Cells with many facets in a Poisson hyperplane tessellation (Q1684661)

From the Introduction: One of the classical models in stochastic geometry to generate a random tessellation is the construction via a Poisson hyperplane process. A Poisson hyperplane process consists of countably many random hyperplanes in \(\mathbb{R}^d\) chosen in such a way, that their distribution is translation invariant, the distribution of the direction of the hyperplanes follows a directional distribution \(\varphi\), that is an even probability measure on the unit sphere which is not concentrated on some great circle, and the number of hyperplanes hitting an arbitrary convex set \(K\) is Poisson distributed. Such a Poisson hyperplane process tessellates \(\mathbb{R}^d\) into countably many convex polytopes, the tiles of the tessellation, see [\textit{R. Schneider} and \textit{W. Weil}, Stochastic and integral geometry. Berlin: Springer (2008; Zbl 1175.60003), Theorem 10.3.2]. The distribution of a tile chosen at random is the distribution of the so-called typical cell \(Z\), a random polytope. The typical cell has been investigated intensively in the past decades, numerous papers have been dedicated to describe quantities associated with this cell. The expected number of facets \(f(Z)\) of the typical cell and the expected volume \(V_d(Z)\) are known [loc. cit., Chapter 10]. But in almost all cases, the distribution of these quantities is out of reach, and even good approximations are extremely difficult and unknown so far. The main theorems fill this gap for the number of facets of \(Z\), giving precise asymptotics for the tails of the distribution. The paper is organized as follows. Section 2 is devoted the so-called complementary theorem, which is a practical disintegration of the distribution of the typical cell \(Z\). In Section 3, the authors provide the required geometric ingredients which deal with the approximation of polytopes by polytopes with fewer facets. Section 4 is devoted to the proof of he main results about cells with many facets. Finally, in Section 5, the authors prove the results which deal with the big cells, that is with a large \(\Phi\) content (the \(\Phi\)-content measures in a certain sense the size of the convex set depending on the directional distribution \(\varphi\) of the hyperplane tessellation).