Stein-Weiss inequality revisit on Heisenberg group

Basic Information

• Paper ID: 2511.00845
• Title: Stein-Weiss inequality revisit on Heisenberg group
• Authors: Chuhan Sun (Zhejiang University), Zipeng Wang (Westlake University)
• Classification: math.CA (Classical Analysis and ODEs)
• Submission Date: November 2, 2025 (arXiv submission)
• Paper Link: https://arxiv.org/abs/2511.00845

Abstract

This paper investigates a class of fractional integral operators defined on the Heisenberg group whose kernels satisfy the Zygmund dilation property. The authors establish a complete characterization between weighted norm inequalities and necessary constraints, with weight functions taking appropriate power function forms. As the main result, necessary and sufficient conditions for the Stein-Weiss inequality on the Heisenberg group are obtained.

Research Background and Motivation

Problem Background

1. Classical Theory Review:
   • Hardy-Littlewood-Sobolev Theorem (1928-1938): Established $L^p \to L^q$ boundedness of fractional integral operators on Euclidean spaces
   • Stein-Weiss Theorem (1958): Provided weighted versions of fractional integral inequalities with weight functions of the form $|x|^{-\gamma}$ and $|x|^{\delta}$
   • Folland-Stein Theorem (1974): Extended the Hardy-Littlewood-Sobolev theorem to the Heisenberg group
2. Limitations of Existing Work:
   • Han-Lu-Zhu (2012) studied the Stein-Weiss inequality on the Heisenberg group but only provided sufficient conditions
   • Existing results lack necessity proofs; the optimality of parameter constraints has not been fully characterized
   • Theoretical research on kernels with multi-parameter structure (satisfying Zygmund dilation) is relatively limited
3. Research Significance:
   • The Heisenberg group is a fundamental non-commutative structure in harmonic analysis with important applications in PDEs and geometric analysis
   • Weighted inequalities play a crucial role in studying singular integral operators and fractional differential equations
   • Multi-parameter structure reflects more refined geometric properties

Research Motivation

This paper aims to:

1. Perfect the theory of Stein-Weiss inequalities on the Heisenberg group by providing necessary and sufficient conditions
2. Study more general kernels (satisfying Zygmund dilation) to extend the existing framework
3. Develop new technical methods for handling the double-weight case ($\gamma + \delta > 0$)

Core Contributions

1. Complete Characterization (Theorem One): Proves that the sufficient conditions in the Han-Lu-Zhu theorem are actually also necessary, establishing necessary and sufficient conditions for the classical Stein-Weiss inequality on the Heisenberg group
2. Generalization to Multi-Parameter Case (Theorem Two): Studies fractional integral operators with kernels $V_{\alpha\beta\vartheta}(u,v,t) = |u|^{\alpha-n}|v|^{\alpha-n}|t|^{\beta-1}\left[\frac{|u||v|}{|t|} + \frac{|t|}{|u||v|}\right]^{-\vartheta}$ and provides necessary and sufficient conditions for $L^p \to L^q$ boundedness
3. Innovative Technical Methods:
   • Develops cone decomposition techniques to handle the double-weight case on $\mathbb{R}^n \times \mathbb{R}^n$
   • Reduces the problem to infinitely many single-parameter operators defined on eccentric cones
   • Proves that operator norms decay exponentially with the cone's eccentricity
4. New Direct Proof: Provides a more direct proof of the sufficiency part compared to the original literature

Detailed Methodology

Task Definition

Research Object: Fractional integral operators defined on the Heisenberg group $\mathbb{R}^{2n+1}$$I_{\alpha\beta\vartheta}f(u,v,t) = \iiint_{\mathbb{R}^{2n+1}} f(\xi,\eta,\tau)V_{\alpha\beta\vartheta}[(u,v,t)\odot(\xi,\eta,\tau)^{-1}]d\xi d\eta d\tau$

where:

• $\odot$ is the Heisenberg group multiplication: $(u,v,t)\odot(\xi,\eta,\tau) = [u+\xi, v+\eta, t+\tau+\mu(u\cdot\eta - v\cdot\xi)]$
• The kernel satisfies the Zygmund dilation property

Weight Functions: $\omega(u,v) = \sqrt{|u|^2+|v|^2}^{-\gamma}$, $\sigma(u,v) = \sqrt{|u|^2+|v|^2}^{\delta}$

Objective: Characterize the necessary and sufficient conditions for the double-weight norm inequality $\|\omega I_{\alpha\beta\vartheta}f\|_{L^q} \leq C \|f\sigma\|_{L^p}$ to hold.

Core Technical Framework

1. Necessity Proof (Section 3)

Strategy: Derive parameter constraints through constructing special test functions and dilation transformations.

