Inverse spectral analysis for finite matrix-valued Jacobi operators

Abstract: Consider the Jacobi operators

c J

given by

(c J y)_{n} = a_{n} y_{n + 1} + b_{n} y_{n} + a_{n - 1}^{*} y_{n - 1}

,

y_{n} i n C^{m}

(here

y_{0} = y_{p + 1} = 0

), where

b_{n} = b_{n}^{*}

and

a_{n} : d e t a_{n} e 0

are the sequences of

m s m

matrices,

n = 1, .., p

. We study two cases: (i)

a_{n} = a_{n}^{*} > 0

; (ii)

a_{n}

is a lower triangular matrix with real positive entries on the diagonal (the matrix

c J

is

(2 m + 1)

-band

m p s m p

matrix with positive entries on the first and the last diagonals). The spectrum of

c J

is a finite sequence of real eigenvalues

l_{1} < ... < l_{N}

, where each eigenvalue

l_{j}

has multiplicity

k_{j} l e m

. We show that the mapping

(a, b) m a p s t o {(l_{j}, k_{j})}_{1}^{N} o p l u s a d d i t i o n a l s p e c t r a l d a t a

is 1-to-1 and onto. In both cases (i), olinebreak (ii), we give the complete solution of the inverse problem.

This page was built for publication: Inverse spectral analysis for finite matrix-valued Jacobi operators