2025-11-15T15:25:11.913329

Nonlocal discrete diffusion equations and the fractional discrete Laplacian, regularity and applications

Ciaurri, Roncal, Stinga et al.
The analysis of nonlocal discrete equations driven by fractional powers of the discrete Laplacian on a mesh of size $h>0$ \[ (-Δ_h)^su=f, \] for $u,f:\mathbb{Z}_h\to\mathbb{R}$, $0<s<1$, is performed. The pointwise nonlocal formula for $(-Δ_h)^su$ and the nonlocal discrete mean value property for discrete $s$-harmonic functions are obtained. We observe that a characterization of $(-Δ_h)^s$ as the Dirichlet-to-Neumann operator for a semidiscrete degenerate elliptic local extension problem is valid. Regularity properties and Schauder estimates in discrete Hölder spaces as well as existence and uniqueness of solutions to the nonlocal Dirichlet problem are shown. For the latter, the fractional discrete Sobolev embedding and the fractional discrete Poincaré inequality are proved, which are of independent interest. We introduce the negative power (fundamental solution) \[ u=(-Δ_h)^{-s}f, \] which can be seen as the Neumann-to-Dirichlet map for the semidiscrete extension problem. We then prove the discrete Hardy--Littlewood--Sobolev inequality for $(-Δ_h)^{-s}$. As applications, the convergence of our fractional discrete Laplacian to the (continuous) fractional Laplacian as $h\to0$ in Hölder spaces is analyzed. Indeed, uniform estimates for the error of the approximation in terms of $h$ under minimal regularity assumptions are obtained. We finally prove that solutions to the Poisson problem for the fractional Laplacian \[ (-Δ)^sU=F, \] in $\mathbb{R}$, can be approximated by solutions to the Dirichlet problem for our fractional discrete Laplacian, with explicit uniform error estimates in terms of~$h$.
academic

Nonlocal discrete diffusion equations and the fractional discrete Laplacian, regularity and applications

Basic Information

  • Paper ID: 1608.08913
  • Title: Nonlocal discrete diffusion equations and the fractional discrete Laplacian, regularity and applications
  • Authors: Óscar Ciaurri, Luz Roncal, Pablo Raúl Stinga, José L. Torrea, Juan Luis Varona
  • Classification: math.AP cs.NA math.CA math.FA math.NA math.PR
  • Publication Date: March 21, 2018 (v2)
  • Paper Link: https://arxiv.org/abs/1608.08913

Abstract

This paper provides a comprehensive analysis of nonlocal discrete equations (Δh)su=f(-\Delta_h)^s u = f driven by fractional powers of the discrete Laplacian, where u,f:ZhRu,f: \mathbb{Z}_h \to \mathbb{R}, 0<s<10 < s < 1, and h>0h > 0 is the mesh size. The study obtains pointwise nonlocal formulas for (Δh)su(-\Delta_h)^s u and establishes nonlocal discrete mean value properties for discrete ss-harmonic functions. The authors prove that (Δh)s(-\Delta_h)^s can be characterized as the Dirichlet-to-Neumann operator of a semi-discrete degenerate elliptic extension problem. Regularity properties and Schauder estimates in discrete Hölder spaces are established, along with existence and uniqueness of solutions to nonlocal Dirichlet problems. Negative powers (fundamental solutions) u=(Δh)sfu = (-\Delta_h)^{-s}f are introduced, and the corresponding discrete Hardy-Littlewood-Sobolev inequality is proved. As applications, the convergence of the fractional discrete Laplacian to the continuous fractional Laplacian in Hölder spaces as h0h \to 0 is analyzed, and it is shown that solutions to the fractional Laplacian Poisson problem can be approximated by solutions to the fractional discrete Laplacian Dirichlet problem.

Research Background and Motivation

Problem Background

The fractional Laplacian operator, as a positive power of the classical Laplacian, has long existed in multiple mathematical fields including potential theory, harmonic analysis, fractional calculus, functional analysis, and probability theory. Over the past decade, nonlocal diffusion equations involving the fractional Laplacian have become one of the most active research topics in the field of partial differential equations.

