Mechanically-Gated Stochastic Resonance for All-Optical Acoustic Sensing in Noisy Environments

Weak signal detection in harsh and noisy environments remains a critical challenge across diverse fields. Stochastic resonance (SR) offers a counterintuitive strategy by exploiting noise to enhance, rather than suppress, weak signals, yet its practical deployment is severely limited by poor frequency selectivity, inadequate adaptability to fluctuating noise levels, and reliance on offline post-processing. Here we report the discovery and experimental demonstration of mechanically-gated stochastic resonance (MGSR), a novel physical mechanism that couples stochastic resonance with intrinsic mechanical resonance. By using mechanical resonance to actively modulate the bistable potential, MGSR achieves targeted frequency-line selection, robust adaptation to variable noise conditions, and markedly improved signal-to-noise ratio (SNR) gain for weak signal detection. We further realize a real-time, electromagnetic-interference-immune all-optical bistable system through a fiber Bragg grating feedback mechanism, enabling hardware-level MGSR free of computational latency. In tests conducted under realistic ocean ambient noise, our system attains a substantial SNR gain of 19.26 dB at the designed frequency, representing a clear advance for hardware-based stochastic resonance acoustic sensors. This work deepens the understanding of SR, establishes a robust all-optical architecture, and opens a scalable pathway toward high-performance, noise-resilient acoustic and vibration sensing for environmental monitoring, structural health diagnosis, and underwater safety.