Wold-type decomposition for Doubly commuting two-isometries
Monojit Bhattacharjee, Rajeev Gupta, Vidhya Venugopal · 2025
Read Wold-type decomposition for Doubly commuting two-isometries by Monojit Bhattacharjee, Rajeev Gupta, Vidhya Venugopal (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-16933.txt.
Introduction
Monojit Bhattacharjee, Rajeev Gupta, and Vidhya Venugopal proved a splitting theorem for pairs of operators in a 2025 paper. They work with two-isometries, operators that behave like isometries up to two steps, in the case where the pair doubly commutes. The authors show any such pair admits a Wold-type decomposition, the classic split of an operator into a pure shift part and a unitary part. The shift part is identified concretely: it acts as multiplication by the coordinate functions on a Dirichlet-type function space over the bidisc.
Catalog record
| CREATOR | Monojit Bhattacharjee, Rajeev Gupta, Vidhya Venugopal |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 35 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-16933
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Wold-type decomposition for Doubly commuting two-isometries Monojit Bhattacharjee, Rajeev Gupta, Vidhya Venugopal Abstract . In this article, we prove that any pair of doubly commuting 2 2 -isometries on a Hilbert space has a Wold-type decomposition. Moreover, the analytic part of the pair is unitary equivalent to the pair of multiplication by coordinate function on a Dirichlet-type space on the bidisc. 2010 Mathematics Subject Classification Primary 46E20, 47B32, 47B38, Secondary 47A50, 31C25 1. Introduction Let \mathcal{H} be a Hilbert space and \mathcal{B}(\mathcal{H}) denote the set of all bounded linear operators on \mathcal{H} . An operator T\in\mathcal{B}(\mathcal{H}) is said to be concave if for every h\in\mathcal{H}, \displaystyle\|T^{2}h\|^{2}-2\|Th\|^{2}+\|h\|^{2}\leq 0. (1) An operator T\in\mathcal{B}(\mathcal{H}) is called a 2 -isometry if for each h\in\mathcal{H}, ( 1 ) holds with equality. A closed subspace \mathcal{W}\subseteq\mathcal{H} is said to be a wandering subspace for T if \mathcal{W}\perp T^{n}\mathcal{W} for all n\geq 1 and an operator T\in\mathcal{B}(\mathcal{H}) is said to have wandering subspace property if \mathcal{H}=\bigvee_{n\geq 0}T^{n}\mathcal{W} for some wandering subspace \mathcal{W} of T . For an operator T\in\mathcal{B}(\mathcal{H}) with the wandering subspace property and a wandering subspace \mathcal{W} , it holds that \mathcal{H}\ominus T\mathcal{H}=\mathcal{W} , which shows that the wandering subspace is uniquely determined. Throughout the article, the subspace \mathcal{H}\ominus T\mathcal{H} shall be denoted by \mathcal{E}_{T}. A key and highly useful result in the study of bounded linear operators on a Hilbert space is the Wold decomposition theorem. The classical version states that any isometry V\in\mathcal{B}(\mathcal{H}) (i.e., V^{*}V=I_{\mathcal{H}} ) is a direct sum of a unitary and an unilateral shift (see [ 22 ] and also see page 3 in [ 10 ] ). The Wold decomposition and results analogous to it have profound impact in Operator theory, Operator algebra and also invariant subspace problem for Hilbert space of holomorphic functions (see [ 5 , 8 , 6 , 21 ] ). Wold-type decompositions of 2 -isometric operators also have been studied by many authors, see [ 16 , 18 , 12 ] . For a 2 -isometry T\in\mathcal{B}(\mathcal{H}) , it is well-known that the subspace \displaystyle\cap_{n\geq 0}T^{n}\mathcal{H} (:= \mathcal{H}_{0} ) is T -reducing and T|_{\mathcal{H}_{0}} is unitary (see [ 12 , Proposition 1.1] ). The operator T\in\mathcal{B}(\mathcal{H}) is said to be analytic provided \mathcal{H}_{0}=\{0\} . In [ 17 ] , the author proves the following Wold-type decomposition for a the class of analytic concave operators. Theorem 1.1 ( [ 17 ] ) . Let T\in\mathcal{B}(\mathcal{H}) be an analytic concave operator. Then \mathcal{H}\ominus T\mathcal{H} is a wandering subspace for T. Later, Shimorin extended this result for general bounded left-invertible operators (see [ 18 , Theorem 3.6] ). For the sake of completeness of this article, we state this result only for the class of 2 -isometries in Theorem 1.2 . Theorem 1.2 ( [ 12 ] ) . Let T be a 2 -isometry on \mathcal{H} . Then there exists a positive \mathcal{B}(\mathcal{E}_{T}) -valued operator measure \mu on the unit circle \mathbb{T} such that \big(T,\mathcal{H}\big)\cong\left(\begin{bmatrix}U&0\\ 0&M_{z}\end{bmatrix},\mathcal{H}_{0}\oplus\mathcal{D}_{\mathcal{E}_{T}}(\mu)\right), where \mathcal{H}_{0} is T -reducing subspace, U is unitary and M_{z} on \mathcal{D}_{\mathcal{E}_{T}}(\mu) is an analytic 2 -isometry. For the definition and properties of \mathcal{D}_{\mathcal{E}_{T}}(\mu) space, we refer the reader to [ 16 , 12 , 14 ] . A natural question which has been an interest to many operator theorist is that ‘Can an analogous result be obtained in multivariate setting?’ There are many interesting results for commuting isometries as well as for the class of pairs of commuting analytic 2 -isometries with some additional assumptions on the pair (see [ 19 , 8 , 4 , 1 , 2 , 9 ] and references therein). It is presumable that some extra assumptions on pair of commuting 2 -isometries are necessary to achieve results analogous to Theorem 1.1 and Theorem 1.2 . In this article, we shall consider the class of pairs of doubly commuting 2 -isometries. Here is the definition of doubly commuting tuple of operators. An n -tuple of commuting bounded linear operators T=(T_{1},\ldots,T_{n}) acting on a Hilbert space \mathcal{H} is said to be doubly commuting if \displaystyle T_{i}^{*}T_{j}=T_{j}T_{i}^{*} for any 1\leq i\neq j\leq n . The natural examples of doubly commuting tuples are those of the co-ordinate wise multiplication operators on the Hardy space, the Bergman space, and the Dirichlet space over the bidisc \mathbb{D}^{2} (defined in Section 2 and also see [ 9 ] ). A Wold-type decomposition for a pair of doubly commuting isometries is derived in [ 19 ] (see also [ 3 , 7 , 15 , 20 ] ).More precisely, Theorem 1.3 . (M. Slocinski) Let V=(V_{1},V_{2}) be a pair of doubly commuting isometries on a Hilbert space \mathcal{H} . Then there exists a unique decomposition \mathcal{H}=\mathcal{H}_{ss}\oplus\mathcal{H}_{su}\oplus\mathcal{H}_{us}\oplus\mathcal{H}_{uu}, where \mathcal{H}_{ij} is joint V -reducing subspace of \mathcal{H} for all i,j=s,u. Moreover, V_{1} on \mathcal{H}_{ij} is a shift if i=s and unitary if i=u and that V_{2} is a shift if j=s and unitary if j=u . In Theorem 4.3 , we establish a Wold-type decomposition for any pair of doubly commuting 2 -isometries acting on a Hilbert space as a generalization of Theorem 1.2 . The essence of analytic structure in it also gives a more finer version of Theorem 5.1 in [ 11 ] . The novelty of this work lies in the treatment of a general pair of doubly commuting 2 -isometries, which are not necessarily cyclic or analytic, and their connection to shifts on Dirichlet spaces. As a corollary of this theorem one can derive Theorem 1.3 . Before we turn our attention to Theorem 4.3 , in Section 2 we note a few notations and preliminaries. 