Well-posedness of the stochastic neural field equation with discontinuous firing rate

Abstract

We study the existence and uniqueness of mild solutions to the deterministic and the stochastic neural field equation with Heaviside firing rate. Since standard well-posedness results do not apply in case of a discontinuous firing rate, we present a monotone Picard iteration scheme to show the existence of a maximal mild solution. Further, we illustrate that general uniqueness does not hold, and therefore investigate uniqueness under suitable additional properties of the solutions. Here a novel criterion, the so-called absolute continuity condition is introduced. Moreover, we observe regularisation by noise: with a suitable choice of spatially correlated additive noise uniqueness is restored without imposing any additional structural assumptions.