Sampling and Recovery of Multidimensional Bandlimited Functions via Frames

Let 0 \in E \subset \mathbb{R}^d be compact such that for all \lambda > 1, E \subset \mathrm{int}(\lambda E). Choose S=(t_n)_{n \in \mathbb{N}} \subset \mathbb{R}^d such that (f_n)_{n \in \mathbb{N}}, defined by f_n(\cdotp) = e^{i \langle \cdotp , t_n \rangle}, is a frame for L_2(E) with frame operator S. Let \lambda_0 >1 with \mathcal{F}^{-1}(g):\mathbb{R}^d \rightarrow \mathbb{R}, \mathcal{F}^{-1}(g) \in C^\infty where \mathcal{F}^{-1}(g)\arrowvert_{E} = 1 and \mathcal{F}^{-1}(g)\arrowvert_{(\lambda_0 E)^c}=0. If \lambda \geq \lambda_0 and f \in PW_E, then

where B_{kn} = \langle S^{-1}f_n , S^{-1}f_k \rangle_E. Convergence of the sum is in L_2(\mathbb{R}^d), hence also uniform. Further, the map B:\ell_2(\mathbb{N}) \rightarrow \ell_2(\mathbb{N}) defined by (y_k)_{k \in \mathbb{N}} \mapsto \big( \sum_{n \in \mathbb{N}} B_{kn} y_n \big)_{k \in \mathbb{N}} is bounded linear, and is an onto isomorphism iff (f_n)_{n \in \mathbb{N}} is a Riesz basis for L_2(E).

If 0 \in E \subset \mathbb{R}^d is compact, then the following are equivalent:
1) E \subset \mathrm{int}(\lambda E) for all \lambda >1.
2) There exists a continuous map \phi:S^{d-1} \rightarrow (0, \infty) such that E = \{t y \phi(y)|y \in S^{d-1} , t \in [0,1]\}.

Choose (t_n)_{n \in \mathbb{N}} \subset \mathbb{R}^d such that (f_n)_{n \in \mathbb{N}}, defined by f_n(\cdotp) = \frac{1}{(2\pi)^{d/2}}e^{i \langle \cdotp , t_n \rangle}, is a frame for L_2([-\pi,\pi]^d). If f \in PW_E, then

The matrix B and the convergence of the sum are as in Theorem 3.1.

Let \Lambda \subset \mathbb{R}^d be countable such that

If E is a subset of the closed unit ball in \mathbb{R}^d and E has positive measure, then \{e^{i\langle \cdot , \lambda \rangle} | \lambda \in \Lambda\} is a frame for L_2(E).

Let (n_k)_{k \in \mathbb{N}} be an enumeration of \mathbb{Z}^d, and S=(t_k)_{k \in \mathbb{N}} \subset \mathbb{R}^d such that

Define e_k,f_k:\mathbb{R}^d \rightarrow \mathbb{C} by e_k (x) = \frac{1}{(2\pi)^{d/2}} e^{i \langle n_k , x \rangle} and \frac{1}{(2\pi)^{d/2}} e^{i \langle t_k , x \rangle}, and let (h_k)_k be the standard basis for \ell_2(\mathbb{N}). Let P_l:\ell_2(\mathbb{N}) \rightarrow \ell_2(\mathbb{N}) be the orthogonal projection onto \mathrm{span}\{h_1,\cdots,h_l\}. If (f_k)_{k \in \mathbb{N}} is a Riesz basis for L_2[-\pi,\pi]^d, then for all f \in PW_{[-\pi,\pi]^d}, we have

where convergence is in L_2 and uniform. Furthermore,

Convergence of the sum is in L_2 and also uniform.

Under the hypotheses of Theorem 5.1,

where convergence of the sum is both L_2 and uniform.

Let (n_k)_{k \in \mathbb{N}} be an enumeration of \mathbb{Z}^d, and let (t_k)_{k \in \mathbb{N}} \subset \mathbb{R}^d. Define e_k,f_k:\mathbb{R}^d \rightarrow \mathbb{C} by e_k (x) = \frac{1}{(2\pi)^{d/2}} e^{i \langle n_k , x \rangle} and f_k(x) = \frac{1}{(2\pi)^{d/2}} e^{i \langle t_k , x \rangle}. Then for any r,s\geq 1, and any finite sequence (a_k)_{k=r}^s, we have

Let (n_k)_{k \in \mathbb{N}} be an enumeration of \mathbb{Z}^d, and let (t_k)_{k \in \mathbb{N}} \subset \mathbb{R}^d such that

Define e_k,f_k:\mathbb{R}^d \rightarrow \mathbb{C} by e_k (x) = \frac{1}{(2\pi)^{d/2}} e^{i \langle n_k , x \rangle} and \frac{1}{(2\pi)^{d/2}} e^{i \langle t_k , x \rangle}. Then the map T:L_2[-\pi,\pi]^d \rightarrow L_2[-\pi,\pi]^d, defined by T e_n = e_n - f_n, satisfies the following estimate:

Let (n_k)_{k \in \mathbb{N}}, (t_k)_{k \in \mathbb{N}} \subset \mathbb{R}^d, and let e_k, f_k and T be defined as in Corollary 5.4. For each l \in \mathbb{N}, define T_l by T_l e_k = e_k - f_k for 1 \leq k \leq l, and T_l e_k = 0 for l < k. If \lim_{k \rightarrow\infty} \Arrowvert n_k - t_k \Arrowvert_\infty = 0, then \lim_{l \rightarrow \infty} T_l = T in the operator norm. In particular, T is a compact operator.

Let (t_n)_{n \in \mathbb{Z}} \subset \mathbb{R} be a sequence of distinct points such that

Then the sequence of functions (f_k)_{k \in \mathbb{Z}}, defined by f_k (x) = \frac{1}{\sqrt{2 \pi}}e^{i t_k x}, is a Riesz basis for L_2[-\pi,\pi].

Let (n_k)_{k \in \mathbb{N}} be an enumeration of \mathbb{Z}^d, and let (t_k)_{k \in \mathbb{N}} \subset \mathbb{R}^d such that

Then the sequence (f_k)_{k \in \mathbb{N}} defined by f_k (x) = \frac{1}{(2\pi)^{d/2}}e^{i \langle x , t_k \rangle} is a Riesz basis for L_2[-\pi,\pi]^d.

Let (a_n)_{n \in \mathbb{Z}^d} \subset \mathbb{R}^d such that

If D_d(L)<1, then \big(\frac{1}{(2 \pi)^d} e^{i \langle a_n , (\cdot) \rangle}\big) is a Riesz basis for L_2[-\pi,\pi]^d with frame bounds (1-D_d(L))^2 and (1+D_d(L))^2.

Let x_d be the unique number satisfying 0 < x_d <1/4 and D_d(x_d)=1. Then

Let (t_n)_{n \in \mathbb{Z}}\subset \mathbb{R} be a sequence (with t_n \neq 0 for n \neq 0) such that (f_n)_n = \big(\frac{1}{\sqrt{2\pi}} e^{i t_n (\cdot)} \big)_n is a Riesz basis for L_2[-\pi,\pi], and let (e_n)_n be the standard exponential orthonormal basis for L_2[-\pi,\pi]. If the map L given by L e_n = f_n satisfies the estimate \Arrowvert I-L \Arrowvert<1, then the biorthogonals G_n of \frac{1}{\sqrt{2\pi}}\mathcal{F}(f_n)(\cdot) = \mathrm{sinc}(\pi(\cdot-t_n)) in PW_{[-\pi,\pi]} are

where

Let (t_n)_{n \in \mathbb{Z}}\subset \mathbb{R} be a sequence such that (f_n)_n (defined above) is a Riesz basis for L_2[-\pi,\pi]. If (e_n)_n is the standard exponential orthonormal basis for L_2[-\pi,\pi] and the map L (defined above) satisfies the estimate \Arrowvert I-L \Arrowvert = \delta <1, then the following are true:

1) For l \geq 0 , the sequence (f_{l,n})_n is a Riesz basis for L_2[-\pi,\pi].
2) For l \geq 0, the map L_l defined by L_l e_n = f_{l,n} satisfies \Arrowvert L_l^{-1} \Arrowvert \leq \frac{1}{1-\delta}.

If (t_n)_n \subset \mathbb{R} satisfies the hypotheses of proposition 7.4, then

in PW_{[-\pi,\pi]}.