Liouville theorem for fully nonlinear elliptic equations with the small oscillation and the periodicity in $x$ and the periodic right hand term

This paper establishes the existence and Liouville-type results for quadratic growth solutions to fully nonlinear elliptic equations with periodic coefficients and right-hand sides, proving that under a small oscillation assumption, such solutions decompose into the sum of a quadratic polynomial and a periodic function.

The Gist

Here is an explanation of the paper, translated from complex mathematical jargon into a story about building a house on a bumpy, repeating landscape.

The Big Picture: The "Perfectly Bumpy" Landscape

Imagine you are an architect trying to design a massive, infinite road (the solution $u$) that stretches forever in every direction. You have a specific rulebook (the equation $F$) that tells you how the road must curve based on the terrain underneath it.

The terrain has two special properties:

1. It repeats: The landscape is like a giant tiled floor. If you walk one mile north, the hills and valleys look exactly the same as they did where you started. This is the Periodicity.
2. It's slightly inconsistent: Even though the tiles repeat, the "texture" of the road rules changes slightly depending on exactly where you are standing on the tile. Sometimes the road wants to curve up a tiny bit more; sometimes a tiny bit less. This is the Oscillation.

The paper asks a big question: If we build a road that grows in a specific way (quadratic growth, meaning it gets steeper like a parabola as you go further out), what does that road actually look like?

The authors prove that no matter how complicated the rules are, as long as the "texture" changes aren't too wild (the "small oscillation" condition), the road will always settle into a very predictable pattern.

The Main Discovery: The "Three-Part Sandwich"

The paper's main result (The Liouville Theorem) says that any such road can be broken down into a simple Three-Part Sandwich:

$$\text{Total Road} = \text{The Big Curve} + \text{The Straight Line} + \text{The Wiggles}$$

1. The Big Curve (Quadratic Polynomial): This is the overall shape. Imagine a giant, smooth bowl or a hill. This represents the long-term trend of the road getting steeper.
2. The Straight Line (Linear Term): This is just a gentle slope, like a ramp.
3. The Wiggles (Periodic Function): This is the fun part. Because the terrain repeats, the road has to "wiggle" up and down to match the bumps in the ground. But these wiggles are perfectly repetitive. They don't get bigger or stranger as you go further out; they just loop forever.

The Analogy: Imagine driving a car on a highway that is being built over a repeating pattern of potholes and speed bumps.

• The Big Curve is the highway curving around a massive mountain.
• The Wiggles are the car bouncing up and down over the repeating speed bumps.
• The paper proves that if the speed bumps aren't too chaotic (small oscillation), the car's path is just the smooth mountain curve plus a predictable, repeating bounce. You don't need to know the exact location of every single pothole to understand the general shape of the drive.

The "Small Oscillation" Rule

Why do the authors need the "small oscillation" condition?

Imagine the terrain is a repeating pattern, but sometimes the "rules" for how the road behaves change drastically from one tile to the next. Maybe on Tile A, the road must be flat, but on Tile B, it must be vertical. If the rules jump around too wildly, the road might not be able to form a smooth, predictable shape. It might get twisted and knotted.

The authors say: "If the rules don't change too much from one spot to the next (small oscillation), then the road will always smooth itself out into that predictable 'Curve + Wiggle' shape."

The "Homogenization" Magic

The paper introduces a clever trick called Homogenization.

Think of it like this: You have a complex, bumpy recipe for baking a cake (the equation with the repeating terrain). It's hard to calculate the exact result for every single crumb.

• The authors create a Simplified Recipe (the "Homogenized Operator").
• They prove that if you bake the cake using the Simplified Recipe, you get the same overall shape as if you had used the complex, bumpy recipe.
• The "wiggles" are the only thing the simplified recipe misses, but the big picture (the curve) is identical.

This allows mathematicians to solve the easy version of the problem and know exactly what the hard version looks like.

Why Does This Matter? (The "So What?")

You might wonder, "Who cares about infinite roads?"

1. Materials Science: This helps engineers understand how materials behave. If you have a material made of repeating atoms (like a crystal), and you want to know how it bends under pressure, this math tells you the "average" behavior without needing to simulate every single atom.
2. Predictability: It proves that even in a chaotic, repeating world, there is an underlying order. If you know the "average" rules, you can predict the long-term behavior of complex systems.
3. Generalization: Previous math only worked for simple, straight-line roads or perfectly smooth tiles. This paper is a giant leap forward, showing that even with "bumpy" rules, the math still holds up.

