Weighted $H^p-L^q$ boundedness of integral operators with rough kernels

Abstract: In this paper, we study integral operators \begin{equation*} T_αf(x)=\int_{\mathbb{R}^{n}}K(x,y) f(y)dy, \end{equation*} with kernels $K(x,y)= k_1( x- A_1y)...k_m( x-A_my),$ where $k_i(x)=\frac{Ω<em>i(x)}{|x|<sup>{n/q_i}}$ and $Ω_i: \mathbb{R}^n\to \mathbb{R}$ are homogeneous functions of degree zero, satisfying a size and a Dini condition, $A</em>{i}$ are certain invertible matrices, and $\frac n{q_1}+\dots+\frac n{q_m}=n-α,$ $0\leq α&lt;n.$ We obtain the $$H<sup>{p}<em>{w^{p}(\mathbb{R}^{{n})-L^{q}}}</em>{w<sup>q}(\mathbb{R}<sup>{n})$$ boundedness of these operators, for a class of Muckenhoupt weights $w$, satisfying the condition \begin{equation*} w(A_ix)\leq cw(x), \end{equation*} $a.e.x\in\mathbb{R} ⁿ$, $1\leq i\leq m$.

• The paper establishes the boundedness of Tα from weighted Hardy spaces to weighted Lebesgue spaces under rough kernel conditions.
• It leverages atomic decomposition and molecular theory to precisely control singular integral operators with non-smooth kernels.
• The approach extends classical results to a multilinear and matrix-weighted setting with explicit exponent thresholds and quantitative bounds.

Weighted $H^p-L^q$ Boundedness for Integral Operators with Rough Kernels

Introduction and Context

The paper addresses the $H^p_w(\mathbb{R}^n) \to L^q_w(\mathbb{R}^n)$ boundedness properties of a class of multilinear integral operators with rough homogeneous kernels under general weighted regimes. Specifically, it focuses on operators of the form

$T_\alpha f(x) = \int_{\mathbb{R}^n} K(x,y) f(y) dy$

where the kernel is given by

$K(x,y) = k_1(x-A_1 y) \cdots k_m(x-A_m y)$

with $k_i(x) = \Omega_i(x)/|x|^{n/q_i}$, each $\Omega_i$ a homogeneous function of degree zero on $\mathbb{R}^n$, and invertible matrices $A_i$. The parameters satisfy

$$\frac{n}{q_1} + \dots + \frac{n}{q_m} = n-\alpha, \quad 0\le \alpha < n.$$

The work extends the classical one-parameter theory of rough singular/fractional integral operators to a multilinear, matrix-weighted, and weighted Hardy space setting.

Prior to this work, weighted bounds for fractional and singular integral operators with homogeneous or rough kernels on Hardy spaces and Lebesgue spaces were established under various smoothness or cancellation assumptions on the kernel and strong conditions on the weight function. Results concerning $H^p-L^q$ boundedness with weights were largely confined to $H^p_w(\mathbb{R}^n) \to L^q_w(\mathbb{R}^n)$0 and smooth/Calderón-Zygmund-type kernels.

Main Results

The central theorem provides $H^p_w(\mathbb{R}^n) \to L^q_w(\mathbb{R}^n)$1 boundedness for the class of integral operators $H^p_w(\mathbb{R}^n) \to L^q_w(\mathbb{R}^n)$2 described above under rough kernel assumptions and with weights $H^p_w(\mathbb{R}^n) \to L^q_w(\mathbb{R}^n)$3 in the Muckenhoupt $H^p_w(\mathbb{R}^n) \to L^q_w(\mathbb{R}^n)$4 class twisted by the matrices $H^p_w(\mathbb{R}^n) \to L^q_w(\mathbb{R}^n)$5. Explicitly, if $H^p_w(\mathbb{R}^n) \to L^q_w(\mathbb{R}^n)$6 satisfies

$H^p_w(\mathbb{R}^n) \to L^q_w(\mathbb{R}^n)$7

and $H^p_w(\mathbb{R}^n) \to L^q_w(\mathbb{R}^n)$8, with detailed conditions on exponents determined by the structure of the operator and kernel integrability, then

$H^p_w(\mathbb{R}^n) \to L^q_w(\mathbb{R}^n)$9

is bounded. The result includes kernels that only satisfy a Dini-type continuity condition and size bounds but may lack any smoothness.

This extends $T_\alpha f(x) = \int_{\mathbb{R}^n} K(x,y) f(y) dy$0-$T_\alpha f(x) = \int_{\mathbb{R}^n} K(x,y) f(y) dy$1 results in the literature to the weighted Hardy-space regime for multilinear operators with singular non-smooth kernels and nontrivial dilation geometry.

