Table of Contents
Fetching ...

The finite Hilbert transform acting on $L^\infty$

Guillermo P. Curbera, Susumu Okada, Werner J. Ricker

TL;DR

The paper analyzes the finite Hilbert transform $T$ acting on $L^\infty(-1,1)$ with values in the Zygmund space $L_{\mathrm{exp}}(-1,1)$, complementing prior work on $T$ from $L\log L$ to $L^1$ and highlighting non-separability effects. It establishes boundedness $T: L^\infty\to L_{\mathrm{exp}}$ and develops the auxiliary operator $\widecheck{T}$ as a functional inverse on the range, linking $T$ and $\widecheck{T}$ via $T\widecheck{T}(f)=f-Q(f)$ and $\widecheck{T}T(f)=f$ for $f\in L^\infty$. The range of $T$ is characterized as a proper subset of $\ker(\varphi_{1/w})$, with $w(x)=\sqrt{1-x^2}$, and a full inverse theory is obtained: $g\in T(L^\infty)$ iff $\widecheck{T}(g)\in L^\infty$ and $\int g/w=0$, enabling inversion $f=\widecheck{T}(g)$ when $g=T(f)$. The paper also proves the optimality of the domain for $T$ taking values in $L_{\mathrm{exp}}$, showing $[T,L_{\mathrm{exp}}]=L^\infty$ and that no continuous linear extension to a larger domain exists, thereby clarifying the operator's domain-range structure in this non-separable setting.

Abstract

The action of the finite Hilbert transform defined on $L^\infty(-1,1)$ and taking its values in the Zygmund space $L_{\textnormal{exp}}(-1,1)$ is studied in detail. This is a reciprocal situation to the investigation recently undertaken in [11] of the finite Hilbert transform defined on the Zygumd space $L\textnormal{log} L(-1,1)$ and taking its values in $L^1(-1,1)$. The fact that both $L^\infty(-1,1)$ and $L_{\textnormal{exp}}(-1,1)$ fail to be separable generates new features not present in[11].

The finite Hilbert transform acting on $L^\infty$

TL;DR

The paper analyzes the finite Hilbert transform acting on with values in the Zygmund space , complementing prior work on from to and highlighting non-separability effects. It establishes boundedness and develops the auxiliary operator as a functional inverse on the range, linking and via and for . The range of is characterized as a proper subset of , with , and a full inverse theory is obtained: iff and , enabling inversion when . The paper also proves the optimality of the domain for taking values in , showing and that no continuous linear extension to a larger domain exists, thereby clarifying the operator's domain-range structure in this non-separable setting.

Abstract

The action of the finite Hilbert transform defined on and taking its values in the Zygmund space is studied in detail. This is a reciprocal situation to the investigation recently undertaken in [11] of the finite Hilbert transform defined on the Zygumd space and taking its values in . The fact that both and fail to be separable generates new features not present in[11].
Paper Structure (6 sections, 18 theorems, 49 equations)

This paper contains 6 sections, 18 theorems, 49 equations.

Key Result

Proposition 3.1

Let $f\in L^1$ and $g\in L\textnormal{log} L$ satisfy $fT(g\chi_A)\in L^1$, for every set $A\in\mathcal{B}$. Then $gT(f)\in L^1$ and the following Parseval formula is valid: In particular, if $f\in L^\infty$ and $g\in L\textnormal{log} L$, then the above assumptions are satisfied and hence, eq-3.1 is valid.

Theorems & Definitions (39)

  • Proposition 3.1
  • Proposition 3.2
  • proof
  • Remark 3.3
  • Proposition 3.4
  • proof
  • Proposition 3.5
  • proof
  • Remark 3.6
  • Lemma 4.1
  • ...and 29 more