Summary in One Sentence

The paper proves that if you are building a massive, curving structure on a repeating, slightly imperfect landscape, the structure will always settle into a simple shape: a smooth curve, a straight slope, and a repeating pattern of bumps, provided the imperfections in the landscape aren't too chaotic.

Technical Summary

Here is a detailed technical summary of the paper "Liouville Theorem for Fully Nonlinear Elliptic Equations with the Small Oscillation and the Periodicity in $x$ and the Periodic Right Hand Term" by Lichun Liang.

1. Problem Statement

The paper investigates the structure of quadratic growth solutions to fully nonlinear elliptic equations in the whole space $\mathbb{R}^n$ and on exterior domains. The primary equation studied is:
$$F(D^2u(x), x) = f(x), \quad x \in \mathbb{R}^n$$
where:

• $F: S^{n \times n} \times \mathbb{R}^n \to \mathbb{R}$ is a fully nonlinear operator.
• $f: \mathbb{R}^n \to \mathbb{R}$ is a periodic function (period $Z^n$).
• The operator $F$ is periodic in the spatial variable $x$.
• The solution $u$ satisfies a quadratic growth condition: $|u(x)| \leq C(1 + |x|^2)$.

Key Assumptions on $F$:

1. Uniform Ellipticity: $F$ satisfies the standard ellipticity condition with constants $0 < \lambda \leq \Lambda < \infty$.
2. Periodicity: $F(M, x+z) = F(M, x)$ for $z \in \mathbb{Z}^n$.
3. Concavity: $F$ is concave with respect to the matrix variable $M$.
4. Small Oscillation: The oscillation of $F$ with respect to $x$, denoted by $\beta(x, x_0)$, is "small" in an integral sense (specifically, satisfying a condition related to the $L^n$ norm over small cubes).

Goal: To establish Liouville-type theorems, proving that any such quadratic growth solution $u$ can be decomposed into a quadratic polynomial, a linear term, and a periodic function. Specifically, the paper aims to show:
$$u(x) = \frac{1}{2}x^T A x + b \cdot x + c + v(x)$$
where $A$ is a constant symmetric matrix, $b \in \mathbb{R}^n$, $c \in \mathbb{R}$, and $v$ is a periodic function with zero mean.

---

2. Methodology

The author employs a combination of techniques from homogenization theory, regularity theory for fully nonlinear equations, and blow-up analysis.

A. Homogenization and Cell Problems

The core of the methodology involves defining a homogenization operator $\bar{F}$.

• Cell Problem: For a fixed matrix $A$, the author considers the periodic corrector problem:
  $$F(A + D^2v(x), x) - \alpha = f(x) - \fint_{Q_1} f$$
  where $v$ is a periodic function with zero mean ($v \in \mathcal{T}$).
• Existence of Correctors: Using the method of continuity and Evans-Krylov theory (for $C^{2,\alpha}$ estimates), the paper proves the existence and uniqueness of the pair $(\alpha, v)$ under the "small oscillation" condition on $F$.
• Definition of $\bar{F}$: The homogenization operator is defined by $\bar{F}(A) = \alpha$. The paper establishes that $\bar{F}$ inherits the uniform ellipticity and concavity properties of $F$.

B. Regularity and Approximation

• Smooth Approximation: Since $F$ and $f$ are only assumed continuous (or Hölder continuous), the author uses mollification to approximate them by smooth functions.
• $W^{2,p}$ Estimates: A crucial step relies on the small oscillation condition (Condition 1.4) to ensure interior $W^{2,p}$ estimates ($p > n$) for viscosity solutions. This allows the use of strong maximum principles and compactness arguments.

C. Blow-Down Analysis (Liouville Theorem Proof)

To prove the structural decomposition of the solution $u$:

1. Rescaling: Define the blow-down sequence $u_R(x) = u(Rx)/R^2$.
2. Compactness: Using the $W^{2,p}$ estimates and the periodicity of the coefficients, the sequence $u_R$ is shown to converge (up to a subsequence) in $C^1_{loc}$ to a quadratic polynomial $Q(x) = \frac{1}{2}x^T A x$.
3. Limiting Equation: The limit $Q$ is shown to satisfy the homogenized equation $\bar{F}(D^2Q) = \fint_{Q_1} f$.
4. Difference Argument: By analyzing the difference $w = u - Q$ and utilizing the concavity of $F$ and the periodicity of $f$, the author reduces the problem to a linear elliptic equation with periodic coefficients.
5. Boundedness: Using the Harnack inequality and properties of difference quotients (following strategies by Caffarelli and Li), it is proven that the remainder term is bounded and periodic.

