New challenges in covariance estimation: multiple structures and coarse\n quantization

2021/06/11 by Johannes Maly, Tianyu Yang, Maly, Johannes +7 · 1 citation
Computer Science · #Blind Source Separation Techniques #Target Tracking and Data Fusion in Sensor Networks #Distributed Sensor Networks and Detection Algorithms

paper · pdf · doi:10.48550/arxiv.2106.06190

Abstract

In this self-contained chapter, we revisit a fundamental problem of\nmultivariate statistics: estimating covariance matrices from finitely many\nindependent samples. Based on massive Multiple-Input Multiple-Output (MIMO)\nsystems we illustrate the necessity of leveraging structure and considering\nquantization of samples when estimating covariance matrices in practice. We\nthen provide a selective survey of theoretical advances of the last decade\nfocusing on the estimation of structured covariance matrices. This review is\nspiced up by some yet unpublished insights on how to benefit from combined\nstructural constraints. Finally, we summarize the findings of our recently\npublished preprint "Covariance estimation under one-bit quantization" to show\nhow guaranteed covariance estimation is possible even under coarse quantization\nof the samples.\n

Cited by

• Compressive Toeplitz Covariance Estimation From Few-Bit Quantized Measurements With Applications to DOA Estimation

Related

• Covariance estimation under one-bit quantization
• Regularized estimation of large covariance matrices
• A Random Matrix--Theoretic Approach to Handling Singular Covariance Estimates
• A Structured Estimator for large Covariance Matrices in the Presence of Pairwise and Spatial Covariates
• New Methods for Handling Singular Sample Covariance Matrices