English

On the rate of convergence in the Hall-Janson coverage theorem

Probability 2025-06-12 v4

Abstract

Consider a spherical Poisson Boolean model ZZ in Euclidean dd-space with d2d \geq 2, with Poisson intensity tt and radii distributed like rYrY with r0r \geq 0 a scaling parameter and YY a fixed nonnegative random variable with finite (2d2)(2d-2)-nd moment (or if d=2d=2, a finite (2+ε)(2 + \varepsilon)-moment condition for some ε>0\varepsilon >0). Let ARdA \subset {\bf R}^d be compact with a nice boundary. Let α\alpha be the expected volume of a ball of radius YY, and suppose r=r(t)r=r(t) is chosen so that αtrdlogt(d1)loglogt\alpha t r^d - \log t - (d-1) \log \log t is a constant independent of tt. A classical result of Hall and of Janson determines the (non-trivial) large-tt limit of the probability that AA is fully covered by ZZ. In this paper we provide an O((loglogt)/logt)O((\log \log t)/\log t) bound on the rate of convergence in that result. With a slight adjustment to r(t)r(t), this can be improved to O(1/logt)O(1/\log t).

Keywords

Cite

@article{arxiv.2405.16461,
  title  = {On the rate of convergence in the Hall-Janson coverage theorem},
  author = {Mathew D. Penrose and Xiaochuan Yang},
  journal= {arXiv preprint arXiv:2405.16461},
  year   = {2025}
}

Comments

24 pages, one figure. Added some extra explanations and a diagram