Degenerate interpolation problems for Nevanlinna pairs (Q809334)

This paper studies interpolation problems for equivalence classes of pairs of matrix functions of the extended Nevanlinna class using two different approaches, namely via the Krein-Langer theory of extensions of symmetric operators and via the de Branges theory of Hilbert spaces of analytic functions. Adaptations of these methods to the study of interpolation problems have been introduced earlier by Alpay-Bruinsma- Dijksma-de Snoo and Dym respectively. The formulation of the interpolation conditions is very general, encoding finitely many directional interpolation conditions of arbitrarily multiplicity in a compact form. The point of formulating the problem for equivalence classes of pairs rather than for functions is to handle functions which formally assume the value infinity identically on a subspace in a rigorous manner. The new feature here is to obtain a linear fractional description for the set of all solutions in the optimal case where the associated Pick matrix may be singular. The parametrization is given as the image of an explicitly computable linear fractional map acting on the set of all Nevanlinna pairs which satisfy a linear side constraint identically. In the last decade or so interpolation theory of this type for matrix functions has been an area of intense activity, due in part to applications in areas such as \(H^{\infty}\) control theory. This paper serves to complete the formalism for one of the several approaches now available.