Uniqueness of stable allocations in dimensions one and two
Let $\Xi$ be a locally finite set of centres in $\mathbb{R}^d$ and let $\alpha\in(0,\infty)$. Hoffman, Holroyd and Peres proved that Lebesgue measure on $\mathbb{R}^d$ has a stable allocation to $\Xi$ with appetite $\alpha$, that it is unique up to null sets when the graphs joining centres at bounded distance are recurrent, and that it is almost surely unique when $\Xi$ comes from a translation-invariant point process. They asked whether some discrete set of centres admits two stable allocations that differ on a set of positive measure. We show that no such set exists in dimensions one and two: for every locally finite $\Xi\subset\mathbb{R}^d$ with $d\le 2$ and every $\alpha\in(0,\infty)$ the stable allocation is unique up to null sets, with no condition on the growth or the local density of $\Xi$. In every dimension the stable allocation is unique when the number of centres within distance $r$ of the origin grows at most slightly faster than $r^2$, for example as $r^2\log r\log\log r$.
Authors
- Yitzchak Shmalo (ORCID: https://orcid.org/0000-0002-2886-3049)
Institutions
- Hebrew University of Jerusalem (IL)
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-11-13
- DOI
- https://doi.org/10.5281/zenodo.24199456
- Primary Topic
- Stochastic processes and statistical mechanics
- Type
- preprint