The limit shape of the zero cell in a stationary Poisson hyperplane tessellation. (Q1879879): Difference between revisions

From MaRDI portal
Import240304020342 (talk | contribs)
Set profile property.
Set OpenAlex properties.
Property / full work available at URL
 
Property / full work available at URL: https://doi.org/10.1214/aop/1079021474 / rank
 
Normal rank
Property / OpenAlex ID
 
Property / OpenAlex ID: W1998334355 / rank
 
Normal rank

Revision as of 20:53, 19 March 2024

scientific article
Language Label Description Also known as
English
The limit shape of the zero cell in a stationary Poisson hyperplane tessellation.
scientific article

    Statements

    The limit shape of the zero cell in a stationary Poisson hyperplane tessellation. (English)
    0 references
    0 references
    0 references
    0 references
    15 September 2004
    0 references
    The authors generalize to higher dimensions, and to not necessarily isotropic stationary Poisson hyperplane process, a proof by \textit{I. N. Kovalenko} [Cybern. Syst. Anal. 33, No. 4, 461--467 (1997); translation from Kibern. Sist. Anal. 1997, No. 4, 3--10 (1997; Zbl 0922.60022)] of D. G. Kendall's conjecture concerning the shape of the zero cell of the random tessellation generated by the process. In the case of a stationary and isotropic Poisson line process in the plane, Kovalenko proved that, given that the area of the zero cell tends to infinity, its shape tends to circularity. Let \(X\) be a stationary Poisson hyperplane process in \(R^d\), \(d\geq 2\), with intensity \(\lambda>0\). An even measure \(\varphi\) on the unit sphere describes the distribution of the unit normals of the hyperplane of \(X\). By Minkowski's existence and uniqueness theorem from the geometry of convex bodies, there exists a unique centrally symmetric convex body \(B\), called the direction body of \(X\), for which \(\varphi\) is the area measure. Let \(Z_0\) be the zero cell of the tessellation induced by \(X\). Then it is shown that there exists a constant \(c_0\), depending only on \(B\), such that the following is true. If \(\varepsilon\in (0,1)\) and \(I=[a,b)\) is any interval with \(a^{1/d}\lambda\geq\sigma_0>0\), then \[ P(r_B(Z_0)\geq\varepsilon\mid V(Z_0)\in I)\leq c \exp\{-c_0 \varepsilon^{d+1} a^{1/d}\lambda\}, \] where \(P\) denotes probability, \[ r_B(Z_0):=\inf \{s/r-1: rB\subset Z_0\subset sB+z,\;z\in R^d,\;r,s>0\}, \] and \(c\) is a constant depending on \(B,\varepsilon,\sigma_0\). As a consequence, the interesting result \[ \lim_{a\to\infty}P(r_B(Z_0)\geq \varepsilon\mid V(Z_0)\geq a)=0 \] for every \(\varepsilon>0\) is obtained, which shows that the conditional law for the shape of \(Z_0\), given a lower bound for \(V(Z_0)\), converges weakly, as that lower bound tends to infinity, to the law concentrated at the shape of the direction body \(B\) associated to the given process
    0 references
    0 references
    asymptotic shape
    0 references
    Crofton cell
    0 references
    Kendall's conjecture
    0 references
    hyperplane tessellation
    0 references
    Poisson hyperplane process
    0 references
    typical cell
    0 references
    zero cell
    0 references

    Identifiers