Autonomous Classification of Acoustic Signals

The goal of this SBIR is to develop an advanced autonomous acoustic classification capability, known as Normal Mode Decomposition (NMD), for use with two acoustic sensors under consideration by PMS-485 for the Shallow Water Surveillance System (SWSS): the Planar Array Prototype (PAP) and the Advanced Deployable System (ADS). The NMD algorithm, originally developed by OASIS under ONR 321-US funding, differs from traditional intel-based classification methods by exploiting acoustic propagation physics and fundamental differences in the way surfaced and submerged sources excite the shallow water waveguide. Under the Phase II effort, the NMD algorithm will be ported to C language and integrated into a prototype SWSS embedded software baseline. Data analysis will be performed in order to quantify NMD performance using existing data sets from the ADS and/or PAP arrays in a suitable, downward-refracting, shallow water environment. Under the Phase II Option, if exercised, a real-time demonstration of the technology will be conducted. Post-analysis of experiment data will be performed to support a formal quantification of baseline performance. The effort will culminate with the delivery of a Phase III development plan that outlines the path for transition of the technology to U. S. Navy use.