A comparative study of stochastic resonance for a model with two pathways by escape times, linear response, invariant measures and the conditional Kolmogorov-Smirnov Test

Tommy LiuarXiv 2018 · DOI · arXiv

Abstract

Stochastic resonance (SR) is a phenomenon where noise enhances the response of a nonlinear system to a weak periodic signal. While well-understood in one-dimensional systems with a single transition path, complex landscapes with multiple pathways present unique challenges. This paper investigates a two-dimensional "Mexican Hat" potential model with two distinct pathways between metastable states. We demonstrate that when forcing causes these pathways to oscillate in synchronization, standard detection metrics—such as signal-to-noise ratio and linear response—fail to identify resonance. To address this, we introduce a "conditional Kolmogorov-Smirnov (KS) test" based on escape time distributions. Our results reveal a novel "double frequency" resonance and provide a robust statistical framework for detecting SR in multi-pathway systems where traditional measures go blind.

Introduction

Stochastic resonance (SR) defies the common intuition that noise is purely detrimental to signal processing. In systems characterized by metastable states separated by a potential barrier, a signal too weak to cause a transition on its own can become effective when supplemented by random fluctuations. By synchronizing the noise-induced jumps with the periodic forcing, the system’s output becomes more coherent.

Most existing literature focuses on one-dimensional models where a single path connects two states. However, real-world systems—ranging from climate patterns to neuronal firing—often involve complex energy landscapes with multiple transition routes. This study explores how the geometry of these pathways affects the detection and manifestation of SR, specifically focusing on the transition from alternating to synchronized barrier oscillations.

The Two-Pathway Model

To study these dynamics, we utilize a two-dimensional potential surface, often visualized as a "Mexican Hat" with two wells and two saddles (mountain passes). The system is described by a diffusion process with drift governed by this potential and subjected to periodic forcing at an angle $\phi$.

Two primary forcing regimes are analyzed:

  1. Alternating Saddles ($\phi = 0^\circ$): The forcing causes one barrier to lower while the other rises, effectively mimicking a standard 1D double-well system.
  2. Synchronized Saddles ($\phi = 90^\circ$): The forcing causes both barriers to rise and fall simultaneously.

In the synchronized case, the symmetry of the system ensures that the long-term probability of finding the particle in either well (the invariant measure) remains a constant 50/50. This symmetry creates a significant hurdle for traditional analysis.

The Failure of Standard Measures

We evaluated six classic measures of stochastic resonance ($M_1$ through $M_6$), including Linear Response, Signal-to-Noise Ratio (SNR), Energy, and Relative Entropy. These metrics typically rely on the time-evolution of the system's average state or its power spectrum.

Our findings show that for the synchronized saddle case ($\phi = 90^\circ$), all six measures fail to detect resonance. Because the invariant measures do not oscillate, the "signal" appears absent from the long-term averages. This suggests that standard tools are "blind" to the underlying synchronization occurring at the level of individual trajectories.

The Conditional Kolmogorov-Smirnov Test

To overcome the limitations of averaging, we propose a method focused on the "path level" of the system: the distribution of individual escape times. We developed a conditional Kolmogorov-Smirnov (KS) test. Unlike a standard KS test, which compares data against a static distribution, this version tests if the observed escape times fit a conditional probability density function that accounts for the specific phase of the forcing at the moment the particle entered the well.

By analyzing the timing of individual jumps rather than the average population density, the conditional KS test can identify whether transitions are being "clocked" by the periodic signal, even when the overall population remains balanced.

Key Results: Double Frequency Resonance

Numerical simulations using Euler-integrated stochastic differential equations (SDEs) revealed a striking phenomenon in the synchronized regime. While 1D systems show resonance at the forcing frequency $\Omega$, the two-pathway synchronized model exhibits resonance at $2\Omega$.

This "Double Frequency" effect occurs because the optimal time for a particle to escape occurs twice per period—once for each available path. Histogram analysis of escape times confirmed clear periodic peaks at $0, T/2, T, 3T/2$, and so on. The conditional KS test successfully identified this resonance with 99% confidence, providing a reliable "footprint" of the multi-pathway dynamics that standard measures missed entirely.

Conclusion and Future Directions

This research demonstrates that the geometry of a potential landscape fundamentally changes the signature of stochastic resonance. In multi-pathway systems, the absence of a signal in the invariant measure does not imply an absence of resonance. The conditional KS test offers a powerful alternative for detecting synchronization in complex systems.

Future work should apply this statistical framework to "sparse" real-world datasets, such as ice-core climate records or biological sensor data, where transitions are infrequent and the underlying landscape may involve multiple competing pathways. Understanding these higher-dimensional resonances is crucial for accurately modeling systems that operate in noisy, complex environments.

Summary updated

Abstract

We consider stochastic resonance for a diffusion with drift given by a potential, which has two metastable states and two pathways between them. Depending on the direction of the forcing, the height of the two barriers, one for each path, will either oscillate alternating or in synchronisation. We consider a simplified model given by a continuous time Markov Chains with two states. This was done for alternating and synchronised wells. The invariant measures are derived for both cases and shown to be constant for the synchronised case. A PDF for the escape time from an oscillatory potential is studied. Methods of detecting stochastic resonance are presented, which are linear response, signal-noise ratio, energy, out-of-phase measures, relative entropy and entropy. A new statistical test called the conditional Kolmogorov-Smirnov test is developed, which can be used to analyse stochastic resonance. An explicit two dimensional potential is introduced, the critical point structure derived and the dynamics, the invariant state and escape time studied numerically. The six measures are unable to detect the stochastic resonance in the case of synchronised saddles. The distribution of escape times however not only shows a clear sign of stochastic resonance, but changing the direction of the forcing from alternating to synchronised saddles an additional resonance at double the forcing frequency starts to appear. The conditional KS test reliably detects the stochastic resonance. This paper is mainly based on the thesis "Stochastic Resonance for a Model with Two Pathways"
BibTeX
@misc{liu2018comparative,
  author = {Tommy Liu},
  title = {{A comparative study of stochastic resonance for a model with two pathways by escape times, linear response, invariant measures and the conditional Kolmogorov-Smirnov Test}},
  year = {2018},
  eprint = {1803.00880},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/1803.00880},
}

Loading related content...

Figure