Convolutions of Orlicz spaces and Orlicz Schatten classes, with applications to Toeplitz operators

Wolfram Bauer, Robert Fulsche, Joachim Toft · 2025

Read Convolutions of Orlicz spaces and Orlicz Schatten classes, with applications to Toeplitz operators by Wolfram Bauer, Robert Fulsche, Joachim Toft (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-2505-01707.txt.

Introduction

Function spaces named after Orlicz stretch the classic scale used to sort functions by size. Wolfram Bauer, Robert Fulsche, and Joachim Toft study how such spaces behave under convolution in a 2025 paper. Convolution blends two functions into one, and knowing which space the blend lands in is a core tool of analysis.

The authors prove blend rules of Young type for classes tied to a broad family of operators. They then turn those rules on a classic operator family from physics and complex analysis. The payoff is a set of sharp conditions telling when those operators fall into the graded classes that mark how close they sit to finite size. Such results feed work on signal models and quantum-style operator theory.

Catalog record

CREATORWolfram Bauer, Robert Fulsche, Joachim Toft
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE127 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-01707

Browse the full library · About the Tower of Babel

Full text

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

Convolutions of Orlicz spaces and Orlicz Schatten classes, with applications to Toeplitz operators Wolfram Bauer, Robert Fulsche, Joachim Toft Abstract . Let Φ \Phi be a Young function. We study convolution properties for symbol classes s A , Φ s_{A,\Phi} , which consist of all a a such that the pseudo-differential operator Op A ⁡ ( a ) \operatorname{Op}_{A}(a) is in the Orlicz Schatten class ℐ Φ ​ ( L 2 ​ ( 𝐑 d ) ) \mathscr{I}_{\Phi}(L^{2}(\mathbf{R}^{d})) . Especially we prove Young type results for such classes. We apply the results on Toeplitz operators and prove Orlicz Schatten properties of such operators. 2010 Mathematics Subject Classification Primary: 46E30, 35S05, 47G30 Secondary: 47B10,47B35 0. Introduction In the paper we deduce continuity properties for convolutions between Orlicz spaces and Orlicz type Schatten-von Neumann symbol classes in pseudo-differential calculus. As special cases, we obtain well-known convolution relations between Lebesgue spaces and classical Schatten-von Neumann classes, in quantum harmonic analysis, given in [ 21 , Proposition 3.2] , by R. Werner. (See also [ 17 ] for an alternative approach, whose style is similar to the present paper.) The Schatten-von Neumann theory offers detailed studies of linear and compact operators acting on a Hilbert space. For example, consider a linear operator T acting on a separable Hilbert space \mathcal{H} . Then T is compact, if and only if there is an orthonormal basis \{f_{j}\}_{j=1}^{\infty} of eigenvectors to |T| , whose corresponding non-negative and non-increasing eigenvalues \lambda(T)=\{\lambda_{j}(T)\}_{j=1}^{\infty} tend to zero as j tends to infinity, i. e. \lim_{j\to\infty}\lambda_{j}(T)=0. (0.1) (For notations, see [ 9 , 16 ] and Section 1 .) The operator T belongs to \mathscr{I}_{p}(\mathcal{H}) , the set of Schatten-von Neumann operators on \mathcal{H} of order p\in(0,\infty) , if and only if \|T\|_{\mathscr{I}_{p}}\equiv\|\lambda(T)\|_{\ell^{p}} is finite. It follows that \lambda_{j}(T)\lesssim j^{-\frac{1}{p}} when T\in\mathscr{I}_{p}(\mathcal{H}) , which indeed gives more concrete information of \lambda_{j} for large j , compared to ( 0.1 ). We also observe that the smaller p , the faster \lambda_{j}(T) are imposed to approach 0 at infinity. This also indicates that by pushing p closer to