Systolic S 1 S^{1} -index and characterization of non-smooth Zoll convex bodies
Stefan Matijević · 2025
Read Systolic S 1 S^{1} -index and characterization of non-smooth Zoll convex bodies by Stefan Matijević (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-2501-13856.txt.
Introduction
Stefan Matijević builds a new yardstick for a prized shape trait: a convex body whose shortest loops all share one length. Such bodies are called Zoll, after a classic case. His 2025 work counts a symmetry index over the space of those shortest loops.
The count turns out to be a symplectic invariant, a number that holds under a broad class of warps. With it, Matijević defines a wider class of such bodies, then shows it matches the classic class when each point bears just one shortest loop. The wider class can also be read off from known capacity numbers. It forms a closed set, so limits of such bodies stay in the class. A ball-like inside forces the trait, he adds.
Catalog record
| CREATOR | Stefan Matijević |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 84 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-13856
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Systolic S 1 S^{1} -index and characterization of non-smooth Zoll convex bodies Stefan Matijević Abstract . We define the systolic S 1 S^{1} -index of a convex body as the Fadell–Rabinowitz index of the space of generalized systoles associated with its boundary. We show that this index is a symplectic invariant. Using the systolic S 1 S^{1} -index, we introduce the notion of generalized Zoll convex bodies and prove that this definition coincides with the classical one when the convex body satisfies the uniqueness of systoles property, that is, when through every point passes at most one systole. Moreover, we show that generalized Zoll convex bodies can be characterized in terms of their Gutt–Hutchings capacities, and we prove that the space of generalized Zoll convex bodies is closed in the space of all convex bodies. As a corollary, we establish that if the interior of a convex body is symplectomorphic to the interior of a ball, then the convex body is generalized Zoll, and in particular Zoll if it satisfies the uniqueness of systoles property. Finally, we discuss several examples. 1 Introduction We say that K\subset\mathbb{R}^{2n} is a convex body if K is the closure of a bounded open convex set. The boundary of such a set is a hypersurface,which is a Lipschitz submanifold of \mathbb{R}^{2n} and, as such, has a tangent plane and an outer normal vector that are defined almost everywhere. Moreover, we can define the tangent cone and the normal cone at every point of \partial K (see [ Cla83 ] ). We define the outward normal cone of \partial K at x\in\partial K as N_{\partial K}(x)=\{\eta\in\mathbb{R}^{2n}\backslash\{0\}\mid\langle\eta,x-y\rangle\geq 0,\quad\forall y\in K\}. If \partial K is differentiable at x , then N_{\partial K}(x)=\mathbb{R}_{+}n , where n is the unit outer normal vector at x . Let J_{0} denote the standard complex structure of \mathbb{R}^{2n} , and let \mathbb{T}:=\mathbb{R}/\mathbb{Z} . Following [ Cla81 ] , we define the set of generalized closed characteristics on \partial K , denoted by \operatorname{Char}(\partial K) , as the set \operatorname{Char}(\partial K):=\{\gamma:\mathbb{T}\to\partial K\mid\dot{\gamma}(t)\in J_{0}N_{\partial K}(\gamma(t)),\ \text{a.e.}\}/\sim, where \gamma:\mathbb{T}\to\partial K is an absolutely continuous loop satisfying the condition that there exist positive constants m_{\gamma},M_{\gamma}>0 such that m_{\gamma}\leq|\dot{\gamma}(t)|\leq M_{\gamma},\quad\text{a.e.