Noise-induced intermittency of a reflexive model with symmetry-induced equilibrium manifold

Abstract

We discuss a route to intermittency based on the concept of reflexivity, namely on the interaction between observer and stochastic reality. A simple model mirroring the essential aspects of this interaction is shown to generate perennial out of equilibrium condition, intermittency and 1/f-noise. In the absence of noise the model yields a symmetry-induced equilibrium manifold with two stable states. Noise makes this equilibrium manifold unstable, with an escape rate becoming lower and lower upon time increase, thereby generating an inverse power law distribution of waiting times. The distribution of the times of permanence in the basin of attraction of the equilibrium manifold are analytically predicted through the adoption of a first-passage time technique. Finally we discuss the possible extension of our approach to deal with the intermittency of complex systems in different fields.