Regulatory effects of time-frequency demodulation mechanisms on spectral and spatial resolution in STFT-BOTDR

Structural health monitoring (SHM) of large-scale infrastructure systems demands high-precision, real-time sensing to enable early warning and mitigate structural failure risks. Short-time Fourier transform-based Brillouin optical time-domain reflectometry (STFT-BOTDR) has shown significant potential in this context. However, its performance is often limited by the absence of well-defined guidelines for configuring key time-frequency parameters. This study systematically examines the effects of window function type, window length, and sliding step size on both spectral and spatial resolutions. Theoretical analysis and experimental validation reveal that while smooth window functions effectively suppress spectral leakage and broaden the spectrum, reducing peak resolvability. Window length governs spectral and spatial resolution, with the full width at half maximum (FWHM) of the Brillouin gain spectrum found to scale inversely with the logarithm of the window length. Furthermore, the sliding step size directly affects the rising edge sharpness in Brillouin frequency shift distributions, thereby impacting spatial precision. Based on these insights, we introduce two optimization criteria: the spectral resolution-driven (SRD) criterion for selecting window length, and the spatial sampling accuracy (SSA) criterion for determining step size. These findings offer practical guidance for optimizing time-frequency parameter configurations to enhance the overall performance of STFT-BOTDR.