The curvature invariant of a Hilbert module over ℂ [z1, ..., zd] (Q4488484)

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 1470854
Language Label Description Also known as
default for all languages
No label defined
    English
    The curvature invariant of a Hilbert module over ℂ [z1, ..., zd]
    scientific article; zbMATH DE number 1470854

      Statements

      The curvature invariant of a Hilbert module over ℂ [z1, ..., zd] (English)
      0 references
      0 references
      5 July 2000
      0 references
      finite free resolution
      0 references
      curvature invariant
      0 references
      Euler characteristic
      0 references
      Gauss-Bonnet-Chern theorem
      0 references
      curvature of a Hilbert module
      0 references
      curvature of a graded Hilbert module
      0 references
      The author introduces a curvature invariant, which essentially differs from the corresponding notion of \textit{M. J. Cowen} and \textit{R. G. Douglas} [see Acta Math. 141, 187-261 (1978; Zbl 0427.47016)]. We shall briefly describe this invariant. Let \(\overline T= (T_1,\dots, T_d)\) be a \(d\)-tuple of mutually commuting operators acting in a Hilbert space \(H\), supposed to be a \(d\)-contraction, i.e., \(\|T_1 x_1+\cdots+ T_d x_d\|^2\leq\|x_1\|^2+\cdots+ \|x_d\|^2\) for all \(x_1,\dots, x_d\in H\). In this case the operator \(\Delta= (\text{\textbf{1}}- T_1 T^*_1-\cdots- T_d T^*_d)^{1/2}\) exists, and one may define the function \(F(z)x= \Delta(\text{\textbf{1}}- T(z)^*)^{- 1}(\text{\textbf{1}}- T(z))^{- 1}\Delta x\), \(x\in\Delta H\), where \(T(z)=\overline z_1 T_1+\cdots+ \overline z_d T_d\) with \(z\) in the unit ball \(B_d= \{z\in\mathbb{C}^d:|z|< 1\}\). Assuming, in addition, that \(\Delta H\) is finite-dimensional, the author shows that for \(\sigma\)-almost every \(w\in\partial B_d\), the limit \(K_0(w)= \lim_{r\uparrow 1}(1- r^2)\text{ trace }F(rw)\) exists, where \(\sigma\) is the natural rotation-invariant probability measure on the sphere \(\partial B_d\). This allows the author to define the curvature invariant \(K(H)\) as the integral \(\int_{\partial B_d} K_0(w) d\sigma(w)\). An important part of this ample paper deals with the problem of computing this invariant. If \(A= \mathbb{C}[z_1,\dots, z_d]\) is the algebra of complex polynomials in \(z_1,\dots, z_d\), then \(H\) can be viewed as a Hilbert \(A\)-module, via the action \(f\cdot x= f(T_1,\dots, T_d)x\), \(x\in H\), \(f\in A\). If \(y_1,\dots, y_d\) is a linear basis of \(\Delta H\), then \(M_H= \{f_1\cdot y_1+\cdots+ f_d\cdot y_d; f_1,\dots, f_d\in A\}\) is a finitely generated \(A\)-module. In particular, \(M_H\) has a finite free resolution, and the Euler characteristic associated with this resolution is, by definition, the Euler characteristic \(\chi(H)\) of \(H\). If the \(d\)-tuple \((T_1,\dots, T_d)\) also possesses a circular symmetry with an additional property (i.e., the spectrum of the corresponding gauge group is bounded below), the author shows that \(K(H)= \chi(H)\), and so in this case \(K(H)\) is an integer. This is an operator-theoretic analogue of the Gauss-Bonnet-Chern theorem, whose proof is based on an asymptotic formula which expresses the curvature of a Hilbert module as the trace of a certain selfadjoint operator. This result also gives an answer to the problem of computing the curvature of a graded Hilbert module. The case of ungraded Hilbert modules remains open.
      0 references

      Identifiers