Detection of echoes using time-frequency analysis techniques

Abstract

The following is a presentation of echo detection techniques based on time-frequency signal analysis for the measuring of thickness in thin multilayer structures. These techniques are shown to provide high-resolution signal characterization in a time-frequency space, and good noise rejection performance. In particular, the short-time Fourier transform, the Gabor expansion, the cross-ambiguity function and the Wigner-Ville distribution are analyzed and compared with techniques such as the logarithmic power spectrum, cepstrum and the segmented chirp Z-Transform. A suitable operating procedure was set up, based on an initial emulation phase in which simulated signals were considered, followed by a second phase in which real signals were processed. The results show the optimum performances of these new techniques compared with the traditional ones and, in particular, that the accurate measurement of thickness can be obtained also when waveform transients partially overlap.