Geometric characterizations of ${\sf PI}$ spaces: an overview of some modern techniques
Emanuele Caputo
TL;DR
The article surveys how doubling metric measure spaces with a $1$-Poincaré inequality ($1$-PI) can be characterized through both dimension-1 (pencil of curves, modulus, obstacle-avoidance) and codimension-1 (separating sets, relative isoperimetric inequalities) geometric$data. It presents two complementary frameworks: a pencil-of-curves/modulus approach and an energy/separating-sets approach, establishing their equivalence and then tying them together via geometric tools like the position function. A Euclidean toy-model demonstration and detailed analysis of separating-set energies in terms of the Riesz kernel and Minkowski content illustrate how these analytic and geometric perspectives yield PI, along with a suite of open questions. The discussion extends to the interplay between obstacle-avoidance and separating sets, and to potential extensions to MCP and GCBA spaces, outlining conditions and constructions that may produce new PI spaces. Overall, the work highlights deep links between variational, geometric measure-theoretic, and metric-analytic methods in the study of non-smooth spaces satisfying a $1$-Poincaré inequality.
Abstract
We survey recent results on the study of metric measure spaces satisfying a Poincaré inequality. We overview recent characterizations in terms of objects of dimension 1, such as pencil of curves, modulus estimates and obstacle-avoidance principles. Then, we turn our attention to characterizations in terms of objects of codimension 1, such as relative isoperimetric inequalities and separating sets, the last one obtained in collaboration with N. Cavallucci in [arXiv:2401.02762]. We propose a strategy to provide examples using our characterization in the toy-model of the Euclidean case. We also discuss a more geometric relation between separating sets and obstacle-avoidance principles, obtained in [IMRN, Vol. 2025, Issue 1, Jan. 2025, rnae276]. Finally, we recall some open questions in the field.
