Author :
Courmontagne, Philippe ; Julien, Gregory ; Bouhier, Marie Edith
Author_Institution :
Equipe Signaux et Syst., Maison des Technol., ISEN Toulon, Toulon, France
Abstract :
The Pulse Compression is a technique widely used in many domains to magnify (point out) a useful signal embedded in a noise. This classical method maximizes the signal to ratio, when the disturbing terms correspond to a white Gaussian noise and when the useful signal is well known. In a transmission context, because of several physical phenomena (propagation, median frequency attenuation, Doppler effect...) the useful signal should be considered as a realization of a stochastic process. Furthermore, the submarine noise, although generally Gaussian, is not really white, even in the useful bandwidth. For these reasons, the classical Pulse Compression is never optimal in a real context. To answer this problem, we have proposed several new pulse compression schemes, based on the use of the Stochastic Matched Filter combined with the Wigner-Ville distribution (the WV-SMF-PC). To enhance results given by the WV-SMF-PC in terms of SNR gain, we propose in this paper to use an atomic approach to compute the time-frequency plane, in place of the classical Wigner-Ville transform. Indeed, the use of such an approach avoids the interferences, due to the Wigner-Ville transform, to appear. Several results are proposed and discussed.
Keywords :
Doppler effect; Wigner distribution; matched filters; pulse compression; transforms; Doppler effect; Ifremer sea trials signals; Wigner-Ville distribution; Wigner-Ville transform; enhanced WV-SMF-PC; interferences; median frequency attenuation; pulse compression; stochastic matched filter; submarine noise; time-frequency plane; white Gaussian noise; Chirp; Matched filters; Noise measurement; Signal to noise ratio; Time frequency analysis; Transforms; Atomic expansion; DCT; Pulse compression; Stochastic Matched Filter; Wigner-Ville transform;