Covariance Matrix Estimation from Correlated Sub-Gaussian Samples

2019/10/16 by Xu Zhang, Wei Cui, Zhang, Xu +3
Computer Science · Engineering · Mathematics · #Blind Source Separation Techniques #Direction-of-Arrival Estimation Techniques #Sparse and Compressive Sensing Techniques #cs.IT #math.IT #math.ST #stat.TH

paper · pdf · doi:10.48550/arxiv.1910.07183

10 pages, 3 figures

arxiv created 2019/10/16 · arxiv updated 2019/10/17

Abstract

This paper studies the problem of estimating a covariance matrix from correlated sub-Gaussian samples. We consider using the correlated sample covariance matrix estimator to approximate the true covariance matrix. We establish non-asymptotic error bounds for this estimator in both real and complex cases. Our theoretical results show that the error bounds are determined by the signal dimension n, the sample size m and the correlation pattern B. In particular, when the correlation pattern B satisfies tr(B)=m, ||B||_F=O(m^1/2), and ||B||=O(1), these results reveal that O(n) samples are sufficient to accurately estimate the covariance matrix from correlated sub-Gaussian samples. Numerical simulations are presented to show the correctness of the theoretical results.

Citations

• Error Bound for Compound Wishart Matrices
• Multiple emitter location and signal parameter estimation
• Fading correlation and its effect on the capacity of multielement antenna systems

Related

• Covariance estimation with nonnegative partial correlations
• Covariance estimation under one-bit quantization
• Better Together: Cross and Joint Covariances Enhance Signal Detectability in Undersampled Data
• Multivariate Analysis of Nonparametric Estimates of Large Correlation Matrices
• Statistical properties of the estimator using covariance matrix