} The equivalence relation \sim is defined as follows. We say that \gamma_{1} is equivalent to \gamma_{2} (i.e., \gamma_{1}\sim\gamma_{2} ) if there exists a bi-Lipschitz homeomorphism \tau:\mathbb{T}\to\mathbb{T} such that \tau(0)=0 and \gamma_{1}(t)=\gamma_{2}(\tau(t)). This homeomorphism must preserve orientation. Let \gamma:\mathbb{T}\to\mathbb{R}^{2n} be an absolutely continuous loop. We define the action of \gamma as \mathcal{A}(\gamma):=\int\limits_{\gamma}\lambda=\frac{1}{2}\int\limits_{\mathbb{T}}\langle\dot{\gamma}(t),J_{0}\gamma(t)\rangle\,dt, where \lambda is any primitive 1-form of the standard symplectic form \omega_{0}=\sum_{j=1}^{n}dx_{j}\wedge dy_{j}. From Stokes’ formula, it follows that the action is invariant under reparametrizations that preserve orientation. Thus, the action is well-defined on the space \operatorname{Char}(\partial K) , is finite, and can easily be shown to be positive for every element of \operatorname{Char}(\partial K) . We define the spectrum of \partial K as \sigma(\partial K):=\{\mathcal{A}(\gamma)\mid\gamma\in\operatorname{Char}(\partial K)\}. The subset of \operatorname{Char}(\partial K) consisting of generalized closed characteristics with minimal action is denoted by \operatorname{Sys}(\partial K) , and it is always non-empty (see [ Cla81 , AO14 ] ). Elements of this set are called generalized systoles. To introduce an S^{1} -action on the spaces \operatorname{Char}(\partial K) and \operatorname{Sys}(\partial K) , we need a choice of parametrization. Assuming the origin is contained in the interior of K , we can define a positively 2 -homogeneous function H_{K}:\mathbb{R}^{2n}\to\mathbb{R} such that H_{K}^{-1}(1)=\partial K . Since H_{K} is convex, the subdifferential of H_{K} at x is defined for every x\in\mathbb{R}^{2n} as \partial H_{K}(x)=\{\eta\in\mathbb{R}^{2n}\mid H_{K}(y)-H_{K}(x)\geq\langle\eta,y-x\rangle,\quad\forall y\in\mathbb{R}^{2n}\}. For every x\in\mathbb{R}^{2n} , the set \partial H_{K}(x) is a non-empty convex compact set, which is precisely \{\nabla H_{K}(x)\} when H_{K} is differentiable at x . Moreover, the multifunction \partial H_{K} is upper semi-continuous. See [ Cla83 ] for the proof of these properties. The function H_{K} allows us to define generalized closed characteristics on \partial K as follows: \operatorname{Char}(\partial K)=\left\{\gamma:\mathbb{T}\to\partial K\text{ absolutely continuous}\mid\dot{\gamma}(t)\in TJ_{0}\partial H_{K}(\gamma(t))\ \text{a.e.},\ \text{for some }T>0\right\}. This definition is equivalent to the previous one in the set-theoretic sense. Indeed, any \gamma\in\operatorname{Char}(\partial K) , according to this definition, has a derivative that is bounded both from below and above. Moreover, N_{\partial K}(x)=\mathbb{R}_{+}\partial H_{K}(x) for all x\in\partial K . Hence, the conclusion follows. In addition, generalized closed characteristics with a fixed action T>0 are precisely those satisfying the equation \dot{\gamma}(t)\in TJ_{0}\partial H_{K}(\gamma(t)),\quad\text{a.e.} This follows from the fact that the 2 -homogeneity of H_{K} implies that H_{K}(x)=\frac{1}{2}\langle\partial H_{K}(x),x\rangle,\quad x\in\mathbb{R}^{2n}, where the equality holds in the set-theoretic sense (for any element of the corresponding subdifferential). Consequently, for \gamma\in\operatorname{Char}(\partial K) , we have \dot{\gamma}(t)\in TJ_{0}\partial H_{K}(\gamma(t)),\quad\text{a.e.}, if and only if \mathcal{A}(\gamma)=T . In particular, we have \operatorname{Sys}(\partial K)=\left\{\gamma:\mathbb{T}\to\partial K\ \text{a. continuous}\mid\dot{\gamma}(t)\in T_{\min}J_{0}\partial H_{K}(\gamma(t))\ \text{a.e.