A Note on Maximal Operators for Moment Curves
Chenjian Wang · 2025
Read A Note on Maximal Operators for Moment Curves by Chenjian Wang (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-2509-01173.txt.
Introduction
Chenjian Wang studies a type of maximal operator linked to moment curves in spaces of dimension three or more. Maximal operators measure the largest average of a function over a family of shapes, and they are central tools in harmonic analysis. Here the averages follow a curved path rather than straight lines or balls.
Wang derives mapping properties that state how these operators act between function spaces, measured in L p norms. In a special case, the estimate obtained is sharp, meaning it cannot be improved. The short note adds a precise piece to the theory of operators tied to curves.
Catalog record
| CREATOR | Chenjian Wang |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 112 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2509-01173
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
A Note on Maximal Operators for Moment Curves Chenjian Wang Abstract . We consider a type of maximal operators associated to moment curves in ℝ d , d ≥ 3 . \mathbb{R}^{d},d\geq 3. We derive L p L^{p} mapping properties for these operators. In a special case, the estimate is sharp. † † footnotetext: Key Words: Geometric maximal operators, Kakeya type problem. Mathematical subject classification: Primary: 42B25-06. 1. Introduction Geometric measure theory is the study of the geometric properties of sets, typically in Euclidean space, through measure theory. An interesting genre of problems in this area is determining the size, usually quantified via Lebesgue measure or dimensions, of sets that are quantitatively known to be large. A prototypical example of such a problem is the notorious Kakeya conjecture which aims to quantify the size of a set that contains a unit line segment in every direction. Such sets are called Kakeya sets. Besicovitch [ 2 ] showed that, contrary to intuition, Kakeya sets can have Lebesgue measure zero. This led to the conjecture which states that Kakeya sets must have full Hausdorff dimension. The problem has been solved for d=2 [ 7 , 6 ] . Recently, Wang and Zahl claimed the resolution of three-dimensional case [ 20 , 21 , 22 ] and the problem is still open in all higher dimensions [ 5 , 23 , 12 ] . The intense interest in the Kakeya conjecture has led to the study of related problems in the “curved” setting. For d=2 , Besicovitch and Rado [ 3 ] and Kinney [ 13 ] constructed a Lebesgue null set that contains all circles with radius r\in[1,2] . Inspired by this, such a set in \mathbb{R}^{d},d\geq 2 (replace circles by r\mathbb{S}^{d-1} for d\geq 3 ) is called Besicovitch-Rado-Kinney set, or a BRK set for short. One can ask a similar question to the Kakeya conjecture: what is the Hausdorff dimension of a BRK set? The problem has been solved for all d\geq 2 : for d\geq 3 in [ 16 ] and d=2 in [ 24 ] . It is now known that all BRK sets in \mathbb{R}^{d},\ d\geq 2 have full Hausdorff dimension. In this note, we consider problems of a similar flavor in \mathbb{R}^{d},d\geq 3 . Let us denote by \gamma(t):=(t,t^{2},...,t^{d}) the standard moment curve in \mathbb{R}^{d} . We want to study sets that contain many affine copies of \gamma . More precisely, we introduce the following BRK-type sets. Throughout the note, we denote (1) s^{\prime}=d+1-s. Definition 1 ( s^{\prime} -parameter moment BRK type sets) . Assume d\geq 3 . For s^{\prime}\in\{1,2,...,d\} , a compact set S_{s^{\prime}}\subseteq\mathbb{R}^{d} is called a s^{\prime} -parameter moment BRK type sets if for all (x_{s+1},...,x_{d-1},x_{d},r)\in[-1/4,1/4]^{s^{\prime}-1}\times[1/2,2]\subseteq\mathbb{R}^{s^{\prime}} , there is (x_{1},...,x_{s})\in\mathbb{R}^{s} such that the (truncated) moment curve (2) H(\mathbf{x},r):=\{\mathbf{x}+r\gamma(t):t\in[-1,1]\}\subseteq S_{s^{\prime}}, where \mathbf{x}=(x_{1},...,x_{d}) . In the following parts of the note, (i) we will denote the range of each coordinate of \mathbf{x} and r by (3) I_{1}:=[-1/4,1/4],\ I_{2}:=[1/2,2]. and \overline{\mathbf{x}}_{s}=(x_{1},...,x_{s}),\quad\underline{\mathbf{x}}_{s^{\prime}-1}=(x_{s+1},...,x_{d}). When s^{\prime} is fixed, we will abbreviate these as \overline{\mathbf{x}} and \underline{\mathbf{x}} . (ii) Denote the isotropic \delta -neighborhood of the moment curve H(\mathbf{x},r) as H_{\delta}(\mathbf{x},r) , More precisely, H_{\delta}(\mathbf{x},r):=\bigcup_{\mathbf{y}\in H(\mathbf{x},r)}B(\mathbf{y},\delta), where \mathbf{y}=(y_{1},...,y_{d})\in\mathbb{R}^{d} and B(\mathbf{y},\delta) is the ball with center \mathbf{y} and radius \delta . It is elementary to see that when r\in[1/2,2] , for all \mathbf{x}\in\mathbb{R}^{d}, \mathcal{L}^{d}(H_{\delta}(\mathbf{x},r))\approx\delta^{d-1}. (iii) All dimensions or \dim(\cdot) refer