On the well-posedness of the rational covariance extension problem (Q4258512)
From MaRDI portal
!
This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:
scientific article; zbMATH DE number 1335839
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | On the well-posedness of the rational covariance extension problem |
scientific article; zbMATH DE number 1335839 |
Statements
28 February 2000
0 references
Schur algorithm
0 references
rational extension
0 references
positive extension
0 references
Carathéodory problem
0 references
rational covariance extension problem
0 references
well-posedness
0 references
Toeplitz matrix
0 references
Schur polynomial
0 references
0.6719372
0 references
0.65061736
0 references
0.64434314
0 references
0.63532114
0 references
On the well-posedness of the rational covariance extension problem (English)
0 references
The rational covariance extension problem is the following: given a finite sequence of real numbers \(c_0,\ldots,c_n\), find all extensions \(c_{n+1},c_{n+2},\ldots\) such that \(c(z)=c_0/2+\sum_{k=1}^\infty c_k z^{-k}\) is rational of degree at most \(n\) and positive real, i.e., \(c(z)\) is analytic in \(|z|\geq 1\) and \(c(z)+c(z^{-1})>0\) for all \(|z|=1\). I. Schur has parametrized all meromorphic solutions (not necessarily of finite degree) in terms of Schur polynomials (polynomials having all their zeros inside the open unit disk) or equivalently in terms of the reflection coefficients or Schur parameters. On the other hand, it is possible to describe all the interpolants of finite degree, but then positivity is not guaranteed. It was only in 1995 that the rational extension problem was completely solved by \textit{C. I. Byrnes}, \textit{A. Lindquist}, \textit{S. V. Gusev} and \textit{A. S. Matveev} [IEEE Trans. Autom. Control 40, No. 11, 1841-1857 (1995; Zbl 0847.93008)]. The current paper investigates the well-posedness of the problem. The main theorem of this paper says that if \(c_0,\ldots,c_n\), is the first row of a symmetric positive definite Toeplitz matrix, then for every Schur polynomial \(\sigma\), there is a unique Schur polynomial \(a\) and a constant \(\rho\), such that \(c(z)=\rho a(z)/\sigma(z)\) solves the rational extension problem. Moreover, the mapping from data to solution is a homeomorphism. The latter confirms the well-posedness of the problem in the sense of Hadamard. For the entire collection see [Zbl 0863.00008].
0 references