This content will become publicly available on July 4, 2024
Beginning with the work of Landau, Pollak and Slepian in the 1960s on time‐band limiting, commuting pairs of integral and differential operators have played a key role in signal processing, random matrix theory, and integrable systems. Previously, such pairs were constructed by ad hoc methods, which essentially worked because a commuting operator of low order could be found by a direct calculation. We describe a general approach to these problems that proves that every point
 Award ID(s):
 2131243
 NSFPAR ID:
 10429004
 Publisher / Repository:
 Wiley Blackwell (John Wiley & Sons)
 Date Published:
 Journal Name:
 Communications on Pure and Applied Mathematics
 ISSN:
 00103640
 Format(s):
 Medium: X
 Sponsoring Org:
 National Science Foundation
More Like this

Commuting integral and differential operators connect the topics of signal processing, random matrix theory, and integrable systems. Previously, the construction of such pairs was based on direct calculation and concerned concrete special cases, leaving behind important families such as the operators associated to the rational solutions of the Korteweg–de Vries (KdV) equation. We prove a general theorem that the integral operator associated to every wave function in the infinitedimensional adelic Grassmannian G r a d of Wilson always reflects a differential operator (in the sense of Definition 1 below). This intrinsic property is shown to follow from the symmetries of Grassmannians of Kadomtsev–Petviashvili (KP) wave functions, where the direct commutativity property holds for operators associated to wave functions fixed by Wilson’s sign involution but is violated in general. Based on this result, we prove a second main theorem that the integral operators in the computation of the singular values of the truncated generalized Laplace transforms associated to all bispectral wave functions of rank 1 reflect a differential operator. A 9 0 ○ rotation argument is used to prove a third main theorem that the integral operators in the computation of the singular values of the truncated generalized Fourier transforms associated to all such KP wave functions commute with a differential operator. These methods produce vast collections of integral operators with prolatespheroidal properties, including as special cases the integral operators associated to all rational solutions of the KdV and KP hierarchies considered by [Airault, McKean, and Moser, Commun. Pure Appl. Math. 30, 95–148 (1977)] and [Krichever, Funkcional. Anal. i Priložen. 12, 76–78 (1978)], respectively, in the late 1970s. Many examples are presented.more » « less

Abstract In the mid 1980s it was conjectured that every bispectral meromorphic function {\psi(x,y)} gives rise to an integral operator {K_{\psi}(x,y)} which possesses a commuting differential operator.This has been verified by a direct computation for several families of functions {\psi(x,y)} where the commuting differential operator is oforder {\leq 6} . We prove a general version of this conjecture for all selfadjoint bispectral functions of rank 1and all selfadjoint bispectral Darboux transformations of the rank 2 Bessel and Airy functions.The method is based on a theorem giving an exact estimate of the second and firstorder terms ofthe growth of the Fourier algebra of each such bispectral function. From it we obtaina sharp upper bound on the order of the commuting differential operator for theintegral kernel {K_{\psi}(x,y)} leading to a fast algorithmic procedurefor constructing the differential operator; unlike the previous examples its order is arbitrarily high.We prove that the above classes of bispectral functions are parametrized by infinitedimensionalGrassmannians which are the Lagrangian loci of the Wilson adelic Grassmannian and its analogsin rank 2.more » « less

How could the Fourier and other transforms be naturally discovered if one didn't know how to postulate them? In the case of the Discrete Fourier Transform (DFT), we show how it arises naturally out of analysis of circulant matrices. In particular, the DFT can be derived as the change of basis that simultaneously diagonalizes all circulant matrices. In this way, the DFT arises naturally from a linear algebra question about a set of matrices. Rather than thinking of the DFT as a signal transform, it is more natural to think of it as a single change of basis that renders an entire set of mutuallycommuting matrices into simple, diagonal forms. The DFT can then be ``discovered'' by solving the eigenvalue/eigenvector problem for a special element in that set. A brief outline is given of how this line of thinking can be generalized to families of linear operators, leading to the discovery of the other common Fouriertype transforms, as well as its connections with group representations theory.more » « less

null (Ed.)The monster sporadic group is the automorphism group of a central charge $c=24$ vertex operator algebra (VOA) or meromorphic conformal field theory (CFT). In addition to its $c=24$ stress tensor $T(z)$, this theory contains many other conformal vectors of smaller central charge; for example, it admits $48$ commuting $c=\frac12$ conformal vectors whose sum is $T(z)$. Such decompositions of the stress tensor allow one to construct new CFTs from the monster CFT in a manner analogous to the GoddardKentOlive (GKO) coset method for affine Lie algebras. We use this procedure to produce evidence for the existence of a number of CFTs with sporadic symmetry groups and employ a variety of techniques, including Hecke operators and modular linear differential equations, to compute the characters of these CFTs. Our examples include (extensions of) nine of the sporadic groups appearing as subquotients of the monster, as well as the simple groups ${}^2\tsl{E}_6(2)$ and $\tsl{F}_4(2)$ of Lie type. Many of these examples are naturally associated to McKay's $\widehat{E_8}$ correspondence, and we use the structure of Norton's monstralizer pairs more generally to organize our presentation.more » « less

Coupled partial differential equations (PDEs) are key tasks in modeling the complex dynamics of many physical processes. Recently, neural operators have shown the ability to solve PDEs by learning the integral kernel directly in Fourier/Wavelet space, so the difficulty for solving the coupled PDEs depends on dealing with the coupled mappings between the functions. Towards this end, we propose a coupled multiwavelets neural operator (CMWNO) learning scheme by decoupling the coupled integral kernels during the multiwavelet decomposition and reconstruction procedures in the Wavelet space. The proposed model achieves significantly higher accuracy compared to previous learningbased solvers in solving the coupled PDEs including GrayScott (GS) equations and the nonlocal mean field game (MFG) problem. According to our experimental results, the proposed model exhibits a 2ˆ „ 4ˆ improvement relative L2 error compared to the best results from the stateoftheart models.more » « less