Self-improving properties of weighted norm inequalities on metric measure spaces

Juha Kinnunen, Juha Lehrbäck, Antti V. Vähäkangas, Dachun Yang · 2025

Read Self-improving properties of weighted norm inequalities on metric measure spaces by Juha Kinnunen, Juha Lehrbäck, Antti V. Vähäkangas, Dachun Yang (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-2501-17651.txt.

Introduction

Juha Kinnunen, Juha Lehrback, Antti Vahakangas, and Dachun Yang prove that some bounds in math lift their own strength. The work came out in 2025. Their stage is a class of spaces with both gap and bulk gauges. In such spaces, the bulk of balls grows in a curbed, two-fold way.

The team studies the Muckenhoupt weight rule. They also study weighted bounds for the Hardy-Littlewood max tool, which takes means of a form over balls. Self-lift means a bound taken at one strength holds at a bit more strength too. The team gives straight proofs for this lift. They use a Whitney cover step and a trick in the style of the Calderon-Zygmund split. Their path skips reverse Holder bounds, the usual side road for such finds. The clean proofs make clear why the lift takes place.

Catalog record

CREATORJuha Kinnunen, Juha Lehrbäck, Antti V. Vähäkangas, Dachun Yang
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE42 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2501-17651

Browse the full library · About the Tower of Babel

Full text

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

Self-improving properties of weighted norm inequalities on metric measure spaces Juha Kinnunen, Juha Lehrbäck, Antti V. Vähäkangas, Dachun Yang Abstract . This work discusses self-improving properties of the Muckenhoupt condition and weighted norm inequalities for the Hardy–Littlewood maximal function on metric measure spaces with a doubling measure. Our main result provides direct proofs of these properties by applying a Whitney covering argument and a technique inspired by the Calderón–Zygmund decomposition. In particular, this approach does not rely on reverse Hölder inequalities. 2020 Mathematics Subject Classification 42B25, 42B35, 46E30, 30L99 1. Introduction Harmonic analysis on metric measure spaces, and more broadly on spaces of homogeneous type, continues to be an active area of research. Coifman and Weiss provided a systematic account of many aspects of harmonic analysis on spaces of homogeneous type in [ 2 ] . The theory of weighted norm inequalities and Muckenhoupt weights has been explored in greater detail in works such as those by Strömberg and Torchinsky in [ 12 ] and by Genebashvili, Gogatishvili, Kokilashvili and Krbec in [ 5 ] . Muckenhoupt weights exhibit a remarkable self-improving property: if a weight belongs to a certain Muckenhoupt class A_{p} , 1<p<\infty , then it also belongs to another Muckenhoupt class A_{p-\varepsilon} with a slightly smaller index, see [ 12 , Lemma 8, p. 5] . This self-improving nature of Muckenhoupt weights arises from the fact that these weights satisfy a reverse Hölder inequality, see [ 12 , Theorem 15, p. 9] for details. Reverse Hölder inequalities are powerful tools in the study of weighted norm inequalities, but the purpose of this work is to discuss an alternative approach to the self-improving property of the Muckenhoupt weights that entirely avoids relying on reverse Hölder inequalities. Thus the novelty in this work lies in the methodology rather than in the actual results. Instead of reverse Hölder inequalities, we utilize the close connection between Muckenhoupt weights and weighted norm inequalities for the Hardy–Littlewood maximal function, see [ 3 , 4 ] . More specifically, let (X,d,\mu) be a metric measure space equipped with a doubling measure \mu and let w be a weight in X ; see Sections 2 and 3 for the definitions. Then the following assertions are equivalent, for 1<p<\infty . (A) There exists a constant C>0 such that (1.1) \int_{X}\bigl({M}^{*}f(x)\bigr)^{p}w(x)\,d\mu(x)\leq C\int_{X}\lvert f(x)\rvert^{p}w(x)\,d\mu(x) for every f\in L^{\infty}(X) with a bounded support. (B) The weight w belongs to the Muckenhoupt class A_{p} . (C) There exists a constant C>0 such that inequality ( 1.1 ) holds for every f\in L^{p}(w\,d\mu) . In ( 1.1 ), the (noncentered) Hardy–Littlewood maximal function of f is defined as {M}^{*}f(x)=\sup_{B\ni x}\frac{1}{\mu(B)}\int_{B}\lvert f(y)\rvert\,d\mu(y), where the supremum is taken over all balls B in X containing x . It was shown in [ 2 ] that {M}^{*} is a bounded operator on L^{p}(X) for every 1<p\leq\infty and of weak type (1,1) under the assumption that \mu is a doubling measure; see also [ 12 , Theorem 3, p. 3] . Observe that assertion (C) above states that the operator {M}^{*} is bounded on L^{p}(w\,d\mu) , for 