Weighted Bounded Variation Revisited

Basic Information

• Paper ID: 2510.14105
• Title: Weighted Bounded Variation Revisited
• Authors: Simon Bortz, Matthew Gossett, Joseph Kasel, Kabe Moen
• Classification: math.CA (Classical Analysis), math.FA (Functional Analysis)
• Publication Date: October 15, 2025 (arXiv preprint)
• Paper Link: https://arxiv.org/abs/2510.14105

Abstract

This paper revisits the theory of weighted bounded variation (BV) functions introduced by Baldi Ba01. The authors impose lower semicontinuity and/or pointwise A₁ conditions on weight functions as required by different theorems. The research is motivated by two primary objectives: establishing weighted Gagliardo-Nirenberg-Sobolev (GNS) inequalities for BV functions and clarifying and extending earlier results on weighted BV spaces. The main contributions include structure theorems under minimal assumptions (lower semicontinuity), smooth approximation results, embedding theorems, weighted GNS inequalities for BV functions, and corresponding weighted isoperimetric inequalities.

Research Background and Motivation

Problem Description

The core problem addressed in this paper concerns the theory of bounded variation function spaces under measure transformations. The bounded variation (BV) space comprises functions whose distributional derivatives are Radon measures, providing a more flexible framework than Sobolev spaces that can handle more singular functions (such as characteristic functions χₑ of sets E with finite perimeter).

Significance

1. Theoretical Value: BV spaces play a central role in generalized solutions of partial differential equations, surface measure theory, and isoperimetric inequalities
2. Applied Value: Important applications in total variation denoising, Mumford-Shah functionals, image processing, and segmentation
3. Extensibility: Weighted BV spaces, as natural extensions of weighted Sobolev spaces W^{1,1}(w), possess significant theoretical importance

Limitations of Existing Methods

Although Baldi's Ba01 work is an important reference in this field, it has the following shortcomings:

1. Structure theorems are limited to A₁* weights, whereas this paper considers only positive lower semicontinuous weights
2. Lack of systematic treatment of finite w-perimeter sets
3. Incomplete density theorems
4. Absence of applications in GNS and isoperimetric inequalities

Core Contributions

1. Structure Theorem: Establishes a structure theorem for BV_(Ω;w) under minimal assumptions (lower semicontinuity), generalizing Baldi's A₁* weight restriction
2. Smooth Approximation: Establishes smooth approximation theorems for weighted BV functions, introducing the w-approximability condition
3. Embedding Theorem: Proves that BV(Ω;w) can be isometrically embedded into higher-dimensional unweighted BV spaces
4. Weighted GNS Inequality: Establishes Gagliardo-Nirenberg-Sobolev inequalities for BV(ℝⁿ;w) functions
5. Weighted Isoperimetric Inequality: Derives corresponding isoperimetric inequalities as corollaries of the GNS inequality

Detailed Methodology

Fundamental Definitions and Notation

Weighted BV Spaces

For a weight w : ℝⁿ → (0,∞] and an open set Ω ⊆ ℝⁿ, define:

• f ∈ BV(Ω;w) if and only if f ∈ L¹(Ω;w) and $\|Df\|_w(Ω) := \sup\left\{\int_Ω f \text{div}\phi dx : \phi \in \text{Lip}_c(Ω;ℝⁿ), |\phi| ≤ w\right\} < ∞$

A₁ Weights

A weight w ∈ A₁ if and only if there exists a constant C > 0 such that for all balls B ⊆ ℝⁿ: $\fint_B w dx ≤ C \inf_{x∈B} w(x)$

Core Technical Methods

1. Structure Theorem (Theorem 1.1)

Theorem: Let w : ℝⁿ → (0,∞] be lower semicontinuous and f ∈ BV_(Ω;w). Then there exist a Radon measure ‖Df‖_w and a ‖Df‖_w-measurable function ν : Ω → ℝⁿ such that:

• |ν(x)| = 1 ‖Df‖_w-a.e.
• For all φ ∈ Lip_c(Ω;ℝⁿ): $\int_Ω f \text{div}\phi dx = -\int_Ω (\phi \cdot ν) \frac{1}{w} d\|Df\|_w$

Key Technique: By establishing the relationship between weighted and unweighted variation measures: d‖Df‖_w = w d‖Df‖, the problem is transformed into classical BV theory.

2. Smooth Approximation Theorem (Theorem 1.2)

Introduces the w-approximability condition: $\lim_{ε→0} \fint_{B(x,ε)} |w(y) - w(x)| dy = 0 \text{ for } \|Df\|-a.e. x$

Theorem: Let w ∈ A₁*, f ∈ BV(Ω;w):

• If f is w-approximable, then there exists a sequence {f_k} ⊆ C^∞(Ω)∩BV(Ω;w) such that f_k → f in L¹(Ω;w) and ‖Df_k‖_w(Ω) → ‖Df‖_w(Ω)
• If f is not w-approximable, then there exists a similar sequence but ‖Df‖_w(Ω) ≤ lim‖Df_k‖w(Ω) ≤ w{A₁}‖Df‖_w(Ω)

3. Isometric Embedding Theorem (Theorem 1.6)

Constructs a mapping J : BV(Ω;w) → BV(Ω_w), where Ω_w = {(x,y) ∈ ℝⁿ×ℝ : x ∈ Ω, 0 < y < w(x)}, defined by Jf(x,y) = f(x).

Theorem: J is an isometric embedding, i.e.:

• ‖f‖{L¹(Ω;w)} = ‖Jf‖{L¹(Ω_w)}
• ‖Df‖_w(Ω) = ‖D(Jf)‖(Ω_w)

Technical Innovations

1. Minimal Assumptions: Requires only lower semicontinuity of weights rather than A₁* conditions, greatly expanding the scope of applicability
2. w-Approximability: A newly introduced technical condition that characterizes the local behavior of weights on singular sets
3. Dimension-Lifting Embedding: Achieves isometric embedding into higher-dimensional unweighted spaces through subgraph construction
4. Unified Framework: Unifies weighted and unweighted BV theory within a single framework

Experimental Setup

Theoretical Verification

This paper is primarily theoretical research, with correctness verified through:

1. Counterexample Construction:
   • Example 4.5: Demonstrates the existence of sets with finite w-perimeter but infinite perimeter
   • Example 4.6: Demonstrates the existence of sets with finite perimeter but infinite w-perimeter
   • Remark 5.8: Shows the necessity of the w-approximability condition
2. Concrete Calculations:
   • w-perimeter calculations for balls verify theoretical results
   • Variation measure calculations for characteristic functions

Weight Class Analysis

Analyzes properties of different types of weights:

• Power weights w(x) = |x|^α
• Piecewise constant weights
• A₁ weight class

Experimental Results

Main Theoretical Results

1. Universality of Structure Theorem: Proves that the structure theorem holds under the weakest condition of merely requiring lower semicontinuity
2. Precision of Smooth Approximation:
   • w-approximable case: Complete convergence ‖Df_k‖_w → ‖Df‖_w
   • Non-w-approximable case: Bounded convergence with error not exceeding w_{A₁} times
3. Isometry of Embedding: Completely preserves L¹ norms and variation measures

Inequality Constants

• GNS Inequality:
  • General case: Constant is C₁w^{2/1*}_{A₁}
  • w^{1/1*}-approximable case: Improved to C₁w^{1/1*}_{A₁}

Application Verification

By substituting the characteristic function χ_E into the GNS inequality, the isoperimetric inequality is naturally obtained, verifying the consistency of the theory.

Related Work

Classical BV Theory

• Evans-Gariepy EG15: Standard theory of unweighted BV spaces
• Ambrosio-Fusco-Pallara AFP00: Modern theory of BV functions

Weighted Function Spaces

• Baldi Ba01: Earliest systematic study of weighted BV spaces
• Camfield Ca08: Coarea formula in weighted BV spaces
• Pérez-Rela PR19: GNS inequalities for weighted Sobolev spaces

Advances in This Paper

Compared to existing work, this paper:

1. Relaxes weight conditions (from A₁* to lower semicontinuity)
2. Introduces new technical tools (w-approximability)
3. Establishes direct connections with unweighted theory (isometric embedding)
4. Provides complete inequality theory

Conclusions and Discussion

Main Conclusions

1. Weighted BV theory can be established under very general weight conditions
2. w-approximability is the key condition for obtaining optimal smooth approximation
3. Weighted BV spaces possess deep geometric connections with unweighted BV spaces
4. Weighted GNS and isoperimetric inequalities provide important tools for applications

Limitations

1. w-Approximability Condition: Although natural, it is not necessary and may be overly restrictive
2. A₁ Weight Restriction: GNS inequalities still require A₁ conditions, failing to completely remove weight restrictions
3. Dimension Increase: The embedding theorem requires lifting to one higher dimension, which may limit applications

Future Directions

1. Seek weaker alternative conditions for w-approximability
2. Study GNS inequalities under more general weights
3. Explore applications in PDEs and variational problems
4. Develop geometric measure theory for weighted BV spaces

In-Depth Evaluation

Strengths

1. Theoretical Depth: Establishes a complete theoretical framework under minimal assumptions, demonstrating profound mathematical expertise
2. Technical Innovation: Both the w-approximability condition and isometric embedding construction are clever technical innovations
3. Systematicity: Forms a complete system from foundational theory to applied inequalities
4. Rigor: Proofs are detailed and rigorous, counterexamples are appropriately constructed, fully demonstrating the boundaries of the theory

Weaknesses

1. Insufficient Application Orientation: As purely theoretical research, it lacks concrete application examples
2. Computational Complexity: Some conditions (such as w-approximability) may be difficult to verify in practice
3. Weight Restrictions: Despite relaxing conditions, GNS inequalities still require A₁ weights

Impact

1. Theoretical Contribution: Provides important supplementation and refinement to weighted function space theory
2. Methodological Value: The introduced technical methods have reference value for related research
3. Application Potential: Provides theoretical foundations for application fields such as image processing and variational problems

Applicable Scenarios

1. Theoretical Research: Function space theory, geometric measure theory research
2. Variational Problems: Optimization problems involving weighted energy functionals
3. Image Processing: Theoretical analysis of weighted total variation models
4. PDE Theory: Study of weighted elliptic and parabolic equations

References

This paper primarily references the following important literature:

• Ba01 A. Baldi. Weighted BV Functions. Houston Journal of Mathematics 27 (2001)
• EG15 L. C. Evans and R. F. Gariepy. Measure Theory and Fine Properties of Functions
• AFP00 L. Ambrosio, N. Fusco, and D. Pallara. Functions of Bounded Variation
• PR19 C. Pérez and E. Rela. Degenerate Poincaré-Sobolev inequalities
• Ca08 C. S. Camfield. Comparison of BV Norms in Weighted Euclidean Spaces

Overall Assessment: This is a high-quality theoretical mathematics paper that makes important contributions to weighted bounded variation function theory. The paper is technically rigorous and highly innovative, laying a solid foundation for further development in this field. Although it emphasizes theory over direct applications, its theoretical value and potential impact are noteworthy.