Skip to main navigation Skip to search Skip to main content

Optimal Functional Inequalities for Fractional Operators on the Sphere and Applications

  • Jean Dolbeault
  • , An Zhang*
  • *Corresponding author for this work
  • Paris-Dauphine University

Research output: Contribution to journalArticlepeer-review

Abstract

This paper is devoted to the family of optimal functional inequalities on the n-dimensional sphere Sn[EQUATION PRESENTED] where Ls denotes a fractional Laplace operator of order s (0, n), q [1, 2) ∪ (2, q+], q+ = 2n n.s is a critical exponent, and dμ is the uniform probability measure on Sn. These inequalities are established with optimal constants using spectral properties of fractional operators. Their consequences for fractional heat flows are considered. If q > 2, these inequalities interpolate between fractional Sobolev and subcritical fractional logarithmic Sobolev inequalities, which correspond to the limit case as q → 2. For q < 2, the inequalities interpolate between fractional logarithmic Sobolev and fractional Poincare inequalities. In the subcritical range q < q+, the method also provides us with remainder terms which can be considered as an improved version of the optimal inequalities. The case s o (.n, 0) is also considered. Finally, weighted inequalities involving the fractional Laplacian are obtained in the Euclidean space, by using the stereographic projection.

Original languageEnglish
Pages (from-to)863-880
Number of pages18
JournalAdvanced Nonlinear Studies
Volume16
Issue number4
DOIs
StatePublished - 1 Nov 2016
Externally publishedYes

Keywords

  • Fractional Heat Flow
  • Fractional Logarithmic Sobolev Inequality
  • Fractional Poincaré Inequality
  • Fractional Sobolev Inequality
  • Hardy-Littlewood-Sobolev Inequality
  • Spectral Gap
  • Stereographic Projection
  • Subcritical Interpolation Inequalities on the Sphere

Fingerprint

Dive into the research topics of 'Optimal Functional Inequalities for Fractional Operators on the Sphere and Applications'. Together they form a unique fingerprint.

Cite this