2. Preliminaries Let \mu_{1} and \mu_{2} to be two positive \mathcal{B}(\mathcal{E}) -valued operator measures defined on the unit circle \mathbb{T} , where \mathcal{E} is a Hilbert space. For i=1,2 , the Poisson integral of \mu_{i} is denoted by P[\mu_{i}] and is given by P[\mu_{i}](z_{i})=\int_{\mathbb{T}}P(z_{i},e^{i\theta})d\mu_{i}(e^{i\theta}) where P(z_{i},e^{i\theta})=\frac{(1-|z_{i}|^{2})}{|e^{i\theta}-z_{i}|^{2}} for any z_{i}\in\mathbb{D} and \theta\in\mathbb{R}. Note that P[\mu_{i}](z_{i}) is a positive operator in \mathcal{B}({\mathcal{E}}). Let the space of all \mathcal{E} -valued analytic functions on the bidisc be denoted by \mathcal{O}(\mathbb{D}^{2},\mathcal{E}) . Now, for any f\in\mathcal{O}(\mathbb{D}^{2},\mathcal{E}) , we define the terms D^{(2)}_{\mu_{1},\mu_{2},i}(f) for i=1,2,3, \displaystyle D^{(2)}_{\mu_{1},\mu_{2},1}(f) \displaystyle:=\lim_{r\rightarrow 1^{-}}\iint_{\mathbb{D}\times\mathbb{T}}\left\langle P[\mu_{1}](z_{1})\frac{\partial f}{\partial z_{1}}(z_{1},re^{it}),\frac{\partial f}{\partial z_{1}}(z_{1},re^{it})\right\rangle_{\mathcal{E}}\,\,dt~~dA(z_{1}) \displaystyle D^{(2)}_{\mu_{1},\mu_{2},2}(f) \displaystyle:=\lim_{r\rightarrow 1^{-}}\iint_{\mathbb{T}\times\mathbb{D}}\left\langle P[\mu_{2}](z_{2})\frac{\partial f}{\partial z_{2}}(re^{it},z_{2}),\frac{\partial f}{\partial z_{2}}(re^{it},z_{2})\right\rangle_{\mathcal{E}}\,\,dA(z_{2})~~dt \displaystyle D^{(2)}_{\mu_{1},\mu_{2},3}(f) \displaystyle:=\iint_{\mathbb{\mathbb{D}}\times\mathbb{D}}\left\langle P[\mu_{1}](z_{1})P[\mu_{2}](z_{2})\frac{\partial^{2}f}{\partial z_{1}\partial z_{2}}(z_{1},z_{2}),\frac{\partial^{2}f}{\partial z_{1}\partial z_{2}}(z_{1},z_{2})\right\rangle_{\mathcal{E}}\,\,dA(z_{1})~~dA(z_{2}), where dt denotes the normalized arc length measure and dA denotes the normalized area measure. Furthermore, we define the Dirichlet-type integral of a function f on bidisc as D^{(2)}_{\mu_{1},\mu_{2}}(f):=D^{(2)}_{\mu_{1},\mu_{2},1}(f)+D^{(2)}_{\mu_{1},\mu_{2},2}(f)+D^{(2)}_{\mu_{1},\mu_{2},3}(f). Combining all these, the Dirichlet-type space over bidisc for the operator-valued measures {\mu_{1},\mu_{2}} on \mathbb{T} , denoted by \mathcal{D}^{2}_{\mathcal{E}}({\mu_{1},\mu_{2}}) , is defined to be the space of all \mathcal{E} -valued analytic functions f such that D^{(2)}_{\mu_{1},\mu_{2}}(f) is finite. The norm of f\in\mathcal{D}^{2}_{\mathcal{E}}(\mu_{1},\mu_{2}) is denoted by \|f\|_{\mu_{1},\mu_{2}} and is given by \|f\|_{\mu_{1},\mu_{2}}^{2}:=\|f\|_{H^{2}(\mathbb{D}^{2})}^{2}+D^{(2)}_{\mu_{1},\mu_{2},1}(f)+D^{(2)}_{\mu_{1},\mu_{2},2}(f)+D^{(2)}_{\mu_{1},\mu_{2},3}(f), (2) The space \mathcal{D}^{2}_{\mathcal{E}}(\mu_{1},\mu_{2}) turns out to be a Hilbert space with respect to the norm given in ( 2 ). In [ 9 ] , the scalar-valued Dirichlet-type space (that is, when \mathcal{E}=\mathbb{C} ) is studied in some detail. Let \mathbb{Z}_{\geq 0} denote the set of all non-negative integers. Throughout the article, for m,p\in\mathbb{Z}_{\geq 0}, we adopt (m\wedge p) as a concise notation for \min\{m,p\}. For any f\in\mathcal{D}^{2}_{\mathcal{E}}({\mu_{1},\mu_{2}}) with f(z_{1},z_{2})=\sum_{m,n=0}^{\infty}a_{m,n}z_{1}^{m}z_{2}^{n} , the following identities can easily