You are here

Autonomous Classification of Acoustic Signals


OBJECTIVE: The objective is to develop signal processing computational techniques and methods for autonomous (no man-in-the-loop) classification of passive acoustic signals from fixed, wide-aperture planar arrays. DESCRIPTION: The Navy needs an innovative approach to autonomous classification of passively detected acoustic data. Specifically, new distributed 2-D planar arrays and potentially, autonomous platforms have tremendous potential for persistent monitoring of surface and subsurface acoustic targets. The Navy already has the capability of collecting large volumes of passive acoustic data; however, the data is analyzed weeks after collection. Today, tactical and strategic forces need immediate processing and classification of passive acoustic data. To meet this need, existing and future distributed systems must be able to acquire, process, detect and autonomously classify surface from subsurface acoustic contact data to support contact messaging to others, at a high confidence level. A wide variety of software tools are available for collecting and analyzing marine acoustic data; but, most existing tools require a trained operator to detect, classify, and track surface and subsurface contacts. Prior year efforts, referenced in Refs (1-4,) evidenced potential for processing and autonomous detection from planar arrays, but the state of computational capability and software design resulted in unacceptable detection reliability. The Navy believes a giant leap-forward is possible and potentially supported by the small business innovation base. Some attractive concepts have focused on physics-based source localization, relying upon the integration of acoustic wave propagation modeling with the spatial filter or beam-former (see Ref. 1) to provide an estimate of source range and depth. Such model-based processing must overcome incomplete knowledge of the physical ocean waveguide, which is required to model the signal (see Ref. 2). A number of approaches have been proposed to meet this challenge, including adaptive methods (see Refs. 3 and 4), and methods that exploit robust properties of the waveguide itself (Ref. 5). New approaches would process fixed wide-aperture planar acoustic array data and autonomously detect, localize, classify, and track, contacts. Navy has been unable to identify traditional sources for integrated approaches that function in the acoustic-only domain. This topic seeks an integrated approach that is better aligned with the passive acoustic processing challenge and specifically addresses three key capabilities: (1) autonomous and robust classification of received acoustic signals differentiating surface contacts from subsurface contacts, (2) suppression of loud fast moving surface-ship acoustic interference in realistic ocean environments, and (3) automated in-situ techniques estimating environmental parameters needed for accurate modeling of the contact and interference signals (as needed in items 1 and 2). Although all proposed methods will be evaluated on their merits, note that contact depth is an undeniable classification feature. The Phase I effort will not require access to classified information. If need be, data of the same level of complexity as classified data will be provided to support Phase I work. The Phase II effort will use more robust data sets; Phase III will migrate to classified data. PHASE I: The company will develop innovative concepts for any or all of the three key technical capabilities described in the description, capable of being integrated into a persistent real-time acoustic autonomous processing system. The company will demonstrate the feasibility of its concepts using unclassified data (either modeled or from realistic ocean environments) in a suitable simulation or analysis of its proposed approach. The small business will provide a Phase II development plan for concepts developed in Phase for two distinct applications: (a) ashore/afloat where size and power constraints are relaxed, and (b) a Distributed/Netted System concept of autonomous undersea surveillance systems where size, low power and endurance requirements are a premium. It will address performance goals, key technical milestones and technical risk reduction. PHASE II: Based on the results of Phase I and using the Phase II development plan, the small business will develop a scaled prototype real-time autonomous signal processing architecture implementing the Phase 1 software concept hosting the innovative software on commercial-off-the-shelf hardware The company will demonstrate performance in the lab and compare it with metrics established in the Phase I effort and against synthetic datasets representative of operational applications. Computational complexity will also be addressed in Phase II. This effort may require access to classified information. The company will prepare a Phase III development plan to transition the technology to Navy use. PHASE III: If Phase II is successful, the company will be expected to support the Navy in transitioning the technology for Navy use. The contractor will be provided classified data as required. Specifically the prototype developed in Phase II will be integrated with a suitable surveillance sonar system and used in a proof-of-concept test to determine overall systems effectiveness including detection and classification performance, reliability, and persistence. In parallel, the company will continue to refine its design for evaluation to determine its effectiveness in operationally relevant environment(s). The company will support the Navy for test and validation to certify and qualify the system for Navy use and will support the transition of the developed technology to appropriate Navy systems. PRIVATE SECTOR COMMERCIAL POTENTIAL/DUAL-USE APPLICATIONS: The automatic classification technology will be useful in Department of Homeland Security (DHS) port protection, Drug Enforcement Agency (DEA) drug interdiction and finding fish and mammals in the commercial and university research sectors. REFERENCES: 1. Baggeroer, B.; Kuperman, W.A.; and Mikhalevsky, P.N."An Overview of Matched Field Methods in Ocean Acoustics."IEEE J. Ocean. Eng. 18 1993: pp. 401424. 2. Shang, E.C. and Wang, Y.Y."Environmental Mismatching Effects on Source localization Processing in Mode Space."J. Acoust. Soc. Am. 89 1991: pp. 2285-2290. 3. Krolik, Jeffrey."Matched-field Minimum Variance Beamforming in a Random Ocean Channel."J. Acoust. Soc. Am. Vol. 92, No. 3, September 1992. 4. Cox, Henry and Pitre, Richard."Robust DMR and Multi-rate Adaptive Beamforming."Conference Record of the Thirty-First Asilomar Conference on Signals, Systems & Computers, 2-5 Nov 1997. 5. Zurk, Lisa; Lee, Nigel; and Ward, James."Source motion mitigation for adaptive matched field processing."J. Acoust. Soc. Am., Vol 113, No. 5, May 2003.
US Flag An Official Website of the United States Government