zero, the class \mathscr{I}_{p}(\mathcal{H}) becomes closer to the set of finite-rank operators on \mathcal{H} . The presentation of Schatten-von Neumann classes in [ 16 ] , by B. Simon, is not limited to \ell^{p} norm estimates on the eigenvalues \lambda_{j} above. Instead, any suitable (quasi-)norm on sequences can be used. In this paper we focus on the case when \{\lambda_{j}(T)\}_{j=1}^{\infty} belongs to (quasi-Banach) Orlicz spaces. We recall that any Young function (or even more general, quasi-Young function) \Phi delivers the Orlicz space \ell^{\Phi}(\mathbf{Z}_{+}) , and that for p\in(0,\infty] one has \displaystyle\ell^{p}(\mathbf{Z}_{+}) \displaystyle=\ell^{\Phi_{[p]}}(\mathbf{Z}_{+})\quad\text{when} (0.2) \displaystyle\Phi_{[p]}(t) \displaystyle=t^{p},\ p<\infty,\quad\text{and}\quad\Phi_{[\infty]}(t)=\begin{cases}0,&t\leq 1,\\[4.30554pt] \infty,&t>1\end{cases} (also in norms). Here \mathbf{Z}_{+} is the set of positive integers. The Schatten-von Neumann class \mathscr{I}_{\Phi}(\mathcal{H}) consists of all T above such that \|T\|_{\mathscr{I}_{\Phi}}\equiv\|\lambda(T)\|_{\ell^{\Phi}} (0.3) is finite. Here the norm on the right-hand side is given in Definition 1.3 in Section 1 . Evidently, ( 0.2 ) gives \mathscr{I}_{p}=\mathscr{I}_{\Phi_{[p]}} , so we regain the usual Schatten-von Neumann classes by choosing the Young functions in such ways. By straight-forward applications of Schwartz’ kernel theorem and Fourier’s inversion formula, we may identify the spaces \mathscr{I}_{\Phi}(L^{2}(\mathbf{R}^{d})) with certain function and distribution spaces of operator kernels, or with symbols to pseudo-differential operators. The Weyl calculus within the theory of pseudo-differential operators is here of special interest as it is often suitable in quantizations. For any a in the Schwartz class \mathscr{S}(\mathbf{R}^{2d}) , the Weyl operator \operatorname{Op}^{w}(a) is the linear and continuous operator on \mathscr{S}(\mathbf{R}^{d}) , given by (\operatorname{Op}^{w}(a)f)(x)=(2\pi)^{-d}\iint_{\mathbf{R}^{2d}}a({\textstyle{\frac{1}{2}}}(x+y),\xi)f(y)e^{i\langle x-y,\xi\rangle}\,dyd\xi,\quad f\in\mathscr{S}(\mathbf{R}^{d}). The definition extends to any a\in\mathscr{S}^{\prime}(\mathbf{R}^{2d}) , in which case \operatorname{Op}^{w}(a) is continuous from \mathscr{S}(\mathbf{R}^{d}) to \mathscr{S}^{\prime}(\mathbf{R}^{d}) . In fact, by Schwartz’ kernel theorem and Fourier’s inversion formula it follows that the map a\mapsto\operatorname{Op}^{w}(a) is a bijection from \mathscr{S}^{\prime}(\mathbf{R}^{2d}) to the set of linear and continuous operators from \mathscr{S}(\mathbf{R}^{d}) to \mathscr{S}^{\prime}(\mathbf{R}^{d}) . In particular it follows that \mathscr{I}_{\Phi}(L^{2}(\mathbf{R}^{d})) can be identified with the Banach space of distributions s_{\Phi}^{w}(\mathbf{R}^{2d})\equiv\{\,a\in\mathscr{S}^{\prime}(\mathbf{R}^{2d})\,;\,\operatorname{Op}^{w}(a)\in\mathscr{I}_{\Phi}(L^{2}(\mathbf{R}^{d}))\,\}, with norm \|a\|_{s_{\Phi}^{w}}\equiv(2\pi)^{\frac{d}{2}}\|\operatorname{Op}^{w}(a)\|_{\mathscr{I}_{\Phi}}. (0.4) Then \mathscr{I}_{p}(L^{2}(\mathbf{R}^{d})) and its norm, correspond to s_{p}^{w}(\mathbf{R}^{2d})\equiv s_{\Phi_{[p]}}^{w}(\mathbf{R}^{2d})=\{\,a\in\mathscr{S}^{\prime}(\mathbf{R}^{2d})\,;\,\operatorname{Op}^{w}(a)\in\mathscr{I}_{p}(L^{2}(\mathbf{R}^{d}))\,\}, and \|a\|_{s_{p}^{w}}\equiv(2\pi)^{\frac{d}{2}}\|\operatorname{Op}^{w}(a)\|_{\mathscr{I}_{p}}, (0.5) respectively. We have included the factor (2\pi)^{\frac{d}{2}} on the right-hand sides of ( 0.4 ) and ( 0.5 ) because then one has \|a\|_{s_{2}^{w}}=\|a\|_{L^{2}} (see e. g. [ 17 ] ). The Young type convolution properties established by R. Werner in [ 21 , Proposition 3.2] are then equivalent to s_{p}^{w}*L^{q}\subseteq