Enhancement signal detection in underwater acoustic noise using level dependent estimation time-frequency de-noising technique

Abstract

ABSTRACTIn sonar and underwater digital communication, optimal signal detection is imperative. In many applications, additive white Gaussian noise (AWGN) is assumed; thus, a linear correlator (LC), which is known to be optimal in the presence of AWGN, is normally used. However, underwater acoustic noise (UWAN) affects the reliability of signal detection in applications in which the noise originates from multiple sources and doesn’t follow the AWGN assumption. As a result, an LC detector performs poorly in tropical shallow waters. Accordingly, this study aims to develop a detection method for improving detection probability () by using a time–frequency denoising method based on the S-transform with multi–level threshold estimation. The UWAN used for the validation is sea truth data collected at Desaru beach on the eastern shore of Johor in Malaysia with the use of broadband hydrophones. The performances of four different detectors, namely, the proposed Gaussian noise injection detector (GNID), a locally optimal (LO) detector, a sign correlation (SC) detector, and a conventional LC detector, are evaluated according to their values. For a time-varying signal, given a false alarm probability of 0.01 and a value of 90 percent, the energy-to-noise ratios of the GNID, LO detector, SC detector, and LC detector are 8.89, 10.66, 12.7, and 12.5 dB, respectively. Among the four detectors, the GNID using the S-transform denoising method achieves the best performance.