hitting time of a subset

Suppose you have the simple random walk on $L=\mathbb{Z}^n,$ and $A \subset L$ is a subset (in my case, the subset is a union of hyperplanes, so in particular infinite). There is a random variable $T(A)$ -- the expected time to get from the origin to $A.$ The question is: is there some standard technology to upper bound the expectation of $T(A)?$ Any particularly recommended things to read?

edited Sep 3 at 16:26

asked Sep 3 at 16:20

Igor Rivin
31.3k●3●37●71

Perhaps this paper will provide ideas: arxiv.org/abs/1206.6568 . "Lyapunov exponents of random walks in small random potential: the lower bound" by Mountford and Mourrat. They compute, under various conditions, "the ﬁrst time at which the random walk [in $\mathbb{Z}^d$] crosses the hyperplane orthogonal to $\ell$ lying at distance $n$ from the origin." – Joseph O'Rourke Sep 3 at 16:58

Thanks! I will check it out... – Igor Rivin Sep 3 at 17:38

There is a book by Sydney Redner: A guide to first-passage processes, Cambridge UP, 2001. It rather discusses physical processes, and continuous random walks, often with n<=3, but may help to find bounds also for the discrete case and general n. – Karl Fabian Sep 3 at 20:36

@Karl: thanks, I will try to look it up... – Igor Rivin Sep 3 at 22:32

Browse other questions tagged pr.probability random-walk or ask your own question.

question feed