Interpolation operators associated with sub-frame sets

From MaRDI portal





Let \(\psi \in L^2(\mathbb{R})\). A normalized tight frame wavelet for a subspace \(X\subset L^2(\mathbb{R})\) is a collection of functions \(\{\psi_{j,k}(x) = 2^{j/2} \psi(2^jx-k): j, k \in\mathbb{Z}\}\), with the property \[ \forall f \in X, \quad f = \sum \langle f, \psi_{j,k} \rangle \psi_{j,k}. \] A measurable set \(E \subset\mathbb{R}\) is called a \textit{sub-frame set} if \(\{\psi_{j,k}: j,k \in\mathbb{Z}\}\) is a normalized tight frame for the closed subspace generated by this sequence, where the function \(\psi\) is of the form \(\widehat \psi = 1/(2\pi) \chi_E\). We say that \(E\) is a \textit{frame set} if this subspace is \(L^2(\mathbb{R})\), and \(E\) is a wavelet set if \(\{\psi_{j,k}\}\) is in fact an orthornormal basis. For any two sub-frame sets, \(E\) and \(F\), which are translation congruent, this congruence extends to an interpolation mapping and we may associate with such a map an interpolation operator \(U\) on \(L^2(\mathbb{R})\). These operators are important in deriving new wavelets by interpolating simple wavelet sets. In the paper under review, the author characterizes when the interpolation operators are elements of the local commutant (i.e., \(C_\psi (D, T) = \{S \in B(L^2(\mathbb{R})): (SD^jT^k - D^jT^kS)\psi =0\), \(j,k \in\mathbb{Z}\}\), where \(D\) is a dilation by 2, and \(T\) is a translation by 1). A corollary of this characterization is that if two sets, \(E\) and \(F\), form an interpolation pair (\(U^2 = \)Id) then one is an MRA wavelet set if and only if the other is.











This page was built for publication: Interpolation operators associated with sub-frame sets

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3151204)