Random-search efficiency in a bounded interval with spatially heterogeneous diffusion coefficient

Abstract

We consider random walkers searching for a target in a bounded one-dimensional heterogeneous environment, in the interval $[0,L]$, where diffusion is described by a space-dependent diffusion coefficient $D(x)$. Boundary conditions are absorbing at the position of the target (set at $x=0$) and reflecting at the border $x=L$. We calculate and compare the estimates of efficiency $\varepsilon_1=\langle 1/ t\rangle$ and $\varepsilon_2=1/\langle t \rangle$. For the Stratonovich framework of the multiplicative random process, both measures are analytically calculated for arbitrary $D(x)$. For other interpretations of the stochastic integrals (e.g., It\^o and anti-It\^o), we get general results for $\varepsilon_2$, while $\varepsilon_1$ is obtained for particular forms of $D(x)$. The impact of the diffusivity profile on these measures of efficiency is discussed. Symmetries and peculiar properties arise when the search starts at the border ($x_0=L$), in particular, heterogeneity spoils the efficiency of the search within the Stratonovich framework, while for other interpretations the searcher can perform better in certain heterogeneous diffusivity profiles.