Paper Info
Title
Coherence Spectrum Estimation From Nonuniformly Sampled Sequences
Abstract
Magnitude squared coherence (MSC) is a useful bivariate spectral measure that finds application in a wide variety of fields. In this paper, we develop a nonparametric Capon-based MSC estimator that utilizes a segmented reformulation of the recently introduced iterative adaptive approach (IAA) to provide high resolution MSC spectrum estimates. The proposed estimator, termed segmented-IAA-MSC (or SIAA-MSC, for short), allows for unevenly sampled data as well as for sequences with arbitrarily missing samples. The estimator first uses segmented-IAA to find accurate estimates of the auto-and cross-covariance matrices of the given sequences. These estimates are then used in a Capon-based MSC estimator reformulated to allow for nonuniformly sampled sequences. To achieve higher statistical accuracy, the estimation problem is formulated so as to allow for overlapped segmentation of the available data. The proposed SIAA-MSC estimator is found to yield improved estimates as compared to the more commonly used least squares Fourier transform (LSFT) based MSC estimator.
Year
DOI
Venue
2010
10.1109/LSP.2010.2040227
IEEE Signal Process. Lett.
Keywords
Field
DocType
covariance matrices,iterative methods,signal sampling,statistical analysis,bivariate spectral measure,coherence spectrum estimation,cross-covariance matrices,iterative adaptive approach,least squares Fourier transform,magnitude squared coherence,nonparametric Capon-based MSC estimator,nonuniformly sampled sequences,statistical accuracy,Capon estimator,coherence spectrum,iterative adaptive approach,missing data,spectral analysis
Least squares,Minimum-variance unbiased estimator,Mathematical optimization,Iterative method,Minimax estimator,Nonparametric statistics,Capon,Missing data,Mathematics,Estimator
Journal
Volume
Issue
ISSN
17
4
1070-9908
Citations 
PageRank 
References 
21
1.49
7
Authors
2
Name
Order
Citations
PageRank
Naveed R. Butt1407.81
Andreas Jakobsson240943.32