Non-linear Gagliardo--Nirenberg inequality involving a second-order elliptic operator in non-divergent form

Basic Information

• Paper ID: 2308.00545
• Title: Non-linear Gagliardo--Nirenberg inequality involving a second-order elliptic operator in non-divergent form
• Authors: Agnieszka Kałamajska (University of Warsaw), Dalimil Peša (Charles University), Tomáš Roskovec (University of South Bohemia)
• Classification: math.AP (Analysis of PDEs)
• Publication Date: Submitted August 2023, revised June 6, 2025
• Paper Link: https://arxiv.org/abs/2308.00545v3

Abstract

This paper establishes a non-linear inequality of the form: $\int_\Omega |\nabla u(x)|^2h(u(x)) dx \leq C\int_\Omega \left(\sqrt{|Pu(x)||T_H(u(x))|}\right)^2h(u(x)) dx + \Theta$

where $\Omega \subset \mathbb{R}^n$ is a bounded Lipschitz domain, $u \in W^{2,1}_{loc}(\Omega)$ is a non-negative function, $P$ is a uniformly elliptic operator in non-divergent form, $T_H(\cdot)$ is a certain transformation of the monotone $C^1$ function $H(\cdot)$ (where $H$ is the antiderivative of the weight function $h$), and $\Theta$ is a boundary term depending on the boundary values of $u$ and $\nabla u$. These results are connected to certain results in probability theory and potential theory, such as certain variants of the Douglas formula.

Research Background and Motivation

Importance of the Problem

1. A priori estimates for elliptic partial differential equations: One of the main motivations is to obtain a priori estimates for solutions of non-linear partial differential equations, particularly for second-order elliptic equations of the form $Pu = f(x)u^{-\gamma}$.
2. Applications in harmonic analysis and potential theory: The research establishes connections with results in probability theory and potential theory, particularly with theories related to analytic semigroup generators.
3. Significance of non-divergent form operators: Second-order operators in non-divergent form play a crucial role in elliptic PDE theory, constitute an important subclass of Feller semigroup generators in probability theory, and appear in the generators of Itô diffusion processes and Kolmogorov backward equations.

Limitations of Existing Methods

Existing similar inequalities are primarily focused on:

• The classical Laplace operator case
• Requiring stronger regularity assumptions (e.g., $u \in W^{2,1}(\Omega) \cap C(\bar{\Omega})$)
• Mainly restricted to divergence form operators

The innovation of this paper lies in handling general elliptic operators in non-divergent form and establishing results under weaker assumptions ($u \in W^{2,1}_{loc}(\Omega)$ and $\tilde{H}(u) \in W^{2,1}(\Omega)$).

Core Contributions

1. Establishes non-linear Gagliardo-Nirenberg type inequalities involving elliptic operators in non-divergent form, generalizing previous results that only applied to the Laplace operator.
2. Proves identities and inequalities under weaker regularity assumptions, allowing functions to take values 0 or boundary values on sets of positive measure in the domain.
3. Provides two methods for simplifying the inequality:
   • Based on sign conditions and Dirichlet boundary conditions
   • Based on Opial-type inequalities
4. Establishes new Opial-type inequalities, which have independent mathematical significance.
5. Reveals deep connections with the Douglas formula, probability theory, and potential theory.

Detailed Methodology

Basic Setup

Domain Assumption (Ω): $\Omega \subseteq \mathbb{R}^n$ ($n \geq 2$) is a bounded Lipschitz domain.

Elliptic Operator: A uniformly elliptic operator in non-divergent form $Pu = \sum_{i,j} a_{i,j}(x)\frac{\partial^2 u}{\partial x_i \partial x_j}(x)$

where the matrix $A(x) = \{a_{i,j}(x)\}$ satisfies the uniform ellipticity condition: $c_A\|\xi\|^2 \leq \xi^T A(x)\xi \leq C_A\|\xi\|^2$

Weight Function System:

• Primary weight function $h: (0,B) \to (0,\infty)$
• First antiderivative $H$ satisfying $H' = h$
• Second antiderivative $\tilde{H}$ satisfying $\tilde{H}'' = h$
• Transformations $T_H(s) = \frac{H(s)}{h(s)}$, $G_H(s) = \frac{H^2(s)}{h(s)}$

Main Theorems

Theorem 3.1 (Fundamental Identity and Inequality): Under assumptions (G) and (u):

1. Identity: $\int_\Omega \|\nabla u\|^2_A h(u(x)) dx = -\int_\Omega Pu H(u) dx - \int_\Omega \text{div}A \cdot \nabla u H(u) dx + \Theta$

where the boundary term $\Theta = \int_{\partial\Omega} n(x)^T A(x)\nabla(\tilde{H}(u)) d\sigma(x)$.

2. Inequality:

• When $\text{div}A \equiv 0$: $\int_\Omega \|\nabla u\|^2_A h(u) dx \leq \int_\Omega |Pu||H(u)| dx + \Theta$
• General case: $\int_\Omega \|\nabla u\|^2_A h(u) dx \leq d_A \int_\Omega G_H(u) dx + 2\int_\Omega |Pu||H(u)| dx + 2\Theta$

Technical Innovations

1. Handling of chain rule: The key insight is that although the pointwise chain rule $P(\tilde{H}(u)) = \tilde{H}'(u)Pu$ fails, one can establish integral form upper bound estimates.
2. Analysis of boundary terms: Through refined trace theory analysis, boundary terms $\Theta$ are handled even when $u \notin W^{1,1}(\Omega)$.
3. Establishment of Opial-type inequalities: New Opial-type inequalities are proven: $\int_{\Omega \cap \{0<u\}} |T_H(u)|^2 h(u) dx \leq C_P C_{\tilde{H}} \int_{\Omega \cap \{0<u\}} \|\nabla u\||T_H(u)|h(u) dx$

