Dynamical sensitivity of recurrence and transience of branching random walks
Abstract
Consider a sequence of i.i.d. random variables Xn where each random variable is refreshed independently according to a Poisson clock. At any fixed time t the law of the sequence is the same as for the sequence at time 0 but at random times almost sure properties of the sequence may be violated. If there are such exceptional times we say that the property is dynamically sensitive, otherwise we call it dynamically stable. In this note we consider branching random walks on Cayley graphs and prove that recurrence and transience are dynamically stable in the sub-and supercritical regime. While the critical case is left open in general we prove dynamical stability for a specific class of Cayley graphs. Our proof combines techniques from the theory of ranching random walks with those of dynamical percolation.