Riesz transforms on compact Lie groups, spheres and Gauss space (Q1416634)

From MaRDI portal
scientific article
Language Label Description Also known as
English
Riesz transforms on compact Lie groups, spheres and Gauss space
scientific article

    Statements

    Riesz transforms on compact Lie groups, spheres and Gauss space (English)
    0 references
    0 references
    15 December 2003
    0 references
    Let \(M\) be a Riemannian manifold without boundary, \(\nabla_M\), \(\text{div}_M\) and \(\Delta_M\) be, respectively, the gradient, the divergence and the Laplacian associated with \(M\). The operator \(R^M=\nabla_M\circ(-\Delta_M)^{-1/2}\) is called the Riesz transform on \(M\). Let \(M=G\) be a compact Lie group endowed with a biinvariant Riemannian metric and let \(\mathfrak G\) be its Lie algebra. Let \(X\) be a left invariant vector field such that \(|X|=1\), where \(|\cdot|\) is the norm induced on \(\mathfrak G\) by the metric of \(G\). The operator \(R^X=X\circ(-\Delta_G)^{-1/2}\) is called the Riesz transform in the direction \(X\). The author gives the following: Let \(1<p<\infty\). Then, it holds \[ \|R^G f\|_p\leq 2(\max(p,p')-1)\|f\|_p\tag{i} \] for any \(f\in L^p(G)\) with \(\int_G f(x) dx=0\); \[ \|R^X f\|_p\leq \cot(\pi/(2\max(p,p')))\|f\|_p\tag{ii} \] for any \(f\in L^p(G)\) with \(\int_G f(x) dx=0\); \[ \|(Id\oplus R^X) f\|_p\leq (\cot(\pi/(2\max(p,p')))^2+1)^{1/2} \|f\|_p\tag{iii} \] for any \(f\in L^p(G)\) with \(\int_G f(x) dx=0\). The above bounds are the best possible, if \(G\) is the \(n\)-dimensional torus, or if \(G=SO(n)\). The author also considers the cases \(M=S^n(\mathbb R^n)\) with the standard metric and \(\mathbb R^n\) endowed with the Gaussian measure \(\gamma(dx)=(2\pi)^{-n/2}e^{-|x|^2/2}dx\). The proofs are based on a probabilistic method, using martingale theory, begun by D. Burkholder, P. A. Meyer, R. F Gundy and N. Varopoulos, etc.
    0 references
    Riemannian manifold
    0 references
    Riesz transform
    0 references
    compact Lie group
    0 references
    Gaussian measure
    0 references
    0 references

    Identifiers

    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references