<title>Estimation of target vibration spectra from laser radar backscatter using time-frequency distributions</title>

Abstract

A time-frequency distribution (TFD) signal processor, developed at the Applied Physics Laboratory, is currently under evaluation using simulated signals and actual laser vibration sensor (LVS) data that we collected on various ship targets. Preliminary results for one instantaneous frequency (IF) estimator implementation, the smoothed cross Wigner-Ville Distribution (XWVD), indicate 8 to 10 dB demodulation (CNR) advantage compared to a digital FM limiter-discriminator. A second approach, using the unsmoothed XWVD TFD, demonstrated a 3-5 dB advantage. Regarding spectral estimation, we are investigating performance of our reduced interference distribution (RID) implementation through comparison with the short-time Fourier transform (STFT). From the LVS data processed, indications are that a significant increase in spectral and temporal resolution exists using our RID approach. Our processor also provided improved detectability over the STFT for transient signals and short-lived sinusoids. Significant correlation between accepted acoustic lines and LVS-derived vibration lines are indicated. Details are presented that describe our signal simulation, the LVS measurements, and signal processing implementations along with assumptions based on measured speckle-induced amplitude modulation.