Dunkl–Riesz Potential in Harmonic Analysis

• Dunkl–Riesz potential is a fractional integral operator that generalizes the classical Riesz potential by incorporating reflection group structures via Dunkl operators.
• It underpins sharp inequalities and embeddings such as the Stein–Weiss and Hardy–Littlewood–Sobolev types using weighted norms and the effective dimension d_k.
• Its diverse representations through convolution, semigroup methods, and wavelet constructions enable precise norm estimates and inversion techniques in harmonic analysis.

The Dunkl–Riesz potential is a fractional integral operator that generalizes the classical Riesz potential by incorporating the structure of finite reflection groups through Dunkl operators and their associated invariant weights and measures. This operator extends traditional harmonic analysis into the context of Dunkl theory, providing a fundamental tool for the study of sharp functional inequalities, weighted norm estimates, and embedding theorems in spaces invariant under reflection symmetry.

1. Fundamental Definitions and Algebraic Structure

Let $R \subset \mathbb R^d\backslash\{0\}$ be a (possibly reduced) root system, $G$ its accompanying reflection group, and $k: R\to [0,\infty)$ a $G$-invariant multiplicity function. The Dunkl operator in the $i$th coordinate is defined by

$$T_i f(x) = \frac{\partial f}{\partial x_i}(x) + \sum_{v\in R^+} k(v)\, v_i \, \frac{f(x)-f(\sigma_v x)}{\langle v, x\rangle}$$

where $R^+$ is a positive subsystem and $\sigma_v$ denotes reflection in the hyperplane orthogonal to $v$.

The associated Dunkl weight is

$$w_k(x) = \prod_{v\in R^+} |\langle v, x\rangle|^{2k(v)}$$

which is homogeneous of degree $G$0, with $G$1. The measure is $G$2, and the effective dimension is $G$3.

The Dunkl kernel $G$4 satisfies

$G$5

and generalizes the exponential in the case $G$6. The Dunkl transform is given by

$G$7

where $G$8 normalizes the transform to an $G$9-isometry.

For $k: R\to [0,\infty)$0, the Dunkl–Riesz potential is defined by

$k: R\to [0,\infty)$1

or equivalently through the Dunkl transform as

$k: R\to [0,\infty)$2

Similar representations are available in the one-dimensional setting, where the operator encodes the full structure of the underlying reflection group $k: R\to [0,\infty)$3 (Dutta et al., 10 Apr 2026, Nowak et al., 2014).

2. Sharp Inequalities and Functional Analysis

Stein–Weiss Inequality (Weighted Hardy–Littlewood–Sobolev)

For parameters $k: R\to [0,\infty)$4, $k: R\to [0,\infty)$5, $k: R\to [0,\infty)$6, $k: R\to [0,\infty)$7, and the critical homogeneity relation

$k: R\to [0,\infty)$8

there exists an optimal constant $k: R\to [0,\infty)$9 such that

$G$0

with extremals existing at critical values (Adhikari et al., 2019, Gorbachev et al., 2017). In the one-dimensional case, necessary and sufficient conditions for the sharp two-weighted $G$1 boundedness are fully characterized (Nowak et al., 2014).

Hardy–Littlewood–Sobolev and Sobolev-Type Embeddings

For $G$2, $G$3, and

$G$4

the Dunkl–Riesz potential maps $G$5 boundedly into $G$6, matching the classical exponents but with $G$7 as the effective dimension (Gorbachev et al., 2017, Abdelkefi et al., 2013). The unweighted Dunkl–Sobolev inequality for $G$8 holds: $G$9 where $i$0.

Both inequalities admit extremal functions, which may be chosen as $i$1-invariant, radial, and positive solutions to an associated Euler–Lagrange equation: $i$2 (Adhikari et al., 2019).

Maximal Functions and Morrey Spaces

The Dunkl fractional maximal function

$i$3

satisfies parallel norm inequalities to $i$4. On Dunkl–Morrey spaces $i$5, Adams-type inequalities of Stein–Weiss-type are valid, with explicit parameter ranges calculated for one-dimensional settings (Dutta et al., 10 Apr 2026).

3. Kernel Representations, Regularity, and Inversion

Integral representations of $i$6 are available:

• Through convolution with a homogeneous radial kernel via the Dunkl translation (Gorbachev et al., 2017, Abdelkefi et al., 2013)
• Through the Dunkl heat semigroup:

$i$7

• By wavelet-based constructions, which yield explicit inverses as limits of truncated hypersingular integrals, with quantitative convergence rates depending on local $i$8-smoothness of the function (Verma et al., 21 May 2025).

In the type A theory, the role of the Riesz kernel is played by distributions $i$9 supported on the positive Weyl chamber, forming a one-parameter convolution group under the Dunkl convolution: $$T_i f(x) = \frac{\partial f}{\partial x_i}(x) + \sum_{v\in R^+} k(v)\, v_i \, \frac{f(x)-f(\sigma_v x)}{\langle v, x\rangle}$$0 and

$$T_i f(x) = \frac{\partial f}{\partial x_i}(x) + \sum_{v\in R^+} k(v)\, v_i \, \frac{f(x)-f(\sigma_v x)}{\langle v, x\rangle}$$1

(Brennecken, 2023, Rösler, 2019).

