Large system spectral analysis of covariance matrix estimation

Husheng Li, H. Vincent Poor

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

Eigendecomposition of estimated covariance matrices is a basic signal processing technique arising in a number of applications, including direction-of-arrival estimation, power allocation in multiple-input/ multiple-output (MIMO) transmission systems, and adaptive multiuser detection. This paper uses the theory of non-crossing partitions to develop explicit asymptotic expressions for the moments of the eigenvalues of estimated covariance matrices, in the large system asymptote as the vector dimension and the dimension of signal space both increase without bound, while their ratio remains finite and nonzero. The asymptotic eigenvalue distribution is also obtained from these eigenvalue moments and the Stieltjes transform, and is extended to first-order approximation in the large sample-size limit. Numerical simulations are used to demonstrate that these asymptotic results provide good approximations for finite systems of moderate size.

Original languageEnglish (US)
Pages (from-to)1395-1422
Number of pages28
JournalIEEE Transactions on Information Theory
Volume55
Issue number3
DOIs
StatePublished - Mar 26 2009

Fingerprint

systems analysis
Covariance matrix
Spectrum analysis
Multiuser detection
Direction of arrival
Signal processing
Computer simulation
simulation

All Science Journal Classification (ASJC) codes

  • Information Systems
  • Library and Information Sciences
  • Computer Science Applications

Cite this

@article{a36a5b079f714598a3569d164de17c5b,
title = "Large system spectral analysis of covariance matrix estimation",
abstract = "Eigendecomposition of estimated covariance matrices is a basic signal processing technique arising in a number of applications, including direction-of-arrival estimation, power allocation in multiple-input/ multiple-output (MIMO) transmission systems, and adaptive multiuser detection. This paper uses the theory of non-crossing partitions to develop explicit asymptotic expressions for the moments of the eigenvalues of estimated covariance matrices, in the large system asymptote as the vector dimension and the dimension of signal space both increase without bound, while their ratio remains finite and nonzero. The asymptotic eigenvalue distribution is also obtained from these eigenvalue moments and the Stieltjes transform, and is extended to first-order approximation in the large sample-size limit. Numerical simulations are used to demonstrate that these asymptotic results provide good approximations for finite systems of moderate size.",
author = "Husheng Li and Poor, {H. Vincent}",
year = "2009",
month = "3",
day = "26",
doi = "https://doi.org/10.1109/TIT.2008.2011440",
language = "English (US)",
volume = "55",
pages = "1395--1422",
journal = "IEEE Transactions on Information Theory",
issn = "0018-9448",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "3",

}

Large system spectral analysis of covariance matrix estimation. / Li, Husheng; Poor, H. Vincent.

In: IEEE Transactions on Information Theory, Vol. 55, No. 3, 26.03.2009, p. 1395-1422.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Large system spectral analysis of covariance matrix estimation

AU - Li, Husheng

AU - Poor, H. Vincent

PY - 2009/3/26

Y1 - 2009/3/26

N2 - Eigendecomposition of estimated covariance matrices is a basic signal processing technique arising in a number of applications, including direction-of-arrival estimation, power allocation in multiple-input/ multiple-output (MIMO) transmission systems, and adaptive multiuser detection. This paper uses the theory of non-crossing partitions to develop explicit asymptotic expressions for the moments of the eigenvalues of estimated covariance matrices, in the large system asymptote as the vector dimension and the dimension of signal space both increase without bound, while their ratio remains finite and nonzero. The asymptotic eigenvalue distribution is also obtained from these eigenvalue moments and the Stieltjes transform, and is extended to first-order approximation in the large sample-size limit. Numerical simulations are used to demonstrate that these asymptotic results provide good approximations for finite systems of moderate size.

AB - Eigendecomposition of estimated covariance matrices is a basic signal processing technique arising in a number of applications, including direction-of-arrival estimation, power allocation in multiple-input/ multiple-output (MIMO) transmission systems, and adaptive multiuser detection. This paper uses the theory of non-crossing partitions to develop explicit asymptotic expressions for the moments of the eigenvalues of estimated covariance matrices, in the large system asymptote as the vector dimension and the dimension of signal space both increase without bound, while their ratio remains finite and nonzero. The asymptotic eigenvalue distribution is also obtained from these eigenvalue moments and the Stieltjes transform, and is extended to first-order approximation in the large sample-size limit. Numerical simulations are used to demonstrate that these asymptotic results provide good approximations for finite systems of moderate size.

UR - http://www.scopus.com/inward/record.url?scp=62749096812&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=62749096812&partnerID=8YFLogxK

U2 - https://doi.org/10.1109/TIT.2008.2011440

DO - https://doi.org/10.1109/TIT.2008.2011440

M3 - Article

VL - 55

SP - 1395

EP - 1422

JO - IEEE Transactions on Information Theory

JF - IEEE Transactions on Information Theory

SN - 0018-9448

IS - 3

ER -