Universal commutative operator algebras and transfer function realizations of polynomials (Q1928565)

The paper is devoted to the universal commutative operator algebra associated with finite-dimensional operator spaces. Let \(E\) be a finite-dimensional operator space of dimension \(n\). Each commuting \(n\)-tuple \(S=(S_{1},\dots,S_{n})\) in \(\mathcal {B} (H)\) defines a linear functional calculus \(\sigma_{S}:E\rightarrow\mathcal {B}(H)\). In the commutative polynomial algebra \(\mathcal{P}(E)\) in elements of \(E\), one can define a matrix norm in the following way: \(\|p\|=\sup{\|p(S)\|}\); the supremum is taken over all commuting tuples \(S\) such that \(\sigma_{S}\) is a matrix contraction. The obtained completion of \(\mathcal {P} (E)\), denoted by \(UC(E)\), is called the universal commutative operator algebra of \(E\). The universal property of \(UC(E)\) is described by the following assertion. There exists a canonical matrix isometric embedding \(L:E\rightarrow UC(E)\) and, if \(\tau:E\rightarrow\mathcal {B}(H)\) is a matrix contraction with commutative range, then there exists a unique matrix contractive unital homomorphism \(\tau:UC(E)\rightarrow\mathcal {B}(H)\) extending \(\tau\).