Key Steps:

(a) Basic Necessary Conditions:

• From self-adjointness, obtain local integrability requirements for weight functions: $\gamma < \frac{2n}{q}$, $\delta < 2n(\frac{p-1}{p})$
• Through variable substitution and volume estimation, obtain the homogeneity condition: $\frac{\alpha+\beta}{n+1} = \frac{1}{p} - \frac{1}{q} + \frac{\gamma+\delta}{2n+2}$

(b) Dilation Parameter Constraints:

• Consider the dilation $(u,v,t) \to (ru,rv,r^{2\lambda}t)$; by taking $\lambda \to 0^+$ or $\lambda \to +\infty$, obtain: $\vartheta \geq \left|\frac{\alpha-n\beta}{n+1} - \frac{\gamma+\delta}{2n+2}\right|$

(c) Non-negativity Condition:

• Construct rectangular test function $f = \sigma^{-\frac{p}{p-1}}\chi_{Q_1\times Q_2 \times I}$
• Through the Lebesgue differentiation theorem, analyze the limiting behavior of the integral as volume contracts
• Obtain $\gamma + \delta \geq 0$

(d) Refined Constraints (Most technically involved): For the case $\gamma \geq 0, \delta \leq 0$, define $A_{\zeta\gamma\delta}^{pq}(Q_1\times Q_2) = \text{vol}\{Q_1\}^{\frac{\zeta}{n}-(\frac{1}{p}-\frac{1}{q})}\text{vol}\{Q_2\}^{\frac{\zeta}{n}-(\frac{1}{p}-\frac{1}{q})} \times \text{(weighted averages)}$

where $\zeta = n[\frac{\alpha+\beta}{n+1}] + \frac{\gamma+\delta}{2n+2}$.

• When $\gamma+\delta=0$, by contracting $Q_1$ obtain $\zeta - \frac{n}{p} < \delta$
• When $\gamma+\delta>0$, utilize cone decomposition $Q_2 = \bigcup_{k\geq 0} Q_2^k$ (dyadic annuli), prove $\sum_{k\geq 0} A_k(\lambda) \text{ continuous at } \lambda=0$ Combined with Hölder continuity and concrete estimates, derive the same constraint

Similar analysis for $\gamma \leq 0, \delta \geq 0$ yields dual constraints.

2. Sufficiency Proof Framework

Core Idea: Decompose the operator into two steps

1. First perform fractional integration in the $t$ direction (using the classical Hardy-Littlewood-Sobolev theorem)
2. Then handle the $(u,v)$ direction, which is the key difficulty

Operator Reconstruction (Section 4):

Using $\vartheta \geq |\frac{\alpha-n\beta}{n+1} - \frac{\gamma+\delta}{2n+2}|$, prove $V_{\alpha\beta\vartheta}(u,v,t) \lesssim |u|^{\zeta-n}|v|^{\zeta-n}|t|^{\frac{\alpha+\beta}{n+1}-\frac{\gamma+\delta}{2n+2}-1}$

Therefore $I_{\alpha\beta\vartheta}f(u,v,t) \lesssim \iint_{\mathbb{R}^{2n}} |u-\xi|^{\zeta-n}|v-\eta|^{\zeta-n}F_{\alpha\beta\gamma\delta}(\xi,\eta,u,v,t)d\xi d\eta$

where $F_{\alpha\beta\gamma\delta}(\xi,\eta,u,v,t) = \int_{\mathbb{R}} f(\xi,\eta,\tau-\mu(u\cdot\eta-v\cdot\xi))|t-\tau|^{[\frac{\alpha+\beta}{n+1}-\frac{\gamma+\delta}{2n+2}]-1}d\tau$

The problem reduces to studying the operator $II_{\zeta}g(u,v) = \iint_{\mathbb{R}^{2n}} g(\xi,\eta)|u-\xi|^{\zeta-n}|v-\eta|^{\zeta-n}d\xi d\eta$

3. Single-Weight Case ($\gamma+\delta=0$, Proposition One)

Method: Iteratively apply the Muckenhoupt-Wheeden theorem

• First apply single-parameter weighted inequality in the $u$ direction
• Then apply in the $v$ direction, using Minkowski's integral inequality to exchange integration order
• The key is verifying that weight functions satisfy the $A_p$ condition (through homogeneity and local integrability)

4. Double-Weight Case ($\gamma+\delta>0$, Proposition Two, Section 5)

This is the most innovative part.

Cone Decomposition Technique:

For each $j \in \mathbb{Z}$, define eccentric cones $\Lambda_j(u,v) = \{(\xi,\eta): 2^{-j} \leq \frac{|u-\xi|}{|v-\eta|} < 2^{-j+1}\}$

Decompose the operator as $II_{\zeta}g = \sum_{j\in\mathbb{Z}} \Delta_j II_{\zeta}g$

where $\Delta_j II_{\zeta}g$ integrates only over $\Lambda_j(u,v)$.