D. Exterior Domains

For equations on exterior domains ($\mathbb{R}^n \setminus \Omega$), the author uses:

• Comparison Principles: Constructing barriers using Pucci's extremal operators.
• Asymptotic Estimates: Deriving decay rates for the difference between the solution and the quadratic polynomial based on the ratio $\Lambda/\lambda$.

---

3. Key Contributions

1. Generalization of Liouville Theorems:
   The paper extends existing Liouville theorems from:

   • Linear equations with periodic coefficients (Li-Wang).
   • Fully nonlinear equations with constant coefficients or periodic right-hand side only (Li-Liang).
   • New Result: It handles the case where both the operator $F$ and the source term $f$ are periodic in $x$, provided the oscillation of $F$ in $x$ is small.

2. Small Oscillation Condition:
   The paper introduces a specific integral condition on the oscillation $\beta(x, x_0)$ (Condition 1.2/1.4). This condition is weaker than requiring $F$ to be constant or Lipschitz in $x$, yet strong enough to guarantee the necessary $W^{2,p}$ regularity and the validity of the homogenization process.

3. Construction of the Homogenization Operator:
   The paper rigorously defines the homogenization operator $\bar{F}$ for fully nonlinear equations with periodic data and proves its structural properties (ellipticity and concavity), which are essential for the asymptotic analysis.

4. Exterior Domain Asymptotics:
   The paper provides precise asymptotic behavior for solutions on exterior domains, establishing decay rates of the form $O(|x|^{1 - (n-1)\frac{\lambda}{\Lambda}})$ under the condition $\frac{\Lambda}{\lambda} < n-1$.

---

4. Main Results

• Theorem 1.2 & Corollary 1.3 (Existence of Correctors):
  Under the small oscillation assumption, for any matrix $A$, there exists a unique periodic function $v$ and a constant $\alpha$ (which defines $\bar{F}(A)$) solving the cell problem.

• Theorem 1.6 (Existence of Quadratic Solutions):
  A quadratic growth solution $u(x) = \frac{1}{2}x^T Ax + b \cdot x + c + v(x)$ exists if and only if $\bar{F}(A) = \fint_{Q_1} f$.

• Theorem 1.7 (Liouville Theorem):
  Any viscosity solution $u$ of $F(D^2u, x) = f$ with quadratic growth must be of the form:
  $$u(x) = \frac{1}{2}x^T Ax + b \cdot x + c + v(x)$$
  where $v$ is periodic, and $A$ satisfies $\bar{F}(A) = \fint_{Q_1} f$.

• Theorem 1.9 & 1.11 (Exterior Domains):
  Existence and uniqueness of solutions on exterior domains with prescribed asymptotic behavior at infinity. If $\Lambda/\lambda < n-1$, the solution converges to the quadratic polynomial plus periodic term with a specific algebraic decay rate.

---

5. Significance

• Theoretical Advancement: This work bridges the gap between linear periodic homogenization and fully nonlinear elliptic theory. It demonstrates that the "small oscillation" technique, previously used for regularity, is also a powerful tool for structural Liouville theorems in the fully nonlinear setting.
• Applications in Homogenization: The results provide the necessary foundation for understanding the limiting behavior of solutions in periodic media, particularly for nonlinear materials where the constitutive laws vary periodically.
• Relaxation of Smoothness: By utilizing viscosity solutions and approximation techniques, the paper relaxes the smoothness requirements on $F$ and $f$ compared to earlier works (e.g., requiring only continuity or Hölder continuity rather than $C^2$ or $C^\infty$), making the results applicable to a broader class of physical models.
• Connection to Variational Problems: The decomposition of solutions into polynomial and periodic parts is intimately linked to the study of minimal solutions in variational problems on tori, offering new insights into the dimension of spaces of polynomial growth solutions.

In summary, Lichun Liang's paper establishes a robust framework for analyzing the asymptotic structure of fully nonlinear elliptic equations with periodic coefficients and data, proving that under a small oscillation condition, the complex behavior of the solution is essentially governed by a homogenized quadratic polynomial and a periodic fluctuation.