Nonlocal random motions: The trapping problem

TL;DR

This paper investigates the challenges of confining nonlocal Le9vy stable processes, like the Cauchy process, within finite regions, highlighting differences from classical Brownian motion confinement.

Contribution

It provides a detailed analysis of trapping nonlocal jump processes, especially the Cauchy process, extending understanding of their confinement properties compared to Gaussian processes.

Findings

Nonlocal jump processes are inherently difficult to confine within finite regions.

The Cauchy process exhibits unique trapping challenges compared to Brownian motion.

Results extend to general non-Gaussian, jump-type stochastic processes.

Abstract

L\'evy stable (jump-type) processes are examples of intrinsically nonlocal random motions. This property becomes a serious obstacle if one attempts to model conditions under which a particular L\'evy process may be subject to physically implementable manipulations, whose ultimate goal is to confine the random motion in a spatially finite, possibly mesoscopic trap. We analyze thisissue for an exemplary case of the Cauchy process in a finiteinterval. Qualitatively, our observations extend to general jump-type processes that are driven by non-gaussian noises, classified by the integral part of the L\'evy-Khintchine formula.For clarity of arguments we discuss, as a reference model, the classic case of the Brownian motion in the interval.