Canonical integral operators on the Fock space II

Xingtang Dong, Kehe Zhu · 2025

Read Canonical integral operators on the Fock space II by Xingtang Dong, Kehe Zhu (2025) 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-2509-16672.txt.

Introduction

Xingtang Dong and Kehe Zhu continue a study of special operators in a 2025 math paper. The operators act on a space of functions on the complex plane. They include classic transforms from physics as special cases. The pair ask when these operators stay bounded and when they shrink sets in a strong sense.

They show that good behavior on a square-based space matches behavior on all related spaces. Their work rests on close study of the operator's kernel. Side results cover averaged forms of the operators. The findings round out a theory used in signal work and quantum models, where such transforms move functions around.

Catalog record

CREATORXingtang Dong, Kehe Zhu
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE62 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2509-16672

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Canonical integral operators on the Fock space II Thanks: Research supported by NNSF of China (Grant Numbers 12271396 and 12271328). Xingtang Dong, Kehe Zhu Abstract . In [ 2 ] we introduced and studied a two-parameter family of integral operators T ( s , t ) T^{(s,t)} on the Fock space F 2 F^{2} of the complex plane. Under the inverse Bargmann transform, these operators include the classical linear canonical transforms in mathematical physics as special cases, so we called T ( s , t ) T^{(s,t)} canonical linear operators on the Fock space. In this paper we continue the study of these operators. We show that when a canonical linear operator T ( s , t ) T^{(s,t)} is compact, it actually belongs to the Schatten class S p S_{p} for all p > 0 p>0 . In this case, we find all singular values, determine the S p S_{p} norm, and obtain a trace formula for T ( s , t ) T^{(s,t)} . We also show that the boundedness (and a natural version of compactness) of T ( s , t ) T^{(s,t)} on F p F^{p} for any given p ∈ ( 0 , ∞ ] p\in(0,\infty] is equivalent to the boundedness (and compactness) of T ( s , t ) T^{(s,t)} on F 2 F^{2} . Our analysis is based on estimates and computations with the integral kernel of T ( s , t ) T^{(s,t)} , which also yield some interesting results about the Berezin transform and the bivariate Berezin transform of T ( s , t ) T^{(s,t)} . 2020 Mathematics Subject Classification Primary 30H20; Secondary 47G10. 1. Introduction Let {\mathbb{C}} be the complex plane and H({\mathbb{C}}) be the linear space of all entire functions on {\mathbb{C}} . For 0<p\leq\infty we use F^{p} to denote the space of all f\in H({\mathbb{C}}) such that the function f(z)e^{-|z|^{2}/2} belongs to L^{p}({\mathbb{C}},dA) , where dA is ordinary area measure on the complex plane. For f\in F^{p} with 0<p<\infty we write \|f\|_{p}^{p}=\frac{p}{2\pi}\int_{{\mathbb{C}}}\left|f(z)e^{-\frac{|z|^{2}}{2}}\right|^{p}\,dA(z). For f\in F^{\infty} we define \|f\|_{\infty}=\sup\left\{|f(z)|e^{-\frac{|z|^{2}}{2}}:z\in{\mathbb{C}}\right\}. The spaces F^{p} are called Fock spaces. When p=2 , F^{2} is a Hilbert space whose