The Turán and Delsarte problems and their duals

Basic Information

• Paper ID: 2510.10172
• Title: The Turán and Delsarte problems and their duals
• Authors: Mihail N. Kolountzakis, Nir Lev, Máté Matolcsi
• Classification: math.CA (Classical Analysis), math.MG (Metric Geometry)
• Publication Date: October 11, 2025
• Paper Link: https://arxiv.org/abs/2510.10172v1

Abstract

This paper investigates two optimization problems for positive definite functions on Euclidean space: the Turán problem and the Delsarte problem, where these functions are subject to restrictions on their support and sign. These problems have been extensively studied due to their connections with tiling and packing problems in geometry. While weak and strong linear duality for these problems are automatic in the finite group setting, the authors establish these properties in the continuous setting, demonstrate the existence of extremal functions for these problems and their duals, and establish tiling-type relationships between the extremal functions of each problem and the extremal measures or distributions of the dual problem. The results are subsequently applied to convex bodies, proving that for convex bodies that cannot tile space, the Delsarte packing bound is strictly superior to the trivial volume packing bound.

Research Background and Motivation

Importance of the Problems

1. Turán Extremal Problem: Studies the maximization of integrals of positive definite functions subject to support constraints, a classical problem in harmonic analysis
2. Delsarte Problem: Has important applications in sphere packing density estimates, kissing number problems, and other geometric problems
3. Duality Theory: While duality is automatic in the finite group setting, it requires deep theoretical analysis in the continuous setting

Limitations of Existing Methods

• Duality may fail in infinite-dimensional linear programming
• The existence of extremal functions is not obvious in the continuous setting
• Lack of a unified theoretical framework for handling Turán and Delsarte problems

Research Motivation

To establish a complete theoretical framework for Turán and Delsarte problems in the continuous setting, including weak and strong duality, existence of extremal functions, and to explore the deep connections with geometric tiling problems.

Core Contributions

1. Proved strong linear duality in the continuous setting: Under appropriate geometric conditions, both Turán and Delsarte problems satisfy $T(U)T'(U) = 1$ and $D(U)D'(U) = 1$
2. Established existence of extremal functions: Proved that both the primal and dual problems admit extremal functions
3. Revealed tiling-type relationships between extremal functions: Such as $f \cdot \alpha = \delta_0$ and $\hat{f} \cdot \hat{\alpha} = \delta_0$
4. Applied to convex body theory: Proved that the Delsarte bound for convex bodies that cannot tile space is strictly superior to the volume bound
5. Connected spectrality, tiling, and optimization problems: Established profound connections between these concepts

Methodology Details

Problem Formulation

Turán Problem: Given an open set $U \subset \mathbb{R}^d$, define the Turán constant as $T(U) = \sup\left\{\int f : f(0) = 1, f = 0 \text{ on } U^c, \hat{f} \geq 0\right\}$

Delsarte Problem: Define the Delsarte constant as $D(U) = \sup\left\{\int f : f(0) = 1, f \leq 0 \text{ on } U^c, \hat{f} \geq 0\right\}$

Construction of Dual Problems

Dual Turán Problem: $T'(U) = \sup\{\hat{\alpha}(\{0\}) : \alpha = \delta_0 + \beta, \text{supp}(\beta) \subset U^c, \hat{\alpha} \geq 0\}$

Dual Delsarte Problem: $D'(U) = \sup\{\hat{\alpha}(\{0\}) : \alpha = \delta_0 + \beta, \beta \geq 0, \text{supp}(\beta) \subset U^c, \hat{\alpha} \geq 0\}$

Technical Innovations

1. Boundary condition handling: Introduces the concept of "continuous boundary," requiring local representability as the graph of a continuous function
2. Approximation techniques: Uses Schwartz function approximation to handle products of continuous functions and tempered distributions
3. Application of Hahn-Banach separation theorem: Establishes duality in the infinite-dimensional setting
4. Translation-bounded measure theory: Handles tempered distributions whose Fourier transforms are measures

Main Theoretical Results

Weak Linear Duality

Theorems 4.3, 5.3: For open sets $U$ satisfying appropriate conditions, $T(U)T'(U) \leq 1, \quad D(U)D'(U) \leq 1$

Strong Linear Duality

Theorems 4.7, 5.4: Under stronger geometric conditions, equality holds: $T(U)T'(U) = 1, \quad D(U)D'(U) = 1$

Existence of Extremal Functions

Theorems 4.9, 5.6: Both the primal and dual problems admit extremal functions.

