Branching-selection particle systems and inverse first passage problems

Abstract

A generalised inverse first passage problem asks whether, given a probability measure $p$ on $[0,\infty ]$, one can find a boundary $b:[0,\infty ]\to \mathbb{R}$ such that the stopping time:\[τ:=\inf\left\{t:Λ\int_0^t ω(W_s-b(s))ds \geq U\right\}\] has distribution $p$, where $U\sim Exp(1)$, $\hat{I}›\in (0,\infty )$ and $\ddot{I}‰$ is a monotonic decreasing function. We construct a branching-selection particle system whose hydrodynamic limit is governed by a free boundary problem and connect this to the generalised inverse first passage problem. In the $N$-particle system, particles move as independent Brownian motions, branch at a prescribed rate, and are removed at a rate proportional to their location relative to a position $b^N(t)$ which is a function of the empirical distribution. We identify the limit of $b^N$ as the solution of the inverse first passage problem.

12 pages, no figures