s_{r}^{w}\quad\text{and}\quad s_{p}^{w}*s_{q}^{w}\subseteq L^{r},\qquad\frac{1}{p}+\frac{1}{q}=1+\frac{1}{r},\ p,q,r\in[1,\infty]. (0.6) In Section 3 we extend ( 0.6 ) and show that if c,c_{0}>0 and the Young functions \Phi_{0} , \Phi_{1} , and \Phi_{2} satisfy st_{1}t_{2}\leq c(\Phi_{0}^{*}(s)\Phi_{1}(t_{1})+\Phi_{0}^{*}(s)\Phi_{2}(t_{2})+\Phi_{1}(t_{1})\Phi_{2}(t_{2})), (0.7) where s,t_{1},t_{2}\in[0,c_{0}] and \Phi_{j}^{*} denotes the conjugate Young function, then s_{\Phi_{1}}^{w}*L^{\Phi_{2}}\subseteq s_{\Phi_{0}}^{w}\quad\text{and}\quad s_{\Phi_{1}}^{w}*s_{\Phi_{2}}^{w}\subseteq L^{\Phi_{0}}. (0.8) By choosing \Phi_{1}=\Phi_{[p]} , \Phi_{2}=\Phi_{[q]} and \Phi_{0}=\Phi_{[r]} , then ( 0.8 ) agrees with ( 0.6 ). In this case, ( 0.7 ) holds true under the conditions on p , q and r in ( 0.6 ). More generally, we show that the mapping properties of the convolutions in ( 0.8 ) still hold true with other pseudo-differential calculi in place of the Weyl quantization, in the definition of s_{\Phi}^{w} -classes. In Section 4 we deduce dilated convolution results for Orlicz Schatten-von Neumann classes. In particular it follows from Theorem 4.1 that if \Phi_{0},\Phi_{1},\Phi_{2} satisfy ( 0.7 ), then a(s\,\cdot\,)*b(t\,\cdot\,)\in s_{\Phi_{0}}^{w},\quad\text{when}\quad a\in s_{\Phi_{1}}^{w},\ b\in s_{\Phi_{2}}^{w},\ \pm s^{-2}\pm t^{-2}=1, for some choices of + and - at each \pm . More generally, in Section 4 we present a multi-linear version of this dilated convolution property. An essential ingredient in the proof of this extension is induction. In order to facilitate the reader to reach the applications, also taking into account that the proof is rather comprehensive, we have pushed it to Appendix A . In Section 5 we present some applications concerning Orlicz Schatten-von Neumann properties for Toeplitz operators. Especially we find that a Toeplitz operator belongs to \mathscr{I}_{\Phi} if its symbol belongs to L^{\Phi} , and that similar conclusions can be performed if a suitable dilation of the symbol belongs to s_{\Phi}^{w} . Finally we remark that in [ 3 ] we perform another approach, based on interpolation techniques, to reach convolution properties for Orlicz spaces and Orlicz Schatten-von Neumann symbol classes. The final results are similar, but contain differences between the different approaches. Especially, in some aspects, the results here are more general, while in other aspects, the analogous results in [ 3 ] are more general. 1. Preliminaries In this section we recall some facts about Orlicz spaces, pseudo-differential operators, Toeplitz operators and Schatten-von Neumann classes of Orlicz types. In the last part we deduce some invariance properties of symbol classes of Orlicz Schatten-von Neumann types. 1.1. Orlicz spaces Since our investigations involve quasi-Banach spaces which fail to be Banach spaces, we first recall the definition of such spaces. Let \mathcal{B} be a vector space. A quasi-norm (of order r\in(0,1] ) on \mathcal{B} is a map f\mapsto\|f\|_{\mathcal{B}} from \mathcal{B} to \mathbf{R} such that \displaystyle\|f\|_{\mathcal{B}} \displaystyle\geq 0, \displaystyle\qquad f \displaystyle\in\mathcal{B}, with equality, if and only if f=0 , \displaystyle\|\alpha f\|_{\mathcal{B}} \displaystyle=|\alpha|\cdot\|f\|_{\mathcal{B}}, \displaystyle\qquad f \displaystyle\in\mathcal{B},\ \alpha\in\mathbf{C}, (1.1) and \displaystyle\|f+g\|_{\mathcal{B}} \displaystyle\leq 2^{\frac{1}{r}-1}\big(\|f\|_{\mathcal{B}}+\|g\|_{\mathcal{B}}\big), \displaystyle\qquad f,g \displaystyle\in\mathcal{B}. (1.2) If the topology on \mathcal{B} is defined through this quasi-norm, then \mathcal{B} is called a quasi-normed space (of order r ), or an

Read the complete text