},\ T_{\min}=\min\sigma(\partial K)\right\}. Now, on the space \operatorname{Char}(\partial K) , there is a natural S^{1} -action defined by \theta\cdot\gamma=\gamma(\cdot-\theta),\quad\theta\in\mathbb{T},\ \gamma\in\operatorname{Char}(\partial K). Since the S^{1} -action preserves the absolute continuity of a loop \gamma:\mathbb{T}\to\mathbb{R}^{2n} , the subsets of generalized closed characteristics with fixed action are S^{1} -invariant. In particular, \operatorname{Sys}(\partial K) is S^{1} -invariant. Given that \partial H_{K} is bounded on \partial K , it follows that the space \operatorname{Sys}(\partial K) is uniformly bounded in the W^{1,\infty} -norm. We endow \operatorname{Sys}(\partial K) with the compact-open topology (i.e., C^{0} -topology). As shown in Proposition 2.1.1 , this space is compact with respect to this topology, which is, in fact, equivalent to the weak ∗ topology of W^{1,\infty} . In contrast, with respect to other topologies, such as the strong topology of W^{1,1} , this space is generally not compact (see Proposition 2.1.2 ). Hence, the uniform norm provides a natural framework for studying this space when one aims to preserve compactness properties, as in the smooth setting. Fadell–Rabinowitz index In this paper, we employ Alexandrov–Spanier cohomology (see [ Spa66 ] ), as it is the cohomology theory used in the definition of the Fadell–Rabinowitz index. Let \mathbb{F} be a field, and let X be a paracompact topological space equipped with an S^{1} -action. Following Borel’s construction, we define the S^{1} -equivariant cohomology of X with coefficients in \mathbb{F} as \operatorname{H}^{*}_{S^{1}}(X;\mathbb{F})=\operatorname{H}^{*}(X\times_{S^{1}}ES^{1};\mathbb{F}), where ES^{1}\to BS^{1} denotes the universal S^{1} -bundle over the classifying space, and X\times_{S^{1}}ES^{1}:=(X\times ES^{1})/S^{1}. For instance, one may take ES^{1}:=S^{\infty}\subset\mathbb{C}^{\infty} and BS^{1}:=\mathbb{C}P^{\infty} . The cohomology of the classifying space BS^{1} is the ring \operatorname{H}^{*}(BS^{1};\mathbb{F})\cong\mathbb{F}[e_{B}] , where e_{B} is a generator of \operatorname{H}^{2}(BS^{1};\mathbb{F}) . The projection map \pi_{2}:X\times_{S^{1}}ES^{1}\to BS^{1},\qquad\pi_{2}([x,y])=[y], induces a homomorphism of cohomology rings \pi^{*}_{2}:\operatorname{H}^{*}(BS^{1};\mathbb{F})\to\operatorname{H}^{*}_{S^{1}}(X;\mathbb{F}). We denote e:=\pi_{2}^{*}e_{B} as the fundamental class associated with the S^{1} -space X . The Fadell–Rabinowitz index of X is then defined by \operatorname{ind}_{\operatorname{FR}}(X;\mathbb{F})=\sup_{i\geq 0}\{i+1\mid e^{i}\neq 0\}, and we set \operatorname{ind}_{\operatorname{FR}}(\emptyset;\mathbb{F})=0 . In what follows, we focus on the case \mathbb{F}=\mathbb{Z}_{2} , in accordance with the results of [ Mat24 ] . Nevertheless, analogous statements remain valid over any field, since [ Mat24 , Theorem A] holds for arbitrary field coefficients. Definition 1.1 . Let K\subset\mathbb{R}^{2n} be a convex body, and let K^{\prime} be any translation of K whose interior contains the origin. We define the systolic S^{1} -index of K as \operatorname{ind}_{\operatorname{sys}}^{S^{1}}(K)=\operatorname{ind}_{\operatorname{FR}}(\operatorname{Sys}(\partial K^{\prime})), where the topology on \operatorname{Sys}(\partial K^{\prime}) is induced by the uniform norm. As the notation suggests, this definition does not depend on the choice of the translation K^{\prime} . This is a consequence of Theorem 1.2