to Hausdorff dimension. (iv) As conventions, A\lesssim B or B\gtrsim A represents “ \exists C>0 (independent of \delta but may depend on other irrelevant parameter such as d ), such that A\leq CB ”. If C depends on some irrelevant parameter \tau , we may also denote A\lesssim C_{\tau}B as A\lesssim_{\tau}B. To state the dimension estimates for the s^{\prime} -parameter moment BRK type sets, we introduce the following definition. Definition 2 ( s -parameter maximal function associated to moment curves) . Fix \delta>0 . For a Schwartz function f:\mathbb{R}^{d}\to\mathbb{C} , d\geq 3, define the s -parameter maximal function associated to moment curves \mathcal{M}^{s}_{\delta}f:I_{1}^{s^{\prime}-1}\times I_{2}\to\mathbb{R}_{+} as (4) \mathcal{M}^{s}_{\delta}f(\underline{\mathbf{x}}_{s^{\prime}-1},r):=\sup_{\mathbf{x}\in\mathbb{R}^{s}}\frac{1}{\mathcal{L}^{d}(H_{\delta}(\mathbf{x},r))}\int_{H_{\delta}(\mathbf{x},r)}|f(\mathbf{y})|d\mathbf{y},s=1,2,...,d where \mathcal{L}^{d} is the d -dimensional Lebesgue measure and \ell is the arc length measure. Recall our notation in ( 1 ). In the parameter space \mathbb{R}^{d+1}=\mathbb{R}^{s}\times\mathbb{R}^{s^{\prime}} , s is the dimension or number of parameters we take supremum over. s^{\prime} is the dimension or number of parameters left, which is also the number of variables of the maximal functions in ( 4 ). By applying Wolff’s circular maximal estimate in [ 24 ] and sharp (exponent) local smoothing estimate by Ko, Lee, and Oh [ 15 ] , the following maximal estimates can be obtained. Theorem 1 (maximal estimates) . Fix d\geq 3 and \varepsilon>0 , then for all p\geq p_{d}, there exists C_{\varepsilon}>0 depending on all the aforementioned quantities and s , such that (5) \left\|\mathcal{M}^{s}_{\delta}f\right\|_{L^{p}(I_{1}^{s^{\prime}-1}\times I_{2})}\leq C_{\varepsilon}\delta^{-\alpha(s)}\left\|f\right\|_{L^{p}{(\mathbb{R}^{d})}}. Here, p_{d}=\begin{cases}3,&\text{if }s=d,\\ 4d-2,&\text{if }s=1,...,d-1,\end{cases} and (6) \alpha(s)=\begin{cases}\varepsilon+\frac{s-2}{p},&s=2,...,d,\\ 0,&s=1.\end{cases} When s=1, C_{\varepsilon} is an absolute number independent of \varepsilon . Note that when s=d , the result p_{d}=3 is better than the general results obtained from local smoothing p_{d}\geq 4d-2 . Remark 1 . When s=1 , since the operator norm is independent of \delta, by a standard limiting argument, one can deduce \left\|\mathcal{M}^{1}f\right\|_{L}^{p}\leq C\left\|f\right\|_{p}, and replace \mathcal{M}^{1}_{\delta} by \mathcal{M}^{1} , where \mathcal{M}^{1}f:=\sup_{\mathbf{x}\in\mathbb{R}^{d}}\int_{H_{(}\mathbf{x},r)}|f(\mathbf{y})|d\ell(\mathbf{y}). To unify notations, we still use \mathcal{M}^{1}_{\delta}. When s=d , the estimate ( 5 ) is sharp in the following sense: (1) (sharpness of the exponent of \delta ) If p\geq 3 , \not\exists\beta<\frac{d-2}{p} such that the following holds \left\|\mathcal{M}^{d}_{\delta}f\right\|_{L^{p}([1/2,2])}\leq C_{\varepsilon}\delta^{-(\varepsilon+\beta)}\left\|f\right\|_{L^{p}{(\mathbb{R}^{d})}}. (2) (sharpness of the range of p ) If p<3 , then an estimate with the form s=d in ( 5 ) cannot hold. By a standard “maximal estimate to dimension estimate” argument (see Section 2 ), the following corollary is straightforward. Recall the notation in Definition 1 . Corollary 2 (dimension estimates) . For d\geq 3 and s^{\prime}\in\{1,2,...,d\} , (7) \dim(S_{s^{\prime}})\geq\min\{s^{\prime}+1,d\}. The dimension estimates ( 7 ) are sharp in the sense that there exist S_{s^{\prime}} such that \dim(S_{s^{\prime}})=\min\{s^{\prime}+1,d\}. All sharp examples will be presented in Section 4 . In particular, for the case s^{\prime}=d , S_{d} has positive Lebesgue measure, see Proposition 1 . Remark 2 . Heuristically, S_{s^{\prime}} is like an s^{\prime} -dimensional union of 1 -dimensional moment curves. Therefore, it is reasonable to guess the resulting set has Hausdorff dimension s^{\prime}+1 . In fact, Corollary 2 is special case of the results in Ham, Ko, Lee, and Oh [ 11 ] . There they generalized the local smoothing in [ 15 ] to fractal setting (similar argument can be found in [ 10 ] ). This provides more flexibility in the choice of curve families, as [ 11 ] allows s^{\prime} to take arbitrary values in (0,d] in an appropriate sense. Remark 3 . The methods applied to obtain maximal estimates for s=d and s=1,...,d-1 are different. It may be interesting that for \mathcal{M}^{d}_{\delta} , the torsion of moment curve is not needed, even the result is sharp under both senses. For \mathcal{M}^{s}_{\delta},s=1,...,d-1 , the torsion is necessary to obtain the sharp local smoothing in [ 15 ] . This is also reflected in the dimension estimates: It is the curvature in s^{\prime} transversal direction(s) that makes S_{s^{\prime}} having dimension at least \min\{s^{\prime}+1,d\} . For \mathcal{M}_{\delta}^{d} which corresponds