Experimental Setup

This is a pure theoretical mathematics paper with no numerical experiments. The theoretical results are verified through the following approaches:

Theoretical Verification Methods

1. Construction of concrete examples:
   • Example 3.7: Construction of $u(x) = (1-|x|)^\alpha$ ($\alpha < 0$), $h(s) = s^\beta$ ($\beta < 0$)
   • Example 3.8: One-dimensional case $u(x) = \text{sgn}(x)|x|^{1/2+\varepsilon} + 1$
2. Verification of conditions: Verification that the assumption $\tilde{H}(u) \in W^{2,1}(\Omega)$ does not imply $u \in W^{2,1}(\Omega)$.
3. Analysis of boundary cases: Analysis of results under various boundary conditions.

Main Results

Establishment of Core Inequalities

Theorem 4.1 (Simplification based on sign conditions): Under additional assumptions, the $G_H$ term can be eliminated: $\int_{\Omega \cap \{u \in (0,B)\}} \|\nabla u\|^2_A h(u) dx \leq \int_{\Omega \cap \{u \in (0,B)\}} |Pu||H(u)| dx$

Theorem 4.4 (Simplification based on Opial inequality): When $\kappa = \|\text{div}A\|_{L^\infty} c_A^{-1} C_P^2 C_{\tilde{H}}^2 < 1$: $\int_{\Omega \cap \{u \in (0,B)\}} \|\nabla u\|^2_A h(u) dx \leq \frac{1}{1-\kappa}\left(\int_{\Omega \cap \{u \in (0,B)\}} |Pu||H(u)| dx + \Theta\right)$

Opial-type Inequalities

Theorem 4.3: New Opial-type inequalities are established, which have independent mathematical value: $\int_{\Omega \cap \{u \in (0,B)\}} G_H(u) dx \leq C_P C_{\tilde{H}} \int_{\Omega \cap \{u \in (0,B)\}} \|\nabla u\||T_H(u)|h(u) dx$

Related Work

Historical Development

1. Mazya's pioneering work (1985): Established inequalities in the one-dimensional case
2. Work by Kalamajska et al. (2012-2019): Developed the theory for the Laplace operator case
3. Metafune and Spina (2008): Related identities in $L^p$ semigroup theory

Connections with Probability Theory

• Douglas formula: Integral identities related to harmonic functions
• Feller semigroups: Non-divergent operators as generators
• Diffusion processes: Infinitesimal generators of Itô diffusions

Connections with Potential Theory

• Sobolev-Bregman forms: Related to boundary integrals
• Harmonic extensions: Poisson integral theory
• Feller kernels: Boundary measure theory

Conclusions and Discussion

Main Conclusions

1. Successfully generalizes the Gagliardo-Nirenberg inequality to the case of elliptic operators in non-divergent form.
2. Establishes the theory under weaker assumptions, allowing more general function classes.
3. Provides practical simplification methods that can eliminate complex terms in concrete applications.
4. Establishes connections with multiple branches of mathematics, revealing deeper mathematical structures.

Limitations

1. Non-negativity assumption: The requirement $u \geq 0$ limits the scope of applications.
2. Complexity of boundary conditions: The handling of boundary terms $\Theta$ remains complex.
3. Restrictiveness of condition (GH): The condition $G_H(s) \leq C_{\tilde{H}}|\tilde{H}(s)|$ does not hold for all functions.

Future Directions

The paper raises four open problems:

1. Extension to non-local operators: Such as fractional Laplace operators
2. Extension to $\Omega = \mathbb{R}^n$: The unbounded domain case
3. Removal of sign conditions: Allowing functions to change sign
4. Douglas-type representation: More precise characterization of boundary terms

In-Depth Evaluation

Strengths

1. Strong theoretical innovation: First systematic treatment of non-linear inequalities for elliptic operators in non-divergent form.
2. Sophisticated technical handling: Cleverly addresses the difficulty of chain rule failure by circumventing pointwise difficulties through integral estimates.
3. Good practical utility of results: Provides multiple simplification methods adapted to different application needs.
4. Rich mathematical connections: Establishes meaningful connections with multiple branches of mathematics.
5. Clear and rigorous writing: The paper has clear structure, detailed proofs, and explicit assumptions.

Weaknesses

1. Limited scope of applications: Non-negativity and other technical assumptions limit the range of applications.
2. Computational complexity: Verifying various assumptions in practical applications may be difficult.
3. Lack of numerical verification: As a pure theoretical work, it lacks numerical examples to verify the sharpness of the theory.

Impact

1. High academic value: Provides new tools for partial differential equation theory.
2. Large application potential: Has promising applications in regularity theory of non-linear elliptic equations.
3. Theoretical completeness: Fills the theoretical gap for the non-divergent operator case.

Applicable Scenarios

1. Non-linear elliptic equations: Particularly equations involving singular non-linear terms
2. Probability theory: Diffusion processes and semigroup theory
3. Potential theory: Harmonic analysis and integral operator theory
4. Mathematical physics: Physical models involving elliptic operators

References

The paper cites 39 important references, covering:

• Classical Sobolev space theory (Leoni, Maz'ja, etc.)
• Development of Gagliardo-Nirenberg inequalities (Gagliardo, Nirenberg, etc.)
• Related results in probability theory (Fukushima, Jacob, etc.)
• Partial differential equation theory (Gilbarg-Trudinger, etc.)

This paper makes important contributions to the field of mathematical analysis, particularly at the intersection of non-linear functional inequalities and elliptic operator theory. Both its theoretical innovation and technical depth reach a high level, laying a foundation for further development in related fields.