Chaotic characteristics analyses of underwater acoustic signals

It is very important to study the dynamic properties of underwater acoustic signals for improving the performance of underwater acoustic systems. This paper studies the chaotic characteristics of noise signals radiated from ships using power spectrum analysis, singular spectrum analysis, Lyapunov exponents and correlation dimension. All studies proved that there exist chaotic characteristic in this type of signal. Experiments showed that the fractal dimensions of different signal sources are different. It would be very useful to be able to classify this type of signal. As a byproduct, a method for estimating Lyapunov exponents based on the idea of system identification is presented in this paper.