Indexed metadata

Packing and covering random trees

Caelan Atamanchuk, Simon Briend, Serte Donderwinkel, Elaine Herrera

Source record

Source: arXiv

Published: Oct 29, 2026

arXiv: 2610.40150

Open original source ↗

Source abstract

We study the sizes of optimal packings and coverings of conditioned branching processes by balls with growing radii. We find a universal limiting constant, and thereby establish universality of packing and covering numbers across different scales and different offspring distributions. To be precise, for μμ an offspring distribution with mean 11 and finite variance σ2σ^2 we let TnT_n be a Bienaymé tree with offspring distribution μμ conditioned to have size nn. We prove that the packing number (resp. covering number) of TnT_n with balls of radii 1≪γn≪n1/21\ll γ_n \ll n^{1/2} is asymptotically equivalent to Cn/(σ2γn2)\mathfrak{C} n /(σ^2 γ_n^2) (resp. Cn/(4σ2γn2)\mathfrak{C} n /(4σ^2 γ_n^2)), for a universal constant C\mathfrak{C}. We also show that the order of limits can be exchanged and that the same limit occurs when we first consider the limiting covering number of TnT_n by balls of fixed radius RR, and thereafter let RR tend to ∞\infty. We then extend these results to the Brownian continuum random tree conditioned to have mass 11. For this model, we prove that the packing (resp. covering) number with balls of radii ε\varepsilon are asymptotically equivalent to C/ε2\mathfrak{C}/\varepsilon^2 (resp. C/(4ε2)\mathfrak{C}/(4\varepsilon^2)), for the same constant C\mathfrak{C}, in contrast with a conjecture by Aldous. Finally, our results answer a question by Devroye, Eide and Pralat on the asymptotic burning number of mean 11 finite variance Bienaymé trees of size nn, by showing that it lies between (c+o(1))n1/3(c+o(1)) n^{1/3} and (2c+o(1))n1/3(2c+o(1))n^{1/3} with high probability, for a constant cc that only depends on the offspring variance.

Evidence graph

No public relationships recorded yet.

Integrity note: This page is a factual metadata record created by deterministic ingestion. It is not a claim that the work moves a mathematical frontier or has been independently verified.