STOCHASTIC MATCHED FIELD PROCESSING USING DIRECTED RIEMANNIAN DISTANCE

Abstract

The Matched Field Processing (MFP) plays a crucial role in modern passive SONAR system since its effectiveness of underwater source localization. Their applications are but not limited to floating boat detecting, submarine detecting, fish finding as well as ocean environmental parameters determining. In the past, the stochastic matched field processing (SMFP) are derived on the basis of Riemannian distances (RDs) which were calculated using isometric mappings (IMs). In this paper, a new STMP is provided using directed RDs which were obtained by solving the geodesic equations directly instead of using IMs. In addition, we exploit the symmetric property of Riemannian manifold to solve the geodesic equation in the simple manner. The performance of the proposed STMP outperformed to that of standard algorithm at the expense of a little more of computation.