Covering of inner subsets by the confined random walk (Work in Progress)

Abstract

We consider the simple random walk that is conditioned to stay in a large domain (D_N = N \cdot D \cap \mathbb{Z}^d ) with typical size N, which is a Doob’s transform of simple random walk by the first eigenvector of the walk killed upon exiting D_N. Considering an inner subset A_N of D_N, we prove asymptotics for the covering time of A_N by the confined walk, meaning the time at which the walk has visited all the points of A_N. We get that the covering time is asymptotically N^d log N with a constant that depends on the minimal value of the eigenvector on A_N. We also prove that restricting to level sets of the eigenvector, we get Gumbel fluctuations and a convergence towards a Poisson point process. This work relies on a previous result from the same author which establishes a coupling between the confined walk and random interlacements, on which we can use the sketch of Belius’ proof for the covering time of SRW on the torus.

Nicolas Bouchot
Nicolas Bouchot
PhD student at Laboratoire de Probabilités, Statistiques et Modélisation (LPSM)

My research focuses on polymer models, random walks in random environments and statistical mechanics.