BZPEERSource author record
Explore this author’s publications, collaborators and research topics. This catalog record has not been claimed by its author.
Catalog footprint
Published work
In this work, we consider symmetric random Toeplitz matrices Tn generated by i.i.d. zero mean random variables Xk satisfying the moment conditions: E∣Xk∣2=1 and $\E|X_1|^n \le n^{\sqrt{n}}$ for all n≥3. We prove that the largest eigenvalue of Tn scaled by nlog(n) converges almost surely to 1.
In this article, we study the fluctuations of the random variable: In(ρ)=N1logdet(ΣnΣn∗+ρIN),(ρ>0) where Σn=n−1/2Dn1/2XnD~n1/2+An, as the dimensions of the matrices go to infinity at the same pace. Matrices Xn and An are respectively random and deterministic N×n matrices; matrices Dn and D~n are deterministic and diagonal, with respective dimensions N×N and n×n; matrix Xn=(Xij) has centered, independent and identically distributed entries with unit variance, either real or complex. We prove that when centered and properly rescaled, the random variable In(ρ) satisfies a Central Limit Theorem and has a Gaussian limit. The variance of In(ρ) depends on the moment $\E X_{ij}^2$ of the variables Xij and also on its fourth cumulant $κ= \E|X_{ij}|^4 - 2 - |\E X_{ij}^2|^2$. The main motivation comes from the field of wireless communications, where In(ρ) represents the mutual information of a multiple antenna radio channel. This article closely follows the companion article "A CLT for Information-theoretic statistics of Gram random matrices with a given variance profile", {\em Ann. Appl. Probab. (2008)} by Hachem et al., however the study of the fluctuations associated to non-centered large random matrices raises specific issues, which are addressed here.
Research graph
Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.
BZPEER is loading the nearby papers, people, topics and institutions for this page.