitening filter in the TDLMS algorithm after applying the transform to completely decorrelate the transformed signal. Linear prediction is adopted for the post-whitening and the prediction coefficients are adapted in the time domain. Furthermore, the mean convergence performance analysis of the proposed POW-TDLMS algorithm is presented. The simulation results show the superiority of the proposed POW-TDLMS algorithm compared to the conventional TDLMS algorithm in terms of the MSE convergence speed and reached steady state.

show abstract

“…In the stochastic modeling of adaptive algorithms, the aim is to derive expressions describing (with certain accuracy) the algorithm behavior under different operating conditions (Duttweiler 2000;Rupp 1993;Doroslovački and Deng 2006;Wagner and Doroslovački 2008a, b;Loganathan et al 2010;de Souza et al 2010b;Kuhn et al 2014a;Haddad and Petraglia 2014). So, from the model expressions, causeand-effect relationships between algorithm parameters and performance metrics can be established, resulting in useful design guidelines (Abadi and Husøy 2009;Kuhn et al 2014b;Kuhn et al 2015;Matsuo and Seara 2016). Moreover, model expressions can help in identifying anomalous and/or undesired behavior, allowing to modify the algorithm to circumvent such an issue (Kolodziej et al 2009).…”

mentioning

confidence: 99%

On the Stochastic Modeling of the NSVR–IAF–PNLMS Algorithm for Correlated Gaussian Input Data

Kuhn

Perez

Souza

et al. 2020

J Control Autom Electr Syst

Self Cite

View full text Add to dashboard Cite

This paper presents a stochastic model for the normalized smoothed variation rate individual-activation-factor proportionate normalized least-mean-square (NSVR-IAF-PNLMS) algorithm. Specifically, taking into account correlated Gaussian input data, model expressions are derived for predicting the mean weight vector, gain distribution matrix, NSVR metric, learning curve, weight-error correlation matrix, and steady-state excess mean-square error. Such expressions are obtained by considering the time-varying characteristics of the gain distribution matrix. Simulation results are shown confirming the accuracy of the proposed model for different operating conditions. Keywords Adaptive filtering • Proportionate normalized least-mean-square algorithm • Stochastic model 1 Introduction Although the least-mean-square (LMS) and the normalized LMS (NLMS) algorithms have been commonly used in many real-world applications (Sayed 2009; Farhang-Boroujeny 2013; Haykin 2014), these algorithms exhibit slow conver-B Rui Seara

show abstract

Analysis of the TDLMS algorithm operating in a nonstationary environment

Cited by 10 publications

References 25 publications

A new pre-whitening transform domain LMS algorithm and its application to speech denoising

A new pre-whitening transform domain LMS algorithm and its application to speech denoising

A New Post-whitening Transform Domain LMS Algorithm

On the Stochastic Modeling of the NSVR–IAF–PNLMS Algorithm for Correlated Gaussian Input Data

Contact Info

Product

Resources

About