Reconstruction of bandlimited functions from space-time samples (Q2662518)

This article is concerned with the problem of reconstructing bandlimited functions from a set of space-time samples. In the classical Whittaker-Shannon sampling theorem, a bandlimited function is constructed from its samples at uniformly distributed discrete set of points. The construction is achieved using translates of the Sinc function \[\mbox{Sinc }x=\left\{ \begin{array}{ll} \frac{\sin \pi x}{\pi x},& x\neq 0 \\ 1& x=0. \end{array} \right.\] In the case where the samples are not uniformly distributed, kernels other than the Sinc function are used. The space of bandlimited functions with bandwidth \(\sigma\) is usually referred to as the Paley-Wiener space and is denoted by \(PW_\sigma.\) The authors consider the following problem: Given a uniformly discrete (u.d.) set of points \(\Lambda\subset {\mathbb R} \) satisfying \[\delta(\Lambda)=\inf_{\lambda, \lambda^\prime\in \Lambda, \lambda\neq \lambda^\prime}\left| \lambda -\lambda^\prime\right| > 0, \] and a kernel \(\left\{\phi_u, u\in I \right\}\) where \(I\) is an interval, under what conditions can a bandlimited function \(f\) be reconstructed in a stable way from the data set \[ \left\{(f\ast \phi_u)(\lambda): \lambda\in \Lambda , u\in I \right\} ? \] By a stable reconstruction, it is meant that \[A\left\| f\right\|_2^2 \leq \int_I \sum_{\lambda\in \Lambda}\left| (f \ast \phi_u )( \lambda) \right|^2 du \leq B\left\| f\right\|_2^2 , \quad \forall f\in PW_\sigma , A, B>0.\] The main result of the paper is to show that stable recovery of bandlimited functions can be achieved for a given u.d set \(\Lambda \subset {\mathbb R} ,\) provided that the kernel \(\left\{\phi_u, u\in I \right\}\) satisfies certain conditions. Having established this result, the authors then proceed to extend the result to the Bernstein space \(B_\sigma\) which is more general than the space \(PW_\sigma.\)