Hee Chun Song
Phys. Oceanogr. Div., KORDI, Ansan P. O. Box 29, Seoul, 425-600, Korea
Inha Univ., Inchon, 402-751, Korea
In a recent paper, matched-field processing (MFP) for localizing a point acoustic source using a vertical array in an oceanic waveguide was extended to treat a moving source problem in the presence of both spatially and temporally white Gaussian noise process. The spatially white noise assumption, however, is not realistic for MFP because noise as well as signal propagates in the ocean obeying the governing wave equation, such as the noise field generated by the action of wind. In this paper an optimum processor is developed for passive localization of a moving source in the presence of correlated noise using normal mode description of the sound field. The approach involves a preliminary processing to transform the problem into a white Gaussian noise problem, followed by applying the white Gaussian noise solution.