A Class of Adaptive Algorithms Based on Entropy Estimation Achieving CRLB for Linear Non-Gaussian Filtering

Adaptive filtering has been extensively studied under the assumption that the noise is Gaussian. The most commonly used least-mean-square-error (LMSE) filter is optimal when the noise is Gaussian. However, in many practical applications, the noise can be modeled more accurately using a non-Gaussian...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on signal processing 2012-04, Vol.60 (4), p.2049-2055
Main Authors: Hualiang Li, Xi-Lin Li, Anderson, M., Adali, T.
Format: Article
Language:English
Subjects:
Adaptive filters
Adaptive signal processing
Applied sciences
Computational modeling
Cramér-Rao bounds
Detection, estimation, filtering, equalization, prediction
Entropy
entropy estimation
Exact sciences and technology
Filtering
Gaussian
Information, signal and communications theory
Least squares approximation
Mathematical models
Maximum likelihood estimation
Noise
Non-Gaussian
non-Gaussian noise
Optimization
Shape
Signal and communications theory
Signal to noise ratio
Signal, noise
Telecommunications and information theory
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Adaptive filtering has been extensively studied under the assumption that the noise is Gaussian. The most commonly used least-mean-square-error (LMSE) filter is optimal when the noise is Gaussian. However, in many practical applications, the noise can be modeled more accurately using a non-Gaussian distribution. In this correspondence, we consider non-Gaussian distributions for the noise model and show that the filter of using entropy bound minimization (EBM) leads to significant performance gain compared to the LMSE filter. The least mean p-norm (LMP) filter using the α-stable distribution to model noise is shown to be the maximum-likelihood solution when using the generalized Gaussian distribution (GGD) to model noise. The GGD model for noise allows us to compute the Cramér-Rao lower bound (CRLB) for the error in estimating the weights. Simulations show that both the EBM and LMP filters achieve the CRLB as the sample size increases. The EBM filter is shown to be less committed with respect to unseen data yielding generally superior performance in online learning when compared to LMP. We also show that, when the noise comes from impulsive α -stable distributions, both the EBM and LMP filters provide better performance than LMSE. In addition, the EBM filter offers the advantage that it does not assume a certain parametric model for the noise, and by proper selection of the measuring functions, it can be adapted to a wide range of noise distributions.
ISSN:1053-587X
1941-0476
DOI:10.1109/TSP.2011.2182345