ASA 128th Meeting - Austin, Texas - 1994 Nov 28 .. Dec 02

4aUW7. Optimum detection of solid spherical elastic objects of uncertain composition in an uncertain reverberation-limited environment.

M. Wazenski

V. Premus

D. Alexandrou

L. W. Nolte

Dept. of Elec. Eng., Duke Univ., Box 90291, Durham, NC 27708-0291

Detection of solid spherical elastic objects is approached via optimal detection and estimation theory in conjunction with acoustic scattering models for both the object and the environment. The objective is the development of full field (amplitude and phase) optimal decision methods for high-frequency active detection of a target in the vicinity of the seafloor. This parametric approach incorporates the inherent uncertainty of both object composition and seafloor geomorphology. The acoustic model used for the spherical object is the well-known modal series solution and is parametrized by object size, density (object and medium), compressional wave speed (object and medium), and shear wave speed. This deterministic model is used to predict the target signature measured at an array of sensors. Acoustic modeling of the seafloor is performed by applying the 3-D Helmholtz--Kirchoff formulation to the anisotropic, power-law description of seafloor relief due to Goff and Jordan and is parametrized by correlation length (two directions), lineation direction, rms height, and fractal dimension. This model is used to predict the spatial and temporal coherence of the scattered sound field. Illustrative detection examples are presented along with receiver operator characteristic (ROC) performance. [Work supported by ONR: Ocean Acoustics.]