inner product is given by \langle f,g\rangle=\int_{{\mathbb{C}}}f(z)\overline{g(z)}\,d\lambda(z), where d\lambda(z)=\frac{1}{\pi}e^{-{|z|^{2}}}\,dA(z) is the standard Gaussian measure on {\mathbb{C}} . Furthermore, F^{2} is a reproducing kernel Hilbert space whose kernel function is K(z,w)=e^{z\overline{w}} . See [ 6 ] for an introduction to the theory of Fock spaces. Recently in [ 2 ] we introduced and studied a two-parameter family of integral operators on L^{2}({\mathbb{C}},d\lambda) and on F^{2} . More specifically, for any (s,t)\in{\mathbb{C}}^{2} with s\not=0 , we define T^{(s,t)}f(z)=\int_{{\mathbb{C}}}K^{(s,t)}(z,w)f(w)\,d\lambda(w), (1) where K^{(s,t)}(z,w)=\frac{1}{\sqrt{s}}\,\exp\left[\frac{tz^{2}-\overline{tw^{2}}+2z\overline{w}}{2s}\right]. Here and throughout the paper, the complex square root is defined as follows: \sqrt{z}=\sqrt{|z|}\,e^{i\theta/2},\qquad z=|z|e^{i\theta},\qquad\theta\in(-\pi,\pi]. Note that K^{(1,0)}(z,w)=e^{z\overline{w}} is the reproducing kernel of F^{2} , and T^{(1,0)}=P is the orthogonal projection from L^{2}({\mathbb{C}},d\lambda) onto F^{2} . One of the main results of [ 2 ] is that T^{(s,t)} is bounded on F^{2} if and only if |s|^{2}\geq|t|^{2}+1 . Furthermore, |s|^{2}=|t|^{2}+1 if and only if T^{(s,t)} is unitary, and |s|^{2}>|t|^{2}+1 if and only if T^{(s,t)} is compact. Thus when T^{(s,t)} is bounded, it is either unitary or compact. Most of [ 2 ] was devoted to the study of T^{(s,t)} when they are unitary. In particular, it was shown in [ 2 ] that, when |s|^{2}=|t|^{2}+1 , the operators T^{(s,t)} on F^{2} are unitarily equivalent to the classical linear canonical transforms on L^{2}({\mathbb{R}},dx) , which are widely used in mathematical physics and engineering applications (see [ 3 ] for example). For this reason we called the two-parameter family of integral operators T^{(s,t)} canonical integral operators on the Fock space. In this paper, we continue this program and study T^{(s,t)} when they are compact. The most important properties of a compact operator are its singular values, which will be our main focus here. The main results of this paper are Theorems A and B below. Theorem A . Let p>0 and (s,t)\in{\mathbb{C}}\times{\mathbb{C}} with |s|^{2}>|t|^{2}+1 . Then the compact operator T^{(s,t)}:F^{2}\to F^{2} belongs to the Schatten class S_{p} . Moreover, (a) The decreasing sequence \left\{\mu_{n}\right\}_{n=0}^{\infty} , where \mu_{n}=\left[\frac{2|s|}{|s|^{2}-|t|^{2}-1+\sqrt{(|s|^{2}-|t|^{2}-1)^{2}+4|s|^{2}}}\right]^{n+(1/2)}, is the sequence of all singular values for T^{(s,t)} . (b) The operator norm and Schatten p -norm of T^{(s,t)} are given by \|T^{(s,t)}\|=\mu_{0},\qquad\|T^{(s,t)}\|_{S^{p}}=\mu_{0}\,\left(1-\mu_{0}^{2p}\right)^{-(1/p)}. (c) The trace of T^{(s,t)} is given by \tr\left(T^{(s,t)}\right)=\frac{1}{\sqrt{s}}\sqrt{\frac{s^{2}}{(s-1)^{2}+|t|^{2}}}. For any fixed w\in{\mathbb{C}} we write K_{w}^{(s,t)}(z)=K^{(s,t)}(z,w),\qquad z\in{\mathbb{C}}. It follows from Lemma 1 of [ 2 ] that the function K_{w}^{(s,t)} belongs to F^{2} if and only if |s|>|t| . Therefore, for |s|>|t| , the integral in ( 1 ) gives rise to an entire function T^{(s,t)}f for every f\in F^{2} . In fact, it is not difficult to see that