4. Weighted and Two-Weight Norm Inequalities

Weighted norm inequalities for the Dunkl–Riesz potential are established using rearrangement-invariant Hardy and Sawyer-type criteria. For weights $$T_i f(x) = \frac{\partial f}{\partial x_i}(x) + \sum_{v\in R^+} k(v)\, v_i \, \frac{f(x)-f(\sigma_v x)}{\langle v, x\rangle}$$2 and exponents $$T_i f(x) = \frac{\partial f}{\partial x_i}(x) + \sum_{v\in R^+} k(v)\, v_i \, \frac{f(x)-f(\sigma_v x)}{\langle v, x\rangle}$$3,

$$T_i f(x) = \frac{\partial f}{\partial x_i}(x) + \sum_{v\in R^+} k(v)\, v_i \, \frac{f(x)-f(\sigma_v x)}{\langle v, x\rangle}$$4

if and only if companion weighted Hardy inequalities on non-increasing rearrangements hold; in particular, this encompasses both radial power weights and general weights satisfying integral conditions (Abdelkefi et al., 2013, Abdelkefi et al., 2013, Gorbachev et al., 2017).

In the one-dimensional ($$T_i f(x) = \frac{\partial f}{\partial x_i}(x) + \sum_{v\in R^+} k(v)\, v_i \, \frac{f(x)-f(\sigma_v x)}{\langle v, x\rangle}$$5) case, sharp necessary and sufficient conditions for two-power-weight $$T_i f(x) = \frac{\partial f}{\partial x_i}(x) + \sum_{v\in R^+} k(v)\, v_i \, \frac{f(x)-f(\sigma_v x)}{\langle v, x\rangle}$$6 boundedness are given explicitly in terms of balance, homogeneity, endpoint, and non-degeneracy constraints on the parameters (Nowak et al., 2014).

5. Specializations: Dunkl–Riesz in Type A, One Dimension, and Bi-Parametric Potentials

For type A root systems, the Dunkl–Riesz potential is represented in terms of weighted measures on the positive Weyl chamber and enjoys full positivity on the generalized Wallach set, with connections to the theory of symmetric cones and Jack polynomials (Rösler, 2019, Brennecken, 2023). The convolution, Laplace transform, and positivity properties are all precisely characterized in this context.

In dimension one, for $$T_i f(x) = \frac{\partial f}{\partial x_i}(x) + \sum_{v\in R^+} k(v)\, v_i \, \frac{f(x)-f(\sigma_v x)}{\langle v, x\rangle}$$7, the potential $$T_i f(x) = \frac{\partial f}{\partial x_i}(x) + \sum_{v\in R^+} k(v)\, v_i \, \frac{f(x)-f(\sigma_v x)}{\langle v, x\rangle}$$8 is equivalently characterized via Hankel–Dunkl transforms, and the limiting cases recover the classical Riesz and Poisson/Gegenbauer–Bessel potential when $$T_i f(x) = \frac{\partial f}{\partial x_i}(x) + \sum_{v\in R^+} k(v)\, v_i \, \frac{f(x)-f(\sigma_v x)}{\langle v, x\rangle}$$9 (Nowak et al., 2014).

The bi-parametric potential $R^+$0, interpolating Bessel and Flett potentials, is constructed via semigroup methods and allows detailed description of the inverse (via hypersingular integrals) and its mapping range characterized through quantitative convergence in $R^+$1 (Verma et al., 21 May 2025).

6. Proof Techniques and Concentration–Compactness

Key analytic mechanisms rely on:

• The absence of classical translation invariance (circumvented by Dunkl translation invariance)
• Dunkl-convolution and semigroup techniques
• Concentration–compactness principles adapted with Dunkl translations and dilation symmetries
• Weighted compactness via Dunkl–heat semigroup and refined Sobolev inequalities involving the supremum of the heat kernel
• Mellin transform and one-dimensional Hardy/Bellman inequalities after radialization

Novel Lemmas in the Dunkl setting address positivity, support properties, and support-reduction under convolution; Young’s inequality and convolution properties for the Dunkl transform play a fundamental role in deriving norm inequalities and sharp constants (Adhikari et al., 2019, Gorbachev et al., 2017, Gorbachev et al., 2017, Dutta et al., 10 Apr 2026).

7. Generalizations and Comparative Analysis

The Dunkl–Riesz potential unifies and extends several classical settings:

• For $R^+$2, the entire theory reduces to classical Euclidean Riesz potentials and corresponding weighted inequalities.
• For type A and suitable values of the multiplicity parameter, it recovers the radial theory on symmetric cones and the Riesz/Wallach distributions.
• The approach generalizes the classical Sobolev and Stein–Weiss inequalities, the convolution group structure for Riesz potentials, and weighted embedding theorems, embedding them into the framework of reflection-invariant Dunkl analysis (Rösler, 2019, Gorbachev et al., 2017).

The resulting machinery is central in modern harmonic analysis on spaces with reflection invariance, with applications to sharp inequalities, extremal function existence, embedding theorems, and semigroup theory. The local and global regularity as well as explicit inversion and approximation rates in $R^+$3 are now accessible via wavelet and hypersingular integral techniques (Verma et al., 21 May 2025).