Principal Lemma:

Define the weight function condition number $A_{\zeta pqr}^{(j:\omega,\sigma)} = \sup_{Q_1\times Q_2: \text{vol}\{Q_1\}^{1/n}/\text{vol}\{Q_2\}^{1/n}=2^{-j}} (\text{weighted condition})$

Prove that there exist $\epsilon>0$ and $r>1$ such that $A_{\zeta pqr}^{(j:\omega,\sigma)} \leq C 2^{-\epsilon|j|}$

Proof Strategy:

• Fix $\text{vol}\{Q_2\}^{1/n}=1$, $\text{vol}\{Q_1\}^{1/n}=\lambda=2^{-j}$
• Divide into three cases: (i) $\gamma\geq 0,\delta\leq 0$; (ii) $\gamma\leq 0,\delta\geq 0$; (iii) $\gamma>0,\delta>0$
• For each case, further subdivide based on the relationship between $\gamma qr$, $\delta\frac{p}{p-1}r$, and $n$
• Using refined constraints from necessity (such as $\zeta - \frac{n}{p} < \delta$), prove $A_{\zeta pqr}^{(j:\omega,\sigma)} \lesssim \lambda^{\epsilon}$ where $\epsilon$ depends on parameter relationships (e.g., $\epsilon = \frac{\gamma+\delta}{2}$ or $\epsilon = \frac{n}{qr}-\frac{\gamma}{2}+\frac{\delta}{2}$, etc.)

Operator Norm Estimation:

Using the Sawyer-Wheeden theorem (sufficient conditions for weighted fractional integrals) and cone decomposition, obtain $\|\omega \Delta_j II_{\zeta}g\|_{L^q} \lesssim 2^{-j[\zeta+\frac{n}{q}]}2^{j[\zeta-\frac{n}{p}+\frac{n}{q}]}A_{\zeta pqr}^{(j:\omega,\sigma)}\|g\sigma\|_{L^p} = A_{\zeta pqr}^{(j:\omega,\sigma)}\|g\sigma\|_{L^p}$

Since $\sum_{j\in\mathbb{Z}} A_{\zeta pqr}^{(j:\omega,\sigma)} < \infty$ (exponential decay), Minkowski's inequality yields the desired result.

Technical Innovation Points

1. Cone Decomposition Method: Systematically decomposes $\mathbb{R}^n\times\mathbb{R}^n$ into infinitely many cones with different eccentricities; the operator on each cone is essentially single-parameter, but requires precise control of norm decay with eccentricity
2. Complete Derivation of Necessary Conditions: Through clever construction of test functions and dyadic decomposition, converts geometric information (volume contraction behavior) into parameter constraints
3. Unified Treatment Framework: Standardizes fractional integral operators with Zygmund dilation-type kernels, revealing the essence of multi-parameter structure
4. Exponential Decay Estimates of Weight Function Condition Numbers: This is key to handling the double-weight case, reflecting the deep connection between geometric structure and analytical properties

Experimental Setup

Note: This is a pure theoretical mathematics paper with no numerical experiments or datasets. All results are rigorous mathematical proofs.

Theoretical Verification Methods

1. Consistency Checks in Special Cases:
   • When $\vartheta = \frac{n+1}{2}-\frac{\rho}{2}$, $\alpha=\frac{\rho}{2}+\frac{n-1}{2}$, $\beta=\frac{\rho}{2}-\frac{n-1}{2}$, Theorem Two reduces to Theorem One
   • When $\gamma+\delta=0$, the double-weight case reduces to the single-weight case
2. Correspondence with Classical Results:
   • When $n=0$ (degenerate to $\mathbb{R}$), results should be consistent with the Stein-Weiss theorem on Euclidean spaces
   • The form of parameter constraints maintains structural similarity with the Folland-Stein theorem

Logical Structure of Proofs

• Section 2: Complete proof of Theorem One (necessity + sufficiency)
• Section 3: Necessity of Theorem Two
• Section 4: Operator reconstruction and single-weight case
• Section 5: Cone decomposition and double-weight case

Each section contains detailed inequality chains and limit analysis.

Experimental Results

Main Theorem Statements

Theorem One (Complete Stein-Weiss Inequality):

Let $0<\rho<n+1$, $\omega(u,v)=\sqrt{|u|^2+|v|^2}^{-\gamma}$, $\sigma(u,v)=\sqrt{|u|^2+|v|^2}^{\delta}$. Then $\|\omega S_{\rho}f\|_{L^q(\mathbb{R}^{2n+1})} \leq C\|f\sigma\|_{L^p(\mathbb{R}^{2n+1})}, \quad 1<p<q<\infty$

if and only if: $\gamma < \frac{2n}{q}, \quad \delta < 2n\left(\frac{p-1}{p}\right), \quad \gamma+\delta \geq 0, \quad \frac{\rho}{n+1} = \frac{1}{p}-\frac{1}{q}+\frac{\gamma+\delta}{2n+2}$

Theorem Two (Multi-Parameter Generalization):

Let $\alpha,\beta\in\mathbb{R}$, $\vartheta\geq 0$, with weight functions as above. Then $\|\omega I_{\alpha\beta\vartheta}f\|_{L^q(\mathbb{R}^{2n+1})} \leq C\|f\sigma\|_{L^p(\mathbb{R}^{2n+1})}, \quad 1<p<q<\infty$

if and only if: