Random normal matrices: eigenvalue correlations near a hard wall

Abstract: We study the eigenvalues of some

n i m e s n

random normal matrices. In our situation, the

n

eigenvalues are separated into two disjoint components by a hard wall. We study the corresponding correlation kernel

K_{n} (z, w)

as

n o i n f t y

in two diagonal regimes (with either

| z - w | = m a t h c a l O (f r a c 1 s q r t n)

or

| z - w | = m a t h c a l O (f r a c 1 n)

), as well as in three off-diagonal regimes (with

| z - w | a s y m p 1

). For some of these regimes, the asymptotics are described by oscillatory theta functions and weighted SzegH{o} kernels. When

z, w

are both at a distance of order

f r a c 1 n

from the hard wall, our results also show a slow decay of

K_{n} (z, w)

as

| z - w |

increases.

This page was built for publication: Random normal matrices: eigenvalue correlations near a hard wall