Chromatic expansion with Bessel operator of fractional order
M. Chegaar and Á. P. Horváth · 2026
Read Chromatic expansion with Bessel operator of fractional order by M. Chegaar and Á. P. Horváth (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-2606-26737.txt.
Introduction
M. Chegaar and Á. P. Horváth build an expansion framework around fractional powers of the Bessel operator. Many problems have radial or spherical shape rather than a grid shape. Optics, acoustics, and radial signal work are examples. Bessel functions are the natural wave shapes for such geometry. Chromatic expansions express a function through its derivative-like samples at chosen points.
The authors define Bessel-chromatic derivatives from the spectral form of the fractional operator. They apply them to weighted spherical means, averages over spheres, at the centre as well as at general points. Several weight classes on finite and infinite ranges are treated. Convergence is studied through classical averaging methods. In the band-limited case, the method reconstructs the spherical means and, under stated conditions, the original function itself. Numerical examples check the accuracy.
Catalog record
| CREATOR | M. Chegaar and Á. P. Horváth |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 68 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2606.26737
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Chromatic expansion with Bessel operator of fractional order M. Chegaar and Á. P. Horváth Abstract . This paper develops a Bessel-chromatic expansion framework associated with fractional powers of the Bessel-Laplace operator. The construction combines methods of weighted polynomial approximation and of fractional differential operators. Using the spectral representation of ( − Δ a ) 1 2 (-\Delta_{a})^{\frac{1}{2}} , we define Bessel-chromatic derivatives and apply them to weighted spherical means both at a general point and at the origin. Different classes of weights on finite and infinite intervals are considered, with particular attention to cases where the inverse Hankel transform is explicit. The convergence of the expansions is studied through Cesàro and de la Vallée Poussin means. In the bandlimited case, the method gives reconstruction formulas for weighted spherical means and, under suitable assumptions, recovery formulas for the original function. Numerical examples illustrate the decay of the Bessel-chromatic coefficients and the accuracy of the corresponding reconstructions. 2020 Mathematics Subject Classification 41A25, 42C10, 44A15, 47G30 1. Introduction Many problems in analysis and applied mathematics involve data whose natural geometry is not Cartesian, but radial, cylindrical, or spherical. This type of structure appears in radial signal processing, numerical differentiation of oscillatory data, wave propagation in symmetric media. With an external field the ordinary Fourier transform is not always the most convenient spectral tool. The Hankel transform, together with Bessel-type differential operators, is often more suitable because it reflects the geometry of the problem more directly; see, for example, [ 29 , 25 , 33 ] . A classical tool for reconstructing bandlimited functions is the Whittaker-Shannon-Kotel’nikov sampling theorem. In its simplest form, under the usual normalization and bandlimiting assumptions, it gives f(t)=\sum_{n\in\mathbb{Z}}f(n)\frac{\sin\pi(t-n)}{\pi(t-n)}. This formula is fundamental in signal processing. However, in numerical applications it may require many samples, and the convergence can be slow when only finite information is available. Taylor expansion has a different advantage: it uses local differential information at one point. But this is also its limitation, since many derivatives may be needed to approximate a function on an interval. Chromatic expansions, introduced by Ignjatović [ 9 ] , offer a useful middle ground. They combine differential information in the physical variable with an orthogonal expansion in the spectral variable. Further developments and applications of chromatic derivatives and chromatic expansions can be found in [ 10 , 11 , 12 , 13 , 14 , 35 , 34 , 8 ] . The main idea can be seen first in the Fourier setting. If p is a polynomial and the inverse Fourier transform is available, then formally p\left(-i\frac{\partial}{\partial x}\right)f(x)=c\int_{\mathbb{R}}p(y)\hat{f}e^{ixy}dy. That is if \mathrm{supp}\hat{f}\subset[a,b] , and \{p_{k}\} are the orthonormal polynomials on [a,b] ; p_{k}\left(-i\frac{\partial}{\partial x}\right)f(0)=cc_{k}(\hat{f}) , i.e. the Fourier coefficients of \hat{f} with respect to \{p_{k}\} . Then the inverse Fourier transform of the series gives the chromatic expansion of f . Chromatic expansions have proven to be useful tool for rapid