English

The extremal position of a branching random walk on the general linear group

Probability 2024-12-11 v2

Abstract

Consider a branching random walk (Gu)uT(G_u)_{u\in \mathbb T} on the general linear group GL(V)\textrm{GL}(V) of a finite dimensional space VV, where T\mathbb T is the associated genealogical tree with nodes uu. For any starting point vV{0}v \in V \setminus\{0\} with v=1\|v\|=1 and x=RvP(V)x = \mathbb R v \in \mathbb P(V), let Mnx=maxu=nlogGuvM^x_n=\max_{|u| = n} \log \| G_u v \| denote the maximal position of the walk logGuv\log \| G_u v \| in the generation nn. We first show that under suitable conditions, limnMnxn=γ\lim_{n \to \infty} \frac{M_n^x }{n} = \gamma almost surely, where γR\gamma \in \mathbb R is a constant. Then, in the case when γ=0\gamma = 0, under appropriate {\textit boundary conditions}, we refine the last statement by determining the rate of convergence at which MnxM_n^x converges to -\infty. We prove in particular that limnMnxlogn=32α\lim_{n \to \infty} \frac{M_n^x}{\log n} = -\frac{3}{2\alpha} in probability, where α>0\alpha >0 is a constant determined by the boundary conditions. Analogous properties are established for the minimal position. As a consequence we derive the asymptotic speed of the maximal and minimal positions for the coefficients, the operator norm and the spectral radius of GuG_u.

Keywords

Cite

@article{arxiv.2206.04941,
  title  = {The extremal position of a branching random walk on the general linear group},
  author = {Ion Grama and Sebastian Mentemeier and Hui Xiao},
  journal= {arXiv preprint arXiv:2206.04941},
  year   = {2024}
}