Underwater pulse position modulation (PPM) transmission uses acoustic pulses since they allow communication, remote control of underwater vehicles, and positioning estimation; the last one when more than one detector is present by using only one technology, that is, the same transmission module. On the other hand, PPM pulses are difficult to detect since they carry information on signal delays, assuring that reliable information delivery both in shallow- and deep-sea channels may be considered a timely challenge due to strong multipath and other interference sources/causes giving rise to signal fluctuation and high noise level. This paper proposes a transceiver architecture characterized by a signal analyzer and an adaptive detection based on the channel features. The signal analysis is based on the Wigner–Ville transform (WVT) that converts a 1-D signal into an image. Edge detection procedures performed on the obtained image so as to discover the interference features and understand the main characteristics of multipath with their adverse effects on detection of pulse position information have been taken into account. Based on this, maximal ratio combining (MRC) and equal gain combining (EGC) for RAKE reception and truncated channel equalization have been considered. Numerical results showing the ability of the proposed transceiver to detect pulses in the presence of multipath and interference generated by external acoustic sources have been carried out also by taking into account the estimation error effects and by comparing the actual performance with ideal cases. Finally, field tests corroborate the reliability of the proposed approach.