Research Shallow Water Acoustic Normal Modes Separation Method

The normal mode separation and information extraction is one of hot topics in the marineacoustics. In this paper, we systematically introduce the existing normal mode separationmethod, and present Waveform Matching-Inverse Dedispersion Transform (WM-IDDT)normal mode separation method. The main points of the dissertation include:(1) In the Pekeris waveguide, the mode depth functions change little along with frequency.And as the frequency band of the signal is relatively narrow, the waveform change ismainly caused by the dispersion of the waveguide. In this case, after performing DDTon a received signal, the resulting signal is composed of source signals with differentamplitudes and positions (corresponding to the arriving time) in reality. With the help ofnumerical optimization algorithm and waveform matching, we can determine therelative amplitude and position of each normal mode. Then the normal mode can beseparated, and the amplitude and the arriving time of each normal mode can be known.With the help of vertical line array, the mode depth function can also be known. Theinverse dedispersion transform is presented, and the real received waveform can berecovered by compensating the dispersion with the parameters got in the DDTprocessing. The recovered signal has no noise, and the received signal is filtered inreality. The whole process is called WM-IDDT.(2) The WM-IDDT normal mode separation method is verified with simulated andexperimental data. The proposed method is proven to have better separating ability viacomparisons between processing results with this method and those with the traditionalnormal mode separation methods.(3) The source ranging method based on the interference structure of normal modes isimproved. The normal modes are extracted with the WM-IDDT method, and theinterference structure of two extracted normal modes with certain orders is thenconstructed. With the help of the phase change of the interference structure and a guidesource, the distance of the target source can be determined. The method is verified withsimulated data and experimental data.