Core Problems

  1. Discretization Challenges: The nonlocal and singular nature of the fractional Laplacian (Δ)s(−\Delta)^s presents major difficulties for numerical approximation
  2. Theoretical Gaps: Lack of a systematic theoretical framework for discrete fractional Laplacian operators
  3. Convergence Analysis: Need for rigorous theoretical guarantees on the convergence of discrete models to continuous models

Research Motivation

  • Establish a complete theoretical system for discrete fractional Laplacian operators
  • Provide effective discrete approximation methods for continuous problems
  • Provide explicit error estimates under minimal regularity assumptions

Core Contributions

  1. Established a complete theoretical framework for discrete fractional Laplacian operators, including pointwise formulas, kernel function estimates, and fundamental properties
  2. Proved regularity theory in discrete Hölder spaces, including Schauder estimates and maximum principles
  3. Established discrete fractional Sobolev embeddings and Poincaré inequalities, providing theoretical foundations for nonlocal Dirichlet problems
  4. Provided discrete Hardy-Littlewood-Sobolev inequalities, completing the discrete fractional integral theory
  5. Proved uniform convergence to the continuous fractional Laplacian, with explicit error estimates in Hölder spaces
  6. Established discrete approximation theory for continuous Poisson problem solutions, providing error estimates in the LL^\infty norm

Methodology Details

Problem Formulation

Study nonlocal discrete diffusion equations on the grid Zh={hj:jZ}\mathbb{Z}_h = \{hj : j \in \mathbb{Z}\}: (Δh)su=f(-\Delta_h)^s u = f where u,f:ZhRu, f: \mathbb{Z}_h \to \mathbb{R}, 0<s<10 < s < 1, and h>0h > 0.

Core Technical Framework

1. Semigroup Method Definition

The fractional power of the discrete Laplacian is defined via the semigroup method: (Δh)suj=1Γ(s)0(etΔhujuj)dtt1+s(-\Delta_h)^s u_j = \frac{1}{\Gamma(-s)} \int_0^\infty \left(e^{t\Delta_h}u_j - u_j\right) \frac{dt}{t^{1+s}}

where wj(t)=etΔhujw_j(t) = e^{t\Delta_h}u_j is the solution to the semi-discrete heat equation:

