Maximal estimates for orthonormal systems of wave equations
Shinya Kinoshita, Hyerim Ko, Shobu Shiraki · 2025
Read Maximal estimates for orthonormal systems of wave equations by Shinya Kinoshita, Hyerim Ko, Shobu Shiraki (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-2508-19446.txt.
Introduction
Shinya Kinoshita, Hyerim Ko, and Shobu Shiraki study maximal estimates for whole families of wave solutions at once, in a 2025 paper. Classic estimates track the largest value one wave reaches; the family version matters for systems with many paired states.
The authors make partial progress in low dimensions. Their method reads the geometry of the wave operator's kernel through a Schatten-class lens. A lemma by Wolff about how thickened spheres overlap does heavy work in the argument. The results extend what was known for single waves to orthonormal families. Such bounds feed into the study of gases of particles that obey wave dynamics.
Catalog record
| CREATOR | Shinya Kinoshita, Hyerim Ko, Shobu Shiraki |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 67 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2508-19446
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Maximal estimates for orthonormal systems of wave equations Shinya Kinoshita, Hyerim Ko, Shobu Shiraki Abstract . This paper investigates maximal estimates of the wave operators for orthonormal families of initial data. We extend the classical maximal estimates for the wave operator by making partial progress on maximal estimates for orthonormal systems in low dimensions. Our novel approach is based on a geometric analysis of the kernel of wave operators within the framework of Schatten 2 2 estimates. In particular, we exploit Wolff’s geometric lemma on the intersection patterns of thickened spheres. 1. Introduction 1.1. Motivation Let n\geq 1 . For s>0 , let H^{s}({\mathbb{R}}^{n}) denote the inhomogeneous Sobolev space of order s equipped with the norm \|g\|_{H^{s}}=\|(1-\Delta)^{s/2}g\|_{L^{2}} . For g\in H^{s}({\mathbb{R}}^{n}) , we consider the wave equation \partial_{t}^{2}u-\Delta u=0 with initial conditions u(x,0)=g(x) and \partial_{t}u(x,0)=0 . The solution u , denoted by \cos(t\sqrt{-\Delta})g , can be represented as \cos(t\sqrt{-\Delta})g(x)=\frac{1}{2}\big(U_{t}+U_{-t}\big)g(x), where U_{t}g(x) is the solution to the half-wave equation (1.1) i\partial_{t}u+\sqrt{-\Delta}u=0 under the initial condition u(x,0)=g(x) . Furthermore, U_{t}g(x) has the explicit representation U_{t}g(x)=(2\pi)^{-n}\int_{{\mathbb{R}}^{n}}e^{i(x\cdot\xi+t|\xi|)}\widehat{g}(\xi)\,\mathrm{d}\xi. Maximal estimates for various operators have drawn considerable attention from different perspectives. A classical result due to [ 10 , 21 ] establishes that maximal estimates for single wave operator: (1.2) \Big\|\sup_{0<t<1}|U_{t}g|\Big\|_{L^{2}({\mathbb{B}}^{n})}\lesssim\|g\|_{H^{s}({\mathbb{R}}^{n})} holds if and only if s>\frac{1}{2} . Here, \mathbb{B}^{n}(x,r) denotes the n -dimensional ball of radius r>0 centered at x and \mathbb{B}^{n}=\mathbb{B}^{n}(0,1) represents the unit ball. Historically, local maximal estimates of this type were first explored in the context of the Schrödinger equation, famously known as Carleson’s problem [ 7 , 11 ] . While our focus here is on the wave equation, interested readers may refer to [ 6 , 12 , 13 ] for recent breakthroughs on the Schrödinger equation