Edge universality for orthogonal ensembles of random matrices (Q836975)

All marginal densities for unitary matrix ensembles can be reproduced in terms of a so called reproducing kernel \(K\). The problem to study marginal distributions is thus replaced by the problem of the analysis of the reproducing kernel \(K\) in the scaling limit. The most general result to prove the edge universality for the kernel \(K\) defined in terms of orthogonal polynomials was obtained by the author in [Commun. Math. Phys. 285, 957--974 (2009; Zbl 1190.82020)], where it was shown that in the one interval case the matrix which is to be controlled is of rank one. This result is used in this paper that up to some small terms the kernel is the same as for Gaussian orthogonal ensembles. This allows to prove the universality of the limiting kernels near the edges. In the proof, the asymptotic of orthogonal polynomials is used instead of the classical asymptotic of Hermite polynomials.