V. Premus
D. Alexandrou
L. W. Nolte
Dept. of Elec. Eng., Duke Univ., Box 90291, Durham, NC 27708-0291
In a reverberation-limited environment, can imperfect knowledge of the spatial correlation properties of the underlying seafloor geomorphology impact active target detection performance? This work seeks to address this question by examining the dependence of the joint probability density function of the scattered acoustic field at low to moderate incidence angles on the seafloor roughness correlation function. The objective is the development of an optimum decision theoretic approach to the problem of high-frequency active detection of a target suspended in the water column in the vicinity of the seafloor. Acoustic modeling is a geologically meaningful seafloor scattering environment is performed by applying the 3-D Helmholtz--Kirchhoff formulation to the anisotropic, power-law description of seafloor relief due to Goff and Jordan. The generality of the Bayesian approach permits uncertainty with respect to geoacoustic environmental properties to be defined probabilistically and incorporated into the detection algorithm. The method attempts to exploit the spatial coherence of the scattered sound field in constructing the joint pdf of the scattered pressure field measured at an N element array of sensors. Upper bounds on detection performance in terms of ROC curves are obtained as a function of environmental uncertainty. [Work supported by ONR, Ocean Acoustics.]