부산대학교 도서관

Time-frequency analysis (TFA) transforms a one-dimensional signal into a two-dimensional image for its TF characteristics, which provides a framework that can be combined with digital image processing. In shallow water, a wide-band acoustic signal received by a hydrophone shows a unique dispersion structure, from which the source and propagation environment information can be obtained. This paper proposes a method combining two-subiteration skeleton extraction algorithm with advanced time-frequency analysis to separate and extract the dispersion curves from data received by a single hydrophone without any prior environmental information. The source location is then determined according to the time difference of dispersion curves.