\(m\)-isometric commuting tuples of operators on a Hilbert space (Q850901): Difference between revisions

\textit{J.\,Agler} and \textit{M.\,Stankus} [Integral Equations Oper.\ Theory 21, 383--429 (1995; Zbl 0836.47008)] defined a Hilbert space operator \(T\) to be an \(m\)-isometry if \[ \sum_{k=0}^m(-1)^k {m \choose k} (T^\ast)^{m-k}T^{m-k}=0. \] The multivariable generalization of these single variable \(m\)-isometries is defined as follows: Let \((T_1,\dots, T_d)\) be a \(d\)-tuple of pairwise commuting operators acting on a Hilbert space \(\mathcal H\). For an operator \(A\) on \(\mathcal H\), let \(Q_T(A)=\sum_{i=1}^dT^\ast_iAT_i\). As \(I-Q_T\) is an operator on \({\mathcal B}({\mathcal H})\), denote \(P_m(T)=(I-Q_T)^m(I)\) for \(m\geq 0\), i.e., \[ P_m(T)=\sum_{j=0}^m(-1)^j{m \choose j}Q^j_T(I). \] A commuting \(d\)-tuple \(T=(T_1,\dots, T_d)\) is said to be an \(m\)-isometry if \(P_m(T)=0\). The authors study some basic properties of \(m\)-isometric tuples of commuting operators and give some examples. In particular, they show that the \(d\)-tuple \(M_z=(M_{z_1},\dots,M_{z_d})\), composed of multiplications by the coordinate functions on the Hardy space \(H^2_d\) of analytic functions in the unit ball \({\mathbb B}_d\) of \({\mathbb C}^d\), is a \(d\)-isometry. As an application of techniques used in the paper, they prove a theorem about cyclic vectors in certain spaces of analytic functions that are properly contained in the Hardy space \(H^2_d\).