1<p<\infty . The equivalence of the assertions (B) and (C) is well known, see [ 12 , Theorem 9, p. 5] . For our purposes, it is convenient to consider also assertion (A) , which is a priori slightly weaker than (C) . That (A) implies (B) is shown in Lemma 4.2 , and the more involved implication from (B) to (C) is proved in Theorem 4.5 by adapting an argument of Lerner [ 10 ] . This approach is direct and, in particular, does not refer to reverse Hölder inequalities. Finally, the implication from (C) to (A) is a triviality. Now, assume that assertion (C) holds, i.e., that {M}^{*} is bounded on some L^{p}(w\,d\mu) , 1<p<\infty , where w is a weight and \mu a doubling measure in X . By the Marcinkiewicz interpolation theorem and the boundedness of the maximal operator {M}^{*} on L^{\infty}(w\,d\mu) , it follows that then {M}^{*} is also bounded on L^{p+\varepsilon}(w\,d\mu) for every \varepsilon>0 . On the other hand, by using the equivalence of the assertions (B) and (C) and the known self-improving property of the Muckenhoupt A_{p} condition, there exits some \varepsilon>0 such that {M}^{*} is bounded on L^{p-\varepsilon}(w\,d\mu) as well. The main goal in this work is to give a direct and essentially self-contained proof for this self-improving property of the weighted norm inequalities. In particular, we show in Lemma 5.6 that if assertion (C) holds, then there exist \varepsilon>0 and C>0 such that (1.2) \int_{X}\bigl({M}^{*}f(x)\bigr)^{p-\varepsilon}w(x)\,d\mu(x)\leq C\int_{X}\lvert f(x)\rvert^{p-\varepsilon}w(x)\,d\mu(x) for every f\in L^{\infty}(X) with a bounded support. The proof of Lemma 5.6 uses a Whitney covering argument and a technique inspired by the Calderón–Zygmund decomposition, based on an unpublished note on Muckenhoupt weights by Keith and Zhong. This approach extends the corresponding Euclidean result with the Lebesgue measure [ 8 , Theorem 8.11] to metric spaces and it does not rely on reverse Hölder inequalities. Similar ideas have been applied in various contexts, for example, in Keith and Zhong [ 7 ] , Kinnunen, Lehrbäck, Vähäkangas and Zhong [ 9 ] and Lewis [ 11 ] . By the equivalence of the assertions (A) , (B) and (C) , we obtain from Lemma 5.6 an independent proof for the self-improving property of the Muckenhoupt classes, see Theorem 5.11 , and we also conclude that the improved weighted norm inequality ( 1.2 ) holds for all functions f\in L^{p-\varepsilon}(w\,d\mu) , see Corollary 5.13 . Before the statements and proofs of our main results and tools in Sections 4 and 5 , we recall preliminaries on metric spaces and on Muckenhoupt weights in Sections 2 and 3 , respectively. In Section 3 we also prove some basic properties of the Muckenhoupt classes for the sake of completeness. 2. Preliminaries Throughout the paper, we assume that X=(X,d,\mu) is a metric measure space equipped with a metric d and a positive complete Borel measure \mu such that 0<\mu(B)<\infty for all open balls B\subset X , where B=B(x,r)=\{y\in X\,:\,d(y,x)<r\} with x\in X and r>0 . The space X is separable under the above assumptions, we refer to [ 1 , Proposition 1.6] . We also assume throughout that \#X\geq 2 and that the measure \mu is doubling, that is, there exists a constant c_{\mu}>1 , called the doubling constant of \mu , such that (2.1) \mu(2B)\leq c_{\mu}\,\mu(B) for all balls B=B(x,r) in X . Here we use for 0<t<\infty the notation tB=B(x,tr) . We let \mathchoice{\mathop{\kern 1.99997pt\vrule width=6.00006pt,height=3.0pt,depth=-2.49997pt\kern-8.00003pt\intop}\nolimits_{\kern-3.27773ptA}}{\mathop{\kern 1.00006pt\vrule width=5.0pt,height=3.0pt,depth=-2.59996pt\kern-6.00006pt\intop}\nolimits_{A}}{\mathop{\kern 1.00006pt\vrule width=5.0pt,height=3.0pt,depth=-2.59996pt\kern-6.00006pt\intop}\nolimits_{A}}{\mathop{\kern 1.00006pt\vrule width=5.0pt,height=3.0pt,depth=-2.59996pt\kern-6.00006pt\intop}\nolimits_{A}}f(y)\,d\mu(y)=\frac{1}{\mu(A)}\int_{A}f(y)\,d\mu(y) be the integral average of f\in L^{1}(A) over a measurable set A\subset X with 0<\mu(A)<\infty . We say that a function f\colon X\to{\mathbb{R}} has a bounded support, if \{x\in X\,:\,f(x)\not=0\} is contained in some ball B of X . The characteristic function of a set E\subset X is denoted by {\mbox{\hbox{$\mathbf{1}$}}}_{\hbox{$\scriptstyle E$}} ; that is, {\mbox{\hbox{$\mathbf{1}$}}}_{\hbox{$\scriptstyle E$}}(x)=1 if x\in E and {\mbox{\hbox{$\mathbf{1}$}}}_{\hbox{$\scriptstyle E$}}(x)=0 if x\in X\setminus E . 3. Muckenhoupt weights For convenience, we recall the standard definition and elementary properties of Muckenhoupt weights in the setting of metric measure spaces. We follow [ 4 ] and also refer to [ 8 ] . Definition 3.1 . A measurable function w\colon X\to{\mathbb{R}}

Read the complete text