Novel Approach for Memory Storage Systems with Chaos-Chaos Intermittency

Abstract

In nonlinear systems with a barrier/threshold, the synchronization under a weak external input signal is strengthened using appropriate additive stochastic noise. This phenomenon is known as stochastic resonance. Recent progress in the application of stochastic resonance has shown that the presence of additive noise enhances memory functions in memory elements with bi-stable oscillations, even under extremely low power consumption. In addition to additive noise, deterministic chaotic behavior induces chaotic resonance, a phenomenon that is similar to stochastic resonance. Chaotic resonance emerges in nonlinear dynamical systems with chaos-chaos intermittency, where the chaotic orbit moves among separated attractor regions through an attractor-merging bifurcation. In previous studies, a higher sensitivity of chaotic resonance compared to that of stochastic resonance was reported. In this context, we hypothesized that memory devices based on chaotic resonance can be used to realize a novel device for storing memory with lower power consumption than in devices based on stochastic resonance. In this study, to prove this hypothesis, we induce the attractor-merging bifurcation in a cubic map system, which is the simplest model for emerging chaotic resonance. We use one approach for adjusting the internal system parameters under noise-free conditions and another for applying stochastic noise, which is similar to the conventional approach using stochastic resonance. By comparing the performance of these approaches, we reveal that the former exhibits a higher memory storing ability than the latter stochastic approach, even under weaker memory storage input signals. This superiority allows the development of memory devices with low power consumption. The method involving chaotic resonance facilitates the improvement of memory devices that were previously limited to the application of stochastic resonance.