A function-theoretic approach to a parametrization of the set of solutions of a completely indeterminate matricial Nehari problem (Q1378132): Difference between revisions

The matrical Nehari problem consists of the following: Let \((\gamma_k)_{k\in\mathbb{N}}\) be a sequence of \(n\times m\) complex matrices. Describe the set \({\mathcal N}[(\gamma_k)_{k\in{\mathcal N}}]\) of all \((\lambda\)-equivalence classes) \(\langle f\rangle_{\lambda}\) from the unit ball of \(L^\infty_{n\times m}({\mathbf T})\) which satisfy \[ {1\over 2\pi} \int_{\mathbf T} z^kf(z)\lambda d(z)= \gamma_k\quad\text{for all }k\in\mathbb{N}, \] where \(\lambda\) is the linear Lebesgue-Borel measure on the unit circle \({\mathbf T}\). The authors give a function-theoretic proof of the following known result: Suppose that the Nehari problem corresponding to \((\gamma_k)_{k\in\mathbb{N}}\) is completely indeterminate. Then there exists a \(\lambda\)-a.e. unique \(A\)-normalized \(\gamma\)-generating \((n,m)\)-type function \(U\) such that \[ {\mathcal N}[(\gamma_k)_{k\in\mathbb{N}}]= \{\langle ({P_-}\underline{\zeta+Q_-}) (Q_+ \zeta+{P_+})^{-1}\rangle_{\underline\lambda}: \zeta\in S_{n\times m}({\mathbf D})\}, \] where \([P_-, Q_-, Q_+, P_+]\) is the \(\gamma\)-generating quadruple associated with \(\langle U\rangle_\lambda\). Owing to their method, they show new results. For example, the unique \(A\)-normalized resolvent matrix \(U\) is constructed by limit procedure. Moreover, some limit relations for the Potapov-Ginzburg transform of \(U\) are obtained.