Wavelet frames for (not necessarily reducing) affine subspaces. II: The structure of affine subspaces (Q630776)

The authors study affine subspaces of \(L^2(\mathbb R)\), i.e., closed linear subspaces of the form \(\{2^{n/2}\psi(2^nt-k): \psi \in \Phi, n,k\in \mathbb Z\}\) where \(\Phi \subset L^2(\mathbb R)\). Such an affine subspace is said to be reducing with respect to translation and dilation if it is invariant under \(D^n\) and \(T^m\), \(n,m\in \mathbb Z\) where \(Df(t)=\sqrt 2f(\sqrt 2 t)\) and \(Tf(t)= f(t-1)\) and the purpose of this paper is to extend as much as possible of the theory of reducing subspaces to non-reducing ones. The authors prove several results on structure of non reducing affine subspaces, e.g., that every affine subspace is the orthogonal direct sum of at most three purely non-reducing ones (i.e ones without any nonzero reducing subspace) while every reducing subspace is the orthogonal direct sum of two purely non-reducing ones. The basic question here is when the orthogonal complement of an affine subspace in another affine subspace is again affine. The authors are not able to achieve their goal of answering the question whether every affine subspace is singly generated but they prove that every affine subspace can be decomposed into a direct sum sum of a singly generated one and one of ``small size''. [For part I, cf. the authors, Appl. Comput. Harmon. Anal. 27, No. 1, 47--54 (2009; Zbl 1169.42017).]