\partial_t w_j = \Delta_h w_j, & \text{on } \mathbb{Z}_h \times (0,\infty) \\ w_j(0) = u_j, & \text{on } \mathbb{Z}_h \end{cases}$$ #### 2. Pointwise Nonlocal Formula **Theorem 1.1**: For $0 < s < 1$ and $u \in \ell_s$: $$(-\Delta_h)^s u_j = \sum_{m \in \mathbb{Z}, m \neq j} (u_j - u_m) K_s^h(j-m)$$ where the discrete kernel function is: $$K_s^h(m) = \frac{4^s \Gamma(1/2 + s)}{\sqrt{\pi}|\Gamma(-s)|} \cdot \frac{\Gamma(|m| - s)}{h^{2s}\Gamma(|m| + 1 + s)}$$ #### 3. Kernel Function Estimates There exist constants $0 < c_s \leq C_s$ such that: $$\frac{c_s}{h^{2s}|m|^{1+2s}} \leq K_s^h(m) \leq \frac{C_s}{h^{2s}|m|^{1+2s}}$$ ### Technical Innovations #### 1. Explicit Expression of Semi-discrete Heat Kernel Precise formula for the semi-discrete heat kernel using modified Bessel functions: $$G(m,t) = e^{-2t}I_m(2t)$$ #### 2. Extension Problem Characterization Characterize $(-\Delta_h)^s$ as the Dirichlet-to-Neumann map of a semi-discrete degenerate elliptic extension problem, establishing connections between local and nonlocal operators. #### 3. Discrete Hölder Space Theory Establish complete theory for discrete Hölder spaces $C_h^{k,\alpha}$, proving regularity of the fractional discrete Laplacian: **Theorem 1.5**: If $u \in C_h^{k,\alpha}$ and $2s < \alpha$, then $(-\Delta_h)^s u \in C_h^{k,\alpha-2s}$. ## Experimental Setup ### Theoretical Verification Methods This paper is primarily theoretical research, with results verified through: 1. **Limiting Behavior Verification**: - As $s \to 0^+$: $(-\Delta_h)^s u_j \to u_j$ - As $s \to 1^-$: $(-\Delta_h)^s u_j \to -\Delta_h u_j$ 2. **Consistency Checks**: Verify consistency between discrete and continuous operators under mesh refinement 3. **Kernel Function Properties**: Verify symmetry, positivity, and asymptotic behavior of kernel functions ### Evaluation Metrics - **Convergence Rate**: Convergence speed in Hölder norm - **Error Estimates**: Explicit error bounds in $L^\infty$ norm - **Regularity Indicators**: Norm estimates in discrete Hölder spaces ## Experimental Results ### Main Theoretical Results #### 1. Uniform Approximation Estimates **Theorem 1.7**: For $U \in C^{0,\alpha}$ with $2s < \alpha$: $$\|(-\Delta_h)^s(r_h U) - r_h((-\Delta)^s U)\|_{\ell^\infty_h} \leq C[U]_{C^{0,\alpha}} h^{\alpha-2s}$$ #### 2. Convergence of Discrete Solutions **Theorem 1.8**: Let $F \in C^{0,\alpha}$ have compact support, $U$ be the solution to the continuous Poisson problem $(-\Delta)^s U = F$, and $u$ be the solution to the corresponding discrete Dirichlet problem. Then: $$\|u - r_h U\|_{\ell^\infty_h(B_R^h)} \leq C\|F\|_{C^{0,\alpha}} R^{2s} h^\alpha$$ #### 3. Hardy-Littlewood-Sobolev Inequality for Fundamental Solutions For $1 < p < q < \infty$ with $1/q \leq 1/p - 2s$: $$\|(-\Delta_h)^{-s}f\|_{\ell^q_h} \leq \frac{C}{h^{1/p-2s-1/q}} \|f\|_{\ell^p_h}$$ ### Key Findings 1. **Optimal Convergence Rate**: Achieves theoretically optimal convergence speed under minimal regularity assumptions 2. **Uniform Estimates**: All constants are independent of mesh size $h$ 3. **Boundary Effects**: The $R^{2s}$ factor appearing in the discrete maximum principle reflects the essential characteristics of nonlocal operators ## Related Work ### Main Research Directions 1. **Continuous Fractional Laplacian Theory**: Caffarelli-Silvestre extension problem, Silvestre regularity theory 2. **Numerical Methods**: Finite difference, finite element, spectral methods and other discretization techniques 3. **Probabilistic Interpretation**: Connections with jump processes and Lévy processes ### Advantages of This Work - First to establish a complete theoretical system for discrete fractional Laplacian operators - Provides rigorous error analysis and convergence proofs - Gives explicit error estimates in $L^\infty$ norm, which is novel in this field ## Conclusions and Discussion ### Main Conclusions 1. Established complete mathematical theory for discrete fractional Laplacian operators 2. Proved uniform convergence of discrete models to continuous models 3. Provided effective numerical approximation methods for fractional Poisson problems ### Limitations 1. **Dimensional Restrictions**: Main results focus on one-dimensional cases; higher-dimensional extensions require new techniques 2. **Parameter Range**: Some results require $s < 1/2$, limiting applicability 3. **Grid Regularity**: Methods apply to uniform grids; non-uniform grids require additional considerations ### Future Directions 1. **Higher-Dimensional Extensions**: Develop theory for multidimensional discrete fractional Laplacian operators 2. **Optimal Regularity**: Determine optimal Hölder regularity of solutions to discrete Dirichlet problems 3. **Adaptive Methods**: Develop adaptive mesh methods based on error estimates ## In-Depth Evaluation ### Strengths 1. **Theoretical Completeness**: Establishes a systematic and complete theoretical framework with rigorous arguments from basic definitions to applications 2. **Technical Innovation**: Combines semigroup methods with Bessel function techniques, providing exact kernel function expressions 3. **Practical Value**: Provides implementable numerical methods with explicit error control 4. **Writing Quality**: Clear paper structure, rigorous proofs, and careful treatment of technical details ### Weaknesses 1. **Computational Complexity**: Kernel functions involve Gamma and Bessel functions, resulting in high computational costs 2. **Application Verification**: Lacks concrete numerical experiments validating theoretical predictions 3. **Comparative Analysis**: Insufficient comparison with other discretization methods ### Impact 1. **Academic Contribution**: Establishes important foundations for discretization theory of nonlocal operators 2. **Application Prospects**: Provides theoretical support for numerical solution of fractional differential equations 3. **Reproducibility**: Complete theoretical results facilitate follow-up research and extensions ### Applicable Scenarios - Numerical solution of fractional diffusion equations - Discretization analysis of nonlocal models - Numerical simulation of jump processes - Non-local filter design in image processing ## References The paper cites important literature in the field, including: - Caffarelli & Silvestre (2007): Extension problem for fractional Laplacian - Silvestre (2007): Regularity theory for fractional Laplacian - Savin & Valdinoci (2011): Sobolev inequalities for nonlocal variational models - Stein & Wainger (2000): Discrete analogues in harmonic analysis --- **Overall Assessment**: This is a high-quality theoretical mathematics paper that systematically establishes the mathematical theory of discrete fractional Laplacian operators, providing important theoretical foundations for numerical analysis of nonlocal differential equations. The paper achieves high standards in both technical depth and theoretical completeness, with significant implications for advancing research in this field.