variant. The analysis of ( 1.2 ) for the wave equation is notably more straightforward, following from relatively simple arguments, whereas the Schrödinger case requires more delicate recent techniques. For a detailed discussion, see Appendix A . The maximal estimates for orthonormal systems (f_{j})_{j} in H^{s}({\mathbb{R}}^{n}) present a natural extension of ( 1.2 ), where we consider infinitely many initial data, rather than a single datum. This generalization can be formulated as (1.3) \Big\|\sup_{0<t<1}\sum_{j}\lambda_{j}|U_{t}f_{j}|^{2}\Big\|_{L^{\beta}(\mathbb{B}^{n})}\lesssim\|\lambda\|_{\ell^{\beta}}. Here, (\lambda_{j})_{j}\in\ell^{\beta}(\mathbb{C}) and \beta\geq 1 . In fact, this is equivalent to \Big\|\sup_{0<t<1}\big(\sum_{j}|U_{t}g_{j}|^{2}\big)^{\frac{1}{2}}\Big\|_{L^{2\beta}(\mathbb{B}^{n})}\lesssim\Big(\sum_{j}\|g_{j}\|_{H^{s}}^{2\beta}\Big)^{\frac{1}{2\beta}} for orthogonal systems (g_{j})_{j} in H^{s}({\mathbb{R}}^{n}) , and it is straightforward to see ( 1.3 ) holds for \beta=1 if s>\frac{1}{2} as a consequence of the aforementioned classical result for a single datum, along with a simple application of the triangle inequality. Furthermore, if the regularity is sufficiently smooth, namely when s>\frac{n}{2} , the Bessel inequality guarantees ( 1.3 ) holds with \beta=\infty , hence for any \beta\geq 1 by the embedding \ell^{\beta}\subset\ell^{\infty} . In fact, assuming (g_{j})_{j} is orthonormal in L^{2}({\mathbb{R}}^{n}) , and thus so is (\widehat{g}_{j})_{j} , we observe (1.4) \displaystyle\sum_{j}\frac{|\lambda_{j}|}{\|\lambda_{j}\|_{\ell^{\infty}}}|U_{t}\langle D\rangle^{-s}g_{j}(x)|^{2}\leq\sum_{j}\Big|\int_{{\mathbb{R}}^{n}}e^{i(x\cdot\xi+t|\xi|)}\langle\xi\rangle^{-s}\widehat{g_{j}}(\xi)\,\mathrm{d}\xi\Big|^{2}\leq\|\langle\cdot\rangle^{-s}\|_{L^{2}}^{2}, where \langle D\rangle:=(1-\Delta)^{1/2} and \langle\xi\rangle=(1+|\xi|^{2})^{1/2} . The right-hand side remains bounded as long as s>\frac{n}{2} . Consequently, this also shows that there is nothing particularly interesting in the case n=1 . Additionally, when s=\frac{n}{2} , the estimate ( 1.3 ) holds for \beta\in[1,\infty) , as a consequence of ( 3.4 ) below together with a minor adjustment at low frequency. With those discussions in mind, in the current manuscript, we aim to establish ( 1.3 ) for the largest possible \beta (depending on s ) for n\geq 2 and s\in(\frac{1}{2},\frac{n}{2}) by making use of the orthogonal structure of the data. 1.2. Previous works The study of Strichartz estimates for orthonormal system has been made by Frank, Lewin, Lieb, and Serringer [ 14 ] and Frank–Sabin [ 16 , 15 ] for some operators such as Schrödinger operators. The result was later extended by Bez et al. [ 1 ] up to a sharp range for non-sharp admissible cases considering homogeneous initial datum. Strichartz estimates for various dispersive operators (cf. [ 5 ] ) and some of the boundary cases have been obtained by [ 3 , 2 ] . The Strichartz estimates for orthonormal system for wave equation seem more difficult and the sharp characterization in terms of the function spaces remains an open problem. In particular, the following diagonal Strichartz estimate has been conjectured (for a more general statement, see [ 5 , 2 ] ). Conjecture 1.1 . Let n\geq 2 and q\geq\frac{2(n+1)}{n-1} . Then, we have \Big\|\sum_{j}\lambda_{j}|U_{t}f_{j}|^{2}\Big\|_{L_{x,t}^{\frac{q}{2}}({\mathbb{R}}^{n+1})}\lesssim\|\lambda\|_{\ell^{\beta}} for