T^{(s,t)}f is a well-defined entire function for every f\in F^{\infty} (and hence for every f\in F^{p} with p>0 ). Since T^{(s,t)}f is a well-defined entire function for every f\in F^{p} when 0<p\leq\infty and |s|>|t| , it is natural for us to consider the boundedness and “compactness” of T^{(s,t)} on F^{p} . There is a minor technical issue we must address, that is, the notion of “compact operators” on the non-Banach spaces F^{p} for 0<p<1 and the non-reflexive Banach spaces F^{1} and F^{\infty} . Thus in the rest of the paper, for convenience and simplicity of terminology, we will say that a bounded linear operator T:F^{p}\to F^{p} , where 0<p\leq\infty , is compact if \|Tf_{n}\|_{p}\to 0 (as n\to\infty ) for every bounded sequence \{f_{n}\}\subset F^{p} that converges to 0 uniformly on compact subsets of {\mathbb{C}} . When 1<p<\infty , this definition of compactness is equivalent to the usual notion of compactness on the Banach space F^{p} . Theorem B . Suppose |s|>|t| and 0<p\leq\infty . Then (a) T^{(s,t)} is bounded on F^{p} if and only if |s|^{2}\geq|t|^{2}+1 ; (b) T^{(s,t)} is compact on F^{p} if and only if |s|^{2}>|t|^{2}+1 . Consequently, the boundedness (and compactness) of T^{(s,t)} on F^{p} is actually independent of p , where 0<p\leq\infty . 2. Estimates for canonical integral kernels For any w\in{\mathbb{C}} we use K_{w} to denote the reproducing kernel of F^{2} at w . Thus K_{w}(z)=K(z,w)=e^{z\overline{w}} . The function k_{w}=K_{w}/\|K_{w}\|_{2} is called the normalized reproducing kernel of F^{2} at w . Each k_{w} is a unit vector in F^{2} ; it is also a unit vector in F^{p} for any 0<p\leq\infty . See [ 6 ] . We begin with the following integral representation for general operators on the Fock space. Proposition 1 . If T is a bounded linear operator on F^{2} , then there exists a unique function H(z,w) on {\mathbb{C}}\times{\mathbb{C}} with the following properties. (i) For fixed z,w\in{\mathbb{C}} the functions u\mapsto H(u,w) and u\mapsto\overline{H(z,u)} are both in F^{2} . (ii) The operator T is represented as Tf(z)=\int_{{\mathbb{C}}}H(z,w)f(w)\,d\lambda(w),\qquad f\in F^{2},\quad z\in{\mathbb{C}}. (2) Proof. Note that for any f\in F^{2} and z\in{\mathbb{C}} we have \displaystyle Tf(z) \displaystyle=\langle Tf,K_{z}\rangle=\langle f,T^{*}K_{z}\rangle \displaystyle=\int_{{\mathbb{C}}}\overline{(T^{*}K_{z})(w)}f(w)\,d\lambda(w) \displaystyle=\int_{{\mathbb{C}}}H(z,w)f(w)\,d\lambda(w), where \displaystyle H(z,w) \displaystyle=\overline{(T^{*}K_{z})(w)}=\overline{\langle T^{*}K_{z},K_{w}\rangle}=\langle K_{w},T^{*}K_{z}\rangle \displaystyle=\langle TK_{w},K_{z}\rangle=(TK_{w})(z). For any fixed w the function z\mapsto H(z,w) is simply TK_{w} in F^{2} . Similarly, for any fixed z\in{\mathbb{C}} , the function w\mapsto\overline{H(z,w)} is simply T^{*}K_{z} in F^{2} . On the other hand, if T admits the integral representation in ( 2 ) with H(z,w) satisfying the conditions in (i), then for any w,z\in{\mathbb{C}} we have \displaystyle(TK_{w})(z) \displaystyle=\int_{{\mathbb{C}}}H(z,u)K_{w}(u)\,d\lambda(u) \displaystyle=\overline{\int_{{\mathbb{C}}}\overline{H(z,u)}K(w,u)\,d\lambda(u)} \displaystyle=\overline{\overline{H(z,w)}}=H(z,w),

Read the complete text