Some Geometric Constants Related to the Midline of Equilateral Triangles in Banach Spaces

Bingren Chen, Zhijian Yang, Qi Liu, Yongjin LiSymmetry 2022 · DOI

TL;DR

This paper introduces and analyzes new geometric constants ($M_1(X)$, $M_2(X)$, $M_3(X)$) for Banach spaces, inspired by existing constants related to triangle geometry. These constants offer novel ways to characterize geometric properties like rotundity and uniform non-squareness, potentially leading to deeper understanding of Banach space geometry.

What Problem Does This Paper Address?

The study of geometric constants in Banach spaces is a rich area, with established constants like the Jordan-von Neumann constant and the James constant providing insights into a space's structure. Recently, constants like $H(X)$ and $A_2(X)$ have been introduced, focusing on geometric configurations within the unit sphere. This paper aims to extend this line of research by introducing new constants that are specifically related to the midlines of equilateral triangles inscribed within the unit sphere of a Banach space. The authors seek to explore the properties of these new constants and their relationships with existing geometric invariants, thereby enriching our understanding of Banach space geometry and potentially revealing new characterizations of important properties.

Methods

The paper introduces three new geometric constants: $M_1(X)$, $M_2(X)$, and $M_3(X)$.

  • $M_1(X)$: This constant is defined as the supremum of the average of two lengths related to an equilateral triangle inscribed in the unit sphere. Specifically, it considers the sum of the lengths of two midlines of such a triangle. The authors analyze its basic properties, its relationship with the hexagon constant $H(X)$, and its connection to uniform non-squareness.
  • $M_2(X)$: This constant is a quadratic form related to $M_1(X)$, defined using the sum of squares of lengths. The paper investigates its properties and its relationship with $M_1(X)$ and the modulus of smoothness $\rho_X(t)$.
  • $M_3(X)$: This constant is a subduplicate form of $M_1(X)$, involving the sum of the lengths themselves. Its properties and connections to $H(X)$ and $M_1(X)$ are explored.

The authors employ standard techniques in geometric functional analysis, including:

  • Supremum calculations: The constants are defined as suprema over specific sets of vectors in the unit sphere.
  • Inequalities: Various inequalities are derived to establish relationships between the new constants and existing ones (e.g., $H(X)$, $\rho_X(t)$).
  • Geometric interpretation: The constants are often explained through geometric constructions, particularly involving equilateral triangles in the Euclidean plane.
  • Characterization of properties: The paper demonstrates how the values of these new constants can characterize properties like rotundity and uniform non-squareness of a Banach space.

Datasets

Not discussed in the paper.

Key Insights and Ideas

  • Bridging Midlines and Geometric Constants: The core innovation is connecting the geometric concept of midlines of inscribed equilateral triangles to quantitative geometric constants in Banach spaces. This provides a new lens through which to view space geometry.
  • $M_1(X)$ as a Measure of "Triangle-ness": The constant $M_1(X)$ appears to capture how "equilateral-like" triangles can be formed within the unit sphere, and its value relates to how "round" or "square-like" the space is.
  • Interconnectedness of Geometric Constants: The paper highlights how these new constants are not isolated but are intricately linked to established constants like $H(X)$ and the modulus of smoothness $\rho_X(t)$. This suggests a deeper, unified theory of geometric constants.
  • New Characterizations of Uniform Non-Squareness: The paper demonstrates that specific bounds on $M_1(X)$, $M_2(X)$, and $M_3(X)$ are sufficient conditions for a Banach space to be uniformly non-square, offering alternative criteria to existing ones.
  • Generalization and Forms: The introduction of $M_2(X)$ (quadratic form) and $M_3(X)$ (subduplicate form) suggests that these geometric concepts can be generalized, leading to a family of related constants with potentially distinct properties.

Key Results

  • For an inner product space $X$, $M_1(X) = \sqrt{3}$.
  • If $M_1(X) < 2$, then $X$ is uniformly non-square.
  • If $M_2(X) < 4$, then $X$ is uniformly non-square.
  • If $M_3(X) < \sqrt{2}$, then $X$ is uniformly non-square.
  • Inequalities relating the new constants to $H(X)$ and $\rho_X(t)$ are established, such as $M_1(X) \leq \frac{H(X)}{2} + 1$ and $M_2(X) \leq 1 + 2\rho_X(1)$.
  • Specific examples are analyzed, such as $l_\infty$ and $\mathbb{R}^2$ with mixed norms, to illustrate the values of these constants. For instance, $M_1(l_\infty) = 2$ and $M_2(l_\infty - l_1) = 4$.

Limitations and Future Work

  • Open Problems: The authors explicitly state that there are "plenty of interesting problems that await discussion."
  • Further Characterizations: A key future direction is to explore how $M_1(X)$, $M_2(X)$, and $M_3(X)$ can be utilized to characterize more geometric properties of Banach spaces beyond rotundity and uniform non-squareness.
  • Broader Applications: The paper suggests that further research could delve deeper into the geometric interpretations and potential applications of these constants.
  • Generalization of Forms: While $M_2(X, p, q)$ is introduced as a generalized form of $M_2(X)$, a more systematic study of these generalized forms and their properties could be a fruitful avenue.

Summary updated

BibTeX
@article{chen2022some,
  author = {Bingren Chen and Zhijian Yang and Qi Liu and Yongjin Li},
  title = {{Some Geometric Constants Related to the Midline of Equilateral Triangles in Banach Spaces}},
  journal = {Symmetry},
  year = {2022},
  doi = {10.3390/sym14020348},
}
Figure