orthonormal systems (f_{j})_{j} in the homogeneous Sobolev space \dot{H}^{s}({\mathbb{R}}^{n}) if \beta\leq\frac{n}{2(n+1)}q . The study of maximal estimates in the context of orthonormal systems is even more recent. Bez, Lee, and Nakamura [ 4 ] were the first to obtain a maximal-in-time estimate for the Schrödinger equation in one dimension, achieving the largest possible \beta when the regularity is at its minimal value, s=\frac{1}{4} . Their method involves first establishing corresponding maximal-in-space estimates and then applying the technique of swapping the roles of space and time to derive the maximal-in-time estimates. Subsequently, Bez and two of the present authors [ 3 ] adopted a more direct approach to extend these results beyond s=\frac{1}{4} , including for fractional Schrödinger operators. They also addressed cases where s is relatively close to \frac{n}{2} in higher dimensions. 1.3. Main results Similar challenges to those in the Strichartz estimates arise when considering the maximal estimates for wave operators. We address these challenges and obtain new, non-trivial results that represent partial progress toward the optimal bounds. To begin, we propose our conjecture concerning the maximal estimates for orthonormal systems in the wave equation framework. The conjecture is based on modifications of the counterexamples from the single-datum setting in [ 9 , 2 ] , following a similar approach to that used in [ 3 ] for the Schrödinger equation. Throughout the paper, we consider the orthonormal function of initial data (f_{j})_{j} in H^{s}(\mathbb{R}^{n}) . However, all our results remain valid for (f_{j})_{j} in the homogeneous Sobolev space \dot{H}^{s}({\mathbb{R}}^{n}) by a similar computation to ( 1.4 ). Conjecture 1.2 . Let n\geq 2 and \frac{1}{2}<s<\frac{n}{2} . The estimate ( 1.3 ) holds for all orthonormal system (f_{j})_{j} in H^{s}({\mathbb{R}}^{n}) if and only if \beta\leq\min\Big\{\frac{n}{n-2s},\frac{n-1}{n+1-4s}\Big\}. One might attempt to approach the problem by adapting the arguments used to establish single-datum results, such as ( 1.2 ) based on the Sobolev embedding in order to address Conjecture 1.2 . However, this would depend on Conjecture 1.1 , and even if the full Conjecture 1.1 were proven, it would still not be sufficient. We will elaborate on this point in Section 3 . We prove ( 1.3 ) with a nontrivial range of \beta=\beta(s) when n=2,3,4 . Theorem 1.3 . Let n=3,4 and \frac{1}{2}<s<\frac{n}{2} . Then ( 1.3 ) holds for all orthonormal system (f_{j})_{j}\in H^{s}({\mathbb{R}}^{n}) provided (1.5) \displaystyle\beta<\min\Big\{\frac{2n-1}{2(n-2s)},\frac{2n-3}{2n-1-4s}\Big\}. See Figure 1.2 below to visualize the improvement compared to Conjecture 1.2 when n=3 . The argument employed in Theorem 1.3 extended to the case n=2 , yielding the bound \beta<\min\{\frac{5}{8-4s},\frac{3}{7-4s}\} (see the Subsection 5.4 for further discussions). Moreover, by exploiting the self-similar structure of the inequality through a certain conic decomposition, we achieve a further improvement beyond the n=2 case of Theorem 1.3 , as illustrated in Figure 1.2 . Theorem 1.4 . Let n=2 and \frac{1}{2}<s<1 . Then, ( 1.3 ) holds for all orthonormal system (f_{j})_{j}\in H^{s}(\mathbb{R}^{2}) provided (1.6) \displaystyle\beta<\min\left\{\frac{9}{14(1-s)},\frac{5}{12-14s}\right\}.