Fast computation of wide-band ambiguity function and matched filtering in active sonars

Ambiguity function is considered mainly in two areas of active sonar processing. First, it is used for evaluating the performance of different types of waveforms. Second, it decides the performance of active sonars using matched filtering. The narrow-band ambiguity function is widely used in many active sonar systems. But, when the signal under analysis is not narrow-band, this particular form of ambiguity function is no longer appropriate. We then have to consider the wide-band ambiguity function (henceforth shortened as WB AF). But such broadband functionals cannot be computed by standard techniques. In this paper, we have studied the WB AF for sonar applications and implemented it using the fast Fourier Mellin transform algorithm. Matched filtering using this implementation is then compared with the conventional scheme using narrow-band assumption. Our simulations show an improvement of 2 dB in the matched filter performance using the proposed method thus clearly highlighting the importance of this time-frequency representation in sonars.