Large information plus noise random matrix models and consistent subspace estimation in large sensor networks

DOI10.1142/S2010326311500067MaRDI QIDQ2884853zbMATH OpenOpenAlexFDO

Authors W. Hachem, Philippe Loubaton, X. Mestre, Jamal Najim, Pascal Vallet

Publication date 18 May 2012

Published in Random Matrices: Theory and Applications (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1106.5119

zbMATH Keywords

consistencyregularizationcovariance matrixquadratic formsingular valueslocalization of the eigenvaluesrandom matrix modelsMUSICsubspace estimationGaussian information plus noise matrixinformation plus noise modellarge sensor networksPoincaré inequality

Mathematics Subject Classification ID

Point estimation (62F10) Analysis of variance and covariance (ANOVA) (62J10) Signal theory (characterization, reconstruction, filtering, etc.) (94A12) Eigenvalues, singular values, and eigenvectors (15A18) Random matrices (algebraic aspects) (15B52)

Abstract: In array processing, a common problem is to estimate the angles of arrival of $K$ deterministic sources impinging on an array of $M$ antennas, from $N$ observations of the source signal, corrupted by gaussian noise. The problem reduces to estimate a quadratic form (called "localization function") of a certain projection matrix related to the source signal empirical covariance matrix. Recently, a new subspace estimation method (called "G-MUSIC") has been proposed, in the context where the number of available samples $N$ is of the same order of magnitude than the number of sensors $M$. In this context, the traditional subspace methods tend to fail because the empirical covariance matrix of the observations is a poor estimate of the source signal covariance matrix. The G-MUSIC method is based on a new consistent estimator of the localization function in the regime where $M$ and $N$ tend to $+infty$ at the same rate. However, the consistency of the angles estimator was not adressed. The purpose of this paper is to prove the consistency of the angles of arrival estimator in the previous asymptotic regime. To prove this result, we show the property that the singular values of M x N Gaussian information plus noise matrix escape from certain intervals is an event of probability decreasing at rate O(1/N^p) for all p. A regularization trick is also introduced, which allows to confine these singular values into certain intervals and to use standard tools as Poincar'e inequality to characterize any moments of the estimator. These results are believed to be of independent interest.

Recommendations

• A subspace estimator for fixed rank perturbations of large random matrices
• Asymptotic properties of subspace estimators
• Robust spiked random matrices and a robust G-MUSIC estimator
• The singular values and vectors of low rank perturbations of large rectangular random matrices
• Spectrum estimation for large dimensional covariance matrices using random matrix theory

Cites work

• A Simple Approach to the Global Regime of Gaussian Ensembles of Random Matrices
• A new application of random matrices: \(\operatorname{Ext} (C_{\text{red}}^*(F_2))\) is not a group
• An inequality for the multivariate normal distribution
• Analysis of the limiting spectral distribution of large dimensional information-plus-noise type matrices
• Deterministic equivalents for certain functionals of large random matrices
• Distributions of Matrix Variates and Latent Roots Derived from Normal Samples
• MUSIC, maximum likelihood, and Cramer-Rao bound
• Modified Subspace Algorithms for DoA Estimation With Large Arrays
• On the Capacity Achieving Covariance Matrix for Rician MIMO Channels: An Asymptotic Approach
• On the empirical distribution of eigenvalues of large dimensional information-plus-noise-type matrices
• The estimation of frequency
• The largest eigenvalues of finite rank deformation of large Wigner matrices: Convergence and nonuniversality of the fluctuations

Cited in

(10)

• A CLT for linear spectral statistics of large random information-plus-noise matrices
• Some remarks on the Dozier-Silverstein theorem for random matrices with dependent entries
• On the inference about the spectral distribution of high-dimensional covariance matrix based on high-frequency noisy observations
• Robust spiked random matrices and a robust G-MUSIC estimator
• Random matrix theory in statistics: a review
• Identifiability of parametric random matrix models
• Cauchy noise loss for stochastic optimization of random matrix models via free deterministic equivalents
• On generalized expectation-based estimation of a population spectral distribution from high-dimensional data
• A subspace estimator for fixed rank perturbations of large random matrices
• Analysis of the limiting spectral measure of large random matrices of the separable covariance type

This page was built for publication: Large information plus noise random matrix models and consistent subspace estimation in large sensor networks

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2884853)