Multilinear matrix weights

TL;DR

This paper fully characterizes matrix weights for bounded multilinear Calderón-Zygmund operators on matrix weighted Lebesgue spaces, introducing new theories and bounds that extend classical results to the multilinear matrix setting.

Contribution

It develops the theory of multilinear singular integrals with matrix values, establishing bounds and a new notion of directional nondegeneracy, advancing the understanding of matrix weights in multilinear analysis.

Findings

Established quantitative bounds in terms of matrix weight characteristics.

Defined a new notion of directional nondegeneracy for multilinear operators.

Recovered the sharpest known bounds in the linear case.

Abstract

In this work we fully characterize the classes of matrix weights for which multilinear Calder\'on-Zygmund operators extend to bounded operators on matrix weighted Lebesgue spaces. To this end, we develop the theory of multilinear singular integrals taking values in tensor products of finite dimensional Hilbert spaces. On the one hand, we establish quantitative bounds in terms of multilinear Muckenhoupt matrix weight characteristics and scalar Fujii-Wilson conditions of a tensor product analogue of the convex body sparse operator, of a convex-set valued tensor product analogue of the Hardy-Littlewood maximal operator, and of a multilinear analogue of the Christ-Goldberg maximal operator. These bounds recover the sharpest known bounds in the linear case. Moreover, we define a notion of directional nondegeneracy for multilinear Calder\'on-Zygmund operators, which is new even in the scalar case. The noncommutavity of matrix multiplication, the absence of duality, and the natural presence of quasinorms in the multilinear setting present several new difficulties in comparison to previous works in the scalar or in the linear case. To overcome them, we use techniques inspired from convex combinatorics and differential geometry.