Relationships Between Extremal Functions

Theorems 4.10, 5.8: If $f$ and $\alpha$ are extremal functions for the primal and dual problems respectively, then:

• $\hat{f} \cdot \hat{\alpha} = \delta_0$
• $f \cdot \alpha = \delta_0$ (in the Delsarte case)

Geometric Applications

Packing Density Estimates

Theorem 6.1: The translational packing density of any set $A$ does not exceed $D(\Delta(A))^{-1}$, where $\Delta(A)$ is the essential difference set.

Characterization of Tiling and Spectrality

Theorems 6.2, 6.3:

• If $A$ tiles space, then $D(\Delta(A)) = m(A)$
• If $A$ is a spectral set, then $D(\Delta(A)) = m(A)$

Complete Characterization for Convex Bodies

Theorem 6.4: For a convex body $A$, the equality $D(\Delta(A)) = m(A)$ holds if and only if $A$ tiles space.

Corollary 6.5: The Delsarte bound for convex bodies that cannot tile space is strictly superior to the volume bound.

Technical Details

Key Lemmas

1. Lemma 4.4: Schwartz function approximation under continuous boundary conditions
2. Lemma 4.5: Establishment of translation-boundedness
3. Lemma 4.6: Convolution relation $\hat{f} * \hat{\alpha} = 1$ a.e.

Proof Strategy

1. Uses separation theorems to establish necessary conditions for duality
2. Establishes existence through approximation and compactness arguments
3. Utilizes analysis of extremality conditions to establish precise relationships between functions

Related Work

Historical Development

• Turán problem originates from the theory of trigonometric series
• Delsarte problem has applications in coding theory and sphere packing
• Complete theory for the finite group case (Matolcsi-Ruzsa 2014)

Relationship to This Work

This paper successfully extends the finite group theory to the continuous setting, resolving long-standing technical difficulties.

Conclusions and Discussion

Main Conclusions

1. Established a complete duality theory for Turán and Delsarte problems in the continuous setting
2. Proved existence of extremal functions and tiling-type relationships between them
3. Obtained profound geometric applications in convex body theory

Limitations

1. Requires relatively strong geometric conditions (such as continuous boundary)
2. Some results may not hold for general open sets
3. Computing specific constant values remains difficult

Future Directions

1. Finding examples of non-Turán domains
2. Extending to more general locally compact abelian groups
3. Exploring connections with other geometric optimization problems

In-Depth Evaluation

Strengths

1. Theoretical completeness: Establishes a complete duality framework in the continuous setting
2. Technical innovation: Skillfully addresses technical difficulties in infinite-dimensional linear programming
3. Geometric insight: Reveals deep connections between optimization problems and geometric properties
4. Practical value: Provides new tools in packing theory

Weaknesses

1. Geometric conditions: Requires relatively strong geometric conditions on open sets, limiting applicability
2. Computational complexity: While establishing the theoretical framework, concrete computation remains difficult
3. Open problems: Some important questions (such as the existence of non-Turán domains) remain unresolved

Impact

This represents significant progress at the intersection of harmonic analysis, convex geometry, and optimization theory, providing powerful theoretical tools for related research.

Application Scenarios

1. Theoretical analysis of sphere packing density
2. Study of geometric properties of convex bodies
3. Extremal problems in harmonic analysis
4. Coding theory and discrete geometry

References

The paper cites important literature in the field, including:

• Original work by Delsarte Del72, DGS77
• Applications by Cohn-Elkies in sphere packing CE03
• Breakthrough results by Viazovska in dimensions 8 and 24 Via17, CKMRV17
• Prior work by the authors on the Fuglede conjecture LM22