Generalized average sampling in shift invariant spaces (Q626139): Difference between revisions

In this paper, the authors study generalized average sampling in shift-invariant spaces with frame generators. In Corollary 3.3, the authors give a classical average sampling theorem that reads as follows: Let \(\{u_n(t):n\in\mathbb Z\}\) be any sequence of average functions with \(\text{supp}\, u_n(t)\subset [n-a, n+b]\) and \(\delta:=\max\{a,b\}\). If \(\sqrt{\delta(a+b)}<\frac{\alpha}{L}\), then there is a frame \(\{S_n(t):n\in\mathbb Z\}\) of \(V(\phi)\) such that \(f(t)=\sum_{n\in\mathbb Z}\langle f, u_n \rangle S_n(t)\), \(f\in V(\phi)\), which converges in \(L^2(\mathbb R)\) and absolutely and uniformly on \(\mathbb R\). In Corollary 3.8, the authors give another average sampling theorem that reads as follows: Assume that \(u_n(t)\) is nondecreasing on \([n-a,n]\) and nonincreasing on \([n,n+b]\). If \(\sqrt{M(a+b)}<\frac{\pi\alpha}{2L}\), where \(M:=\sup_{n\in\mathbb Z}\{b\int_n^{n+b}u_n(t)dt, a\int_{n-a}^n u_n(t)dt\}\), then there is a frame \(\{S_n(t):n\in\mathbb Z\}\) of \(V(\phi)\) such that \(f(t)=\sum_{n\in\mathbb Z}\langle f, u_n \rangle S_n(t)\), \(f\in V(\phi)\), which converges in \(L^2(\mathbb R)\) and absolutely and uniformly on \(\mathbb R\).