Technical Approach

The analysis uses the atomic decomposition and molecular theory of weighted Hardy spaces, as developed by Fefferman-Stein, Coifman, Latter, and their generalization to weights in the $T_\alpha f(x) = \int_{\mathbb{R}^n} K(x,y) f(y) dy$2 class. A core step is to prove that, under the specified conditions on the kernel and matrices, $T_\alpha f(x) = \int_{\mathbb{R}^n} K(x,y) f(y) dy$3 is uniformly bounded on weighted atoms.

A sharp extension theorem for linear operators acting on Hardy spaces is established, whereby $T_\alpha f(x) = \int_{\mathbb{R}^n} K(x,y) f(y) dy$4 is bounded $T_\alpha f(x) = \int_{\mathbb{R}^n} K(x,y) f(y) dy$5 if and only if it is uniformly bounded on $T_\alpha f(x) = \int_{\mathbb{R}^n} K(x,y) f(y) dy$6-atoms adapted to $T_\alpha f(x) = \int_{\mathbb{R}^n} K(x,y) f(y) dy$7, for admissible exponents. The theorem is optimized by using precise properties of $T_\alpha f(x) = \int_{\mathbb{R}^n} K(x,y) f(y) dy$8 weights under matrix action and careful handling of the scaling behavior of the kernel components. The Dini-type and $T_\alpha f(x) = \int_{\mathbb{R}^n} K(x,y) f(y) dy$9 conditions on the $K(x,y) = k_1(x-A_1 y) \cdots k_m(x-A_m y)$0 functions ensure generalized Hörmander continuity enabling boundedness even in the rough case.

For the off-support part of the atom, vanishing moment structure and growth/oscillation conditions on the kernel allow cancellation to occur, giving suitable decay and letting weighted norm estimates close via careful integration and maximal function control.

The proofs make significant use of multilinear/core Young and reverse Hölder inequalities, and a nuanced analysis of the interaction of the kernel decay and weight geometry via the $K(x,y) = k_1(x-A_1 y) \cdots k_m(x-A_m y)$1 conditions, particularly in contexts where the matrices $K(x,y) = k_1(x-A_1 y) \cdots k_m(x-A_m y)$2 may be orthogonal or other structured invertible maps.

Numerical Significance and Implications

The paper provides explicit exponent thresholds and quantitative criteria for boundedness in terms of the parameters of the kernel and the weights. The uniform atom estimate establishes boundedness constants that are independent of atom support, thus ensuring the optimality of the atomic decomposition. The proof strategy, based around extension via atomic control, allows further generalization to other related settings such as Hardy spaces associated to different function systems or geometry.

A key consequence is that for multilinear operators with highly non-smooth, non-scalar dilated kernels, boundedness can be achieved in the Hardy regime for a wide class of $K(x,y) = k_1(x-A_1 y) \cdots k_m(x-A_m y)$3-type weights—as long as these weights are stable under pullback by matrices $K(x,y) = k_1(x-A_1 y) \cdots k_m(x-A_m y)$4. This is not simply an extension of prior $K(x,y) = k_1(x-A_1 y) \cdots k_m(x-A_m y)$5 or $K(x,y) = k_1(x-A_1 y) \cdots k_m(x-A_m y)$6 boundedness: the multilinear/matrix structure introduces fundamentally new difficulties in controlling the weighted norm and kernel oscillation simultaneously that are nontrivial for $K(x,y) = k_1(x-A_1 y) \cdots k_m(x-A_m y)$7.

Theoretical and Practical Outlook

The results have strong implications for the general theory of harmonic analysis in the weighted setting, particularly relevant to multi-parameter and non-commutative integral operators. The flexibility with respect to roughness and non-standard weights opens avenues for the study of more general pseudo-differential operators and singular integrals with anisotropic, non-scalar geometries. These estimates can also form the foundation for non-Euclidean analysis and the regularity theory of PDEs in highly heterogeneous (weight-dependent) media.

Future extensions may aim at vector-valued extensions, multilinear commutators with BMO or other function spaces, or at sharp weighted bounds with endpoint extrapolation, as well as the connection with boundedness in metric space Hardy spaces or on groups with dilation structure.

Conclusion

This work establishes a general theory for the boundedness of multilinear (fractional-type) integral operators with rough homogeneous kernels between weighted Hardy and Lebesgue spaces, with comprehensive control via matrix-adapted Muckenhoupt weights. By systematically using atomic/molecular decompositions and detailed analysis of kernel roughness and weight structure, it extends and refines a sequence of important results in modern harmonic analysis, setting a flexible technical framework for further investigations in the field.