On the structure of the Gram matrix for Gabor systems generated by B-splines.
Martin Buck, Christina Frederick, Kasso A. Okoudjou, Alexander Stangl · 2026
Read On the structure of the Gram matrix for Gabor systems generated by B-splines. by Martin Buck, Christina Frederick, Kasso A. Okoudjou, Alexander Stangl (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-16986.txt.
Introduction
Buck and co-authors study the Gram matrix of Gabor systems built from B-splines. A Gabor system tiles the time frequency plane with shifts of one window function over a lattice. They show that, with the right ordering, certain submatrices of the Gram matrix have a block Toeplitz structure. That structure lets them import spectral results from the theory of Toeplitz matrices. They derive spectral facts for finite sub blocks of the Gram matrix. They then apply the results to systems generated by B-splines of any order.
Catalog record
| CREATOR | Martin Buck, Christina Frederick, Kasso A. Okoudjou, Alexander Stangl |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 48695 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.16986
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
On the structure of the Gram matrix for Gabor systems generated by B-splines. Martin Buck, Christina Frederick, Kasso A. Okoudjou, Alexander Stangl Abstract We consider the Gabor system 𝒢 ( g , a ℤ × b ℤ ) \mathcal{G}(g,a\mathbb{Z}\times b\mathbb{Z}) generated by a continuous, compactly supported function g g over the time-frequency lattice generated by the parameters a a and b b . We show that, under an appropriate ordering of the Gabor elements, certain submatrices of the Gram matrix of 𝒢 ( g , a ℤ × b ℤ ) \mathcal{G}(g,a\mathbb{Z}\times b\mathbb{Z}) exhibit a block-Toeplitz structure. This structural property enables us to derive spectral results for finite sub-blocks of the Gram matrix by appealing to the spectral theory of Toeplitz matrices. In particular, we apply our results to the Gram matrix of Gabor systems generated by the N N th-order B-spline. 1 Introduction 1.1 Background on the frame set problem Gabor analysis originated from two distinct perspectives: the operator-theoretic framework of J. von Neumann 44 in 1932 and the communication-theoretic approach of D. Gabor 20 in 1946. Both authors independently investigated the completeness of the system \mathcal{G}(g,\mathbb{Z}^{2})=\{g_{(n,k)}(\cdot)=g(\cdot-n)e^{2\pi ik\cdot}:(n,k)\in\mathbb{Z}^{2}\}, (1) utilizing the Gaussian g(x)=e^{-\pi x^{2}} for its optimal localization in the time-frequency plane. This dual legacy highlights the fundamental nature of the integer lattice \mathbb{Z}^{2} in tiling phase space (see 34 for a detailed history). More generally, the Gabor system generated by a non-zero window function g\in L^{2}(\mathbb{R}) along a (countable) discrete set of time-frequency shifts \Gamma\subset\mathbb{R}^{2} , \mathcal{G}(g,\Gamma)=\{g_{\gamma}:=g(\cdot-\gamma_{1})e^{2\pi i\gamma_{2}\cdot}:\,\gamma=(\gamma_{1},\gamma_{2})\in\Gamma\}, (2) is said to be a frame provided that there exist constants 0<\alpha\leq\beta<\infty such that for each f\in L^{2}(\mathbb{R}) \alpha\left\lVert f\right\rVert^{2}\leq\sum_{\gamma\in\Gamma}\left\lvert\left\langle f,g_{\gamma}\right\rangle\right\rvert^{2}\leq\beta\left\lVert f\right\rVert^{2}. (3) It follows that there exists an associated set of functions \left\{h_{\gamma}\right\}_{\gamma\in\Gamma} which is also a Gabor frame for L^{2}(\mathbb{R}) for which the following reconstruction formulas hold for each f\in L^{2}(\mathbb{R}) : f=\sum_{\gamma\in\Gamma}\left\langle f,h_{\gamma}\right\rangle g_{\gamma}=\sum_{\gamma\in\Gamma}\left\langle f,g_{\gamma}\right\rangle h_{\gamma}. (4) When \Gamma=a\mathbb{Z}\times b\mathbb{Z} is a separable lattice, it can be shown that the set of functions \left\{h_{\gamma}\right\}_{\gamma\in\Gamma} is a Gabor frame generated by a window function h\in L^{2}(\mathbb{R}) , and called a dual window to g . In this case, \left\{h_{\gamma}\right\}_{\gamma\in\Gamma} is a dual Gabor frame to \left\{g_{\gamma}\right\}_{\gamma\in\Gamma} The full understanding of all window functions g , and all positive parameters a,b such that \mathcal{G}(g,a\mathbb{Z}\times b\mathbb{Z}) is a Gabor frame for L^{2}(\mathbb{R}) , is still incomplete. Nonetheless, the last few decades of research have resulted in an impressive body of work on Gabor systems. For example, there exists a Nyquist-type condition dating back to the work of Daubechies 16 , in which it is proved that when ab>1 is rational then \mathcal{G}(g,a\mathbb{Z}\times b\mathbb{Z}) is incomplete, for the irrational case we refer to 4 . Furthermore, if \mathcal{G}(g,a\mathbb{Z}\times b\mathbb{Z}) is a Gabor frame for L^{2}(\mathbb{R}) then ab\leq 1 . The critical case ab=1 corresponds to the case when it is possible for \mathcal{G}(g,a\mathbb{Z}\times b\mathbb{Z}) to be an orthonormal or a Riesz basis. In such a case, the window function is necessarily poorly localized in the time-frequency plane as dictated by the Balian-Low Theorem 5 . We refer to 34 for an history of the density condition in Gabor analysis. We also refer to some of the foundational work of I. Daubechies, who in the setting of coherent state frames in the Bargmann-Fock conjectured that a Gabor system G(g,a,b) on the rectangular lattice a\mathbb{Z}\times b\mathbb{Z} is a frame if and only if ab<1 for the Gaussian, 18 . Still, it is in general unknown how to characterize the so-called full frame set \mathcal{F}(g) of a given window g\in L^{2}(\mathbb{R}) \mathcal{F}(g)=\left\{\;\Gamma\subset\mathbb{R}^{2}:\mathcal{G}(g,\Gamma)=\left\{g_{(\gamma_{1},\gamma_{2})}\right\}_{(\gamma_{1},\gamma_{2})\in\Gamma}\text{ is a frame over }L^{2}(\mathbb{R})\;\right\}. (5) Restricting \Gamma to the lattice \Lambda=a\mathbb{Z}\times b\mathbb{Z} for positive constants a,b , we arrive at what is referred to as the frame set problem . Insights offered by this variant of the full frame set problem are hoped to enable further investigation of the more general problem. Celebrated results in the early 90s characterized the frame set for the Gaussian window as the open region \{(a,b)\in\mathbb{R}^{2}_{+}:\,0<ab<1\} ; see 42 ; 41 ; 39 . More recently, a similar result was obtained for classes of functions, including totally positive functions, 27 ; 36 ; 26 ; 29 . For an overview of the state of this fascinating problem we refer to 11 ; 28 . For some historical perspective, we refer to the work of Daubechies, Grossmann and Meyer, 17 , on painless non-orthogonal expansions from which one concludes that the rectangle (0,2)\times(0,1/2) belongs to the frame set for many window functions g . For g\in L^{2}(\mathbb{R}),g\neq 0 , and f\in L^{2}(\mathbb{R}) , let V_{g}f be the function defined on \mathbb{R}^{2} by V_{g}f(x,\omega)=\int_{\mathbb{R}}f(t)\overline{g(t-x)}e^{-2\pi it\omega}\,dt. The modulation space M^{1}(\mathbb{R}) , also known as the Feichtinger algebra S_{0}(\mathbb{R}) , is the dense Banach subspace of L^{2}(\mathbb{R}) consisting of all functions g such that V_{g}g\in L^{1}(\mathbb{R}^{2}) 6 ; 25 . It was proved in 19 that if the window function g is in the modulation space M^{1}(\mathbb{R}) then its frameset \mathcal{F}(g) is an open subset of \{(a,b)\in\mathbb{R}^{2}_{+}:\,0<ab<1\} . Examples of functions in M^{1}(\mathbb{R}) are the B-splines of order N=2 or higher, which are even, compactly supported functions defined recursively as s_{1}(x)=\chi_{[-1/2,1/2]}(x),\qquad s_{N}(x)=s_{1}(x)*s_{N-1}(x). (6) They are of special interest to study for a number of reasons: B-spline wavelets, which have a construction similar to Gabor systems, have been enormously successful in finite element methods 46 ; 35 ; 47 ; 45 . Gabor systems corresponding to B-spline window functions have also shown promise in acoustic scattering problems 37 . The frame set for the first order B-spline s_{1} was completely characterized in 15 , and work on higher order B-spline has gradually accumulated in the past decade (see 3 2 11 10 12 22 21 24 38 36 for details). As noted above, \mathcal{F}(s_{2}) is an open set in \{(a,b)\in\mathbb{R}^{2}_{+}:\,0<ab<1\} 19 , but a full characterization of this set remains an open question. The authors of 3 introduced a framework for determining the frame sets of compactly supported functions, including the B-spline of order N\geq 2 , that unified many known results on the frame sets of B-spline. Similarly, the authors of 2 used a similar linear algebra based approach to shed new light on the set \mathcal{F}(s_{2}) and enlarged the known frame set region by deriving sufficient conditions for invertibility of the linear system relating a frame with its dual. The key idea is based on the following result (see ( 13 , Theorem 2.10) ): for g,h\in L^{2}(\mathbb{R}) and a,b>0 , the Bessel sequences \mathcal{G}(g,a\mathbb{Z}\times b\mathbb{Z}) and \mathcal{G}(h,a\mathbb{Z}\times b\mathbb{Z}) are dual frames for L^{2}(\mathbb{R}) if and only if \sum_{k\in\mathbb{Z}}\overline{g(x-\ell/b+ka)}h(x+ka)=b\delta_{\ell,0},\forall\ell\in\mathbb{Z}\mbox{ and a.e. }x\in\left[-\frac{a}{2},\frac{a}{2}\right]. (7) The authors of 2 showed the existence of a bounded compactly supported function h\in L^{2}(\mathbb{R}) that solves ( 7 ) when (a,b) belongs to a family of sets parameterized by m . To do this, the authors simplified and rewrote ( 7 ) as a matrix-vector equation using the compact support of s_{2} and derived sufficient conditions for invertibility of the system. 1.2 Contributions This paper aims to investigate the frameset of continuous, compactly supported functions. To this end, we initiate an analysis of the Gram matrix of the corresponding Gabor systems. We build upon the structural investigations introduced