recovery of functions as well as for numerical differentiation, see e.g. [ 10 , 12 ] . Investigation of the underlying spaces is also important, see e.g. [ 11 ] . Multidimensional extension can be found e.g. in [ 14 ] , and the problem is discussed in case of generalized functions in [ 34 ] . The questions of different weighted cases and different integral transforms naturally arose and were examined in numerous papers, see e.g. [ 35 , 8 ] . The relationship of wavelets and chromatic expansions is discussed in [ 8 ] . For the summary of recent results, see [ 13 ] . In this paper we use the Bessel differential operator, which in one dimension is B_{\alpha}:=\frac{\partial^{2}}{\partial x^{2}}+\frac{2\alpha+1}{x}\frac{\partial}{\partial x}, and the corresponding Hankel transform, \mathcal{H}_{\alpha} . Since \mathcal{H}_{\alpha}((-B_{\alpha}f))(y)=y^{2}\mathcal{H}_{\alpha}(f)(y), in Bessel setup one can choose the entire Bessel functions, j_{\alpha}(\sqrt{\lambda}x) as eigenfunction of the differential operator and so the kernel of the integral operator, cf. e.g. [ 33 ] , or taking the square-root of the -B_{\alpha} operator, one can arrive to a situation which is very similar to the sketched one. Here we choose the second solution. In several variables we work on the positive octant \mathbb{R}^{d}_{+} with the Bessel-Laplace operator \Delta_{a}=\sum_{i=1}^{d}B_{\alpha_{i}},\qquad a_{i}=2\alpha_{i}+1, and with the corresponding d -dimensional Hankel transform. This choice is motivated by the intention to provide estimates of the of convergence rate of chromatic expansions. At first we mention, that similarly to the Fourier case, a Shannon-type sampling formula can be derived (see e.g. [ 16 ] and [ 15 ] ), but it possesses rather poor convergence properties. We give a brief description of it in the Appendix. On the other hand, since Bessel and Hankel operators act on functions defined on the positive octant, the functions under study can be extended evenly, and the investigation can be transferred to a symmetric interval with symmetric weight. Our estimations on the rate of convergence are based on the smoothness properties of the functions on the image side. Since smoothness can be lost during even extension, this method generally results in weaker convergence. Below we define spherical Bessel chromatic expansion on the positive octant of the d -dimensional space. It approximates the weighted spherical mean of a function. In some cases the original function can be recovered as well. To define chromatic derivatives and series we need some facts about Bessel translation, fractional Bessel operators and Hankel transform. These basic facts are contained in the second section. The main goal of the investigation is to estimate the speed of convergence. To this, we use some results of weighted polynomial approximation. Although, as mentioned in the remarks at the end of the sections, our method and theorems can be extended to general classes of weights, we focus on weights with explicit inverse Hankel transform. It is known that the weighted spherical mean is the transmutation operator intertwining the d -dimensional and 1 -dimensional Bessel-Laplace operators, see [ 29 ] . In the third section we define spherical Bessel chromatic derivatives by the extension of this theorem to fractional operators. In the unweighted bandlimited case the original function can also be recovered from spherical expansion. In the fourth section, considering the weighted spherical mean at zero, we simplify the method. Although the possibility of global reconstruction is lost, the method has wide applicability as in the unweighted (Laplace) case, see e.g. [ 5 ] . We mention here, that the d -dimensional Laplace and Bessel-Laplace operators are closely related, see e.g. [ 3 ] . Thus, the results of this section can be transferred to the square root of the Laplace operator and the Fourier transform. Of course, in this case, it is more natural to consider -i\frac{\partial}{\partial x} instead of (-\Delta)^{\frac{1}{2}} . A short fifth section has also been added, which contains the method of approximation of functions with even entire functions of exponential type \nu . Here we gave an example to the joint application of the different approximations. Finally, the Appendix, besides Shannon sampling, contains some numerical illustrations of Bessel chromatic expansion. 2. Notation, Preliminaries 2.1. Spaces We introduce the next weighted spaces. On \mathbb{R}^{d}_{+}:=\{(x_{1},\dots,x_{d}):x_{i}>0\hskip 4.0pti=1,\dots,d\} we introduce the L^{p}_{a}(\mathbb{R}^{d}_{+}) space, 1\leq p<\infty with norm (1) \|f\|_{p,a}^{p}=\int_{\mathbb{R}^{d}_{+}}|f(x)|^{p}x^{a}dx, where a=(a_{1},\dots,a_{d}) , a_{i}>0 , i=1,\dots,d ; x^{a}=\prod_{i=1}^{d}x_{i}^{a_{i}} . For p=\infty