SBIR/STTR Award attributes
Present-day Surveillance Towed Array Sensor System (SURTASS) ships perform detection and classification operations, including the computationally expensive task of processing collected data. Since greater computing is available at on-shore processing stations, improved target detection and localization are possible if the terrestrial stations perform the processing. However, the large volume of data requires large communications bandwidth between the ship and shore. If this data volume can be reduced, then the shore facilities can take on increased processing, reducing the need for costly custom hardware on the ship.Metron proposes an innovative, physics-based data compression approach that transforms hydrophone data into a related mathematical matrix space that possesses structure related to ambient noise and any acoustic sources, allowing improved communications. The approach allocates bits to represent the transformed signals more efficiently than traditional linear methods and preserves nearly full-fidelity data representation with fewer bits. By reducing