Quadratic Time-Frequency Transforms-Based Brillouin Optical Time-Domain Reflectometry

Linear and quadratic Time-Frequency (T-F) transforms are proposed for the signal processing of the Brillouin optical time-domain reflectometry, to improve the system performance, in terms of the transition responsivity and frequency resolution. Various T-F transforms, including linear T-F transform (short time Fourier transform) and quadratic transforms (Choi-Williams, Zhao-Atlas-Marks, smoothed pseudo Wigner-Ville, S-method, and adaptive spectrogram) are applied to the experimental backscattered time-domain spontaneous Brillouin signals. Multiple T-F approaches can be jointly applied to efficiently improve the time and frequency resolutions simultaneously. Results show that the SWPV transform provides the best transition responsivity and frequency resolution simultaneously among the six T-F transforms because of its smoothing operations on frequency and time axis for reducing the cross-term interference.