A wait-and-reset strategy to synchronizing chaotic systems

A convenient way of synchronizing chaotic systems is to continuously feed one output of a driving system to the response system as one of its inputs. However, the success of this method depends crucially on the fact that the sub-lyapunov exponents of the reduced system are negative, which may not hold for all of the system variables. We propose that this shortcoming sometimes can be overcome by not constantly monitoring the response system. We show that this somewhat counter-intuitive mechanism works because it has taken advantage of the fact that accompanying a chaotic attractor usually is a swirling motion in the phase space, which sometimes can be favorably utilized to suppress the exponential divergence typical of a chaotic system when resetting is applied regularly at a suitably chosen interval of time.