The bi-dimensional Directed IDLA forest
Nicolas Chenavier, David Coupier, Arnaud Rousselle
TL;DR
This paper develops three infinite-IDLA models on ${\mathbb Z}^2$ with sources along the vertical axis, proving invariance under vertical translations, stabilization, mixing, and shape theorems. It introduces a directed infinite-volume IDLA forest $\mathcal{F}_\infty$ built from three aggregates $A_n[\infty]$, $A_n^*[\infty]$, $A_n^\dag[\infty]$, and shows that the forest is stationary, spans ${\mathbb Z}^2$, and comprises a countable union of directed trees rooted on $I(\infty)$. The work overcomes challenges from the loss of Abelian property and long-range dependencies by establishing weak stabilization and robust bounds, enabling a consistent horizontal limit and a well-defined forest. It also provides detailed shape theorems for the aggregates (including Poisson variants) and discusses open questions about the forest’s finer properties and connections to the classical IDLA tree $\mathcal{T}_\infty$. Overall, the results contribute a stationary, translation-invariant framework for directed IDLA-like growth and a new random forest object with potential links to scaling limits and the Brownian web.
Abstract
We investigate three types of Internal Diffusion Limited Aggregation (IDLA) models. These models are based on simple random walks on $\mathbf{Z}^2$ with infinitely many sources that are the points of the vertical axis $I(\infty)=\{0\}\times\mathbf{Z}$. Various properties are provided, such as stationarity, mixing, stabilization and shape theorems. Our results allow us to define a new directed (w.r.t. the horizontal direction) random forest spanning $\mathbf{Z}^2$, based on an IDLA protocol, which is invariant in distribution w.r.t. vertical translations.
