Eigenvalue asymptotics for a class of multi-variable Hankel matrices

Abstract: A one-variable Hankel matrix

H_{a}

is an infinite matrix

H_{a} = [a (i + j)]_{i, j g e q 0}

. Similarly, for any

d g e q 2

, a

d

-variable Hankel matrix is defined as

H_{m a t h b f a} = [m a t h b f a (m a t h b f i + m a t h b f j)]

, where

m a t h b f i = (i_{1}, d o t s, i_{d})

and

m a t h b f j = (j_{1}, d o t s, j_{d})

, with

i_{1}, d o t s, i_{d}, j_{1}, d o t s, j_{d} g e q 0

. For

g a m m a > 0

, A. Pushnitski and D. Yafaev proved that the eigenvalues of the compact one-variable Hankel matrices

H_{a}

with

a (j) = j^{- 1} (l o g j)^{- g a m m a}

, for

j g e q 2

, obey the asymptotics

l a m b d a_{n} (H_{a}) s i m C_{g} a m m a n^{- g a m m a}

, as

n o + i n f t y

, where the constant

C_{g} a m m a

is calculated explicitly. This paper presents the following

d

-variable analogue. Let

g a m m a > 0

and

a (j) = j^{- d} (l o g j)^{- g a m m a}

, for

j g e q 2

. If

m a t h b f a (j_{1}, d o t s, j_{d}) = a (j_{1} + d o t s + j_{d})

, then

H_{m a t h b f a}

is compact and its eigenvalues follow the asymptotics

l a m b d a_{n} (H_{m a t h b f a}) s i m C_{d, g a m m a} n^{- g a m m a}

, as

n o + i n f t y

, where the constant

C_{d, g a m m a}

is calculated explicitly.

This page was built for publication: Eigenvalue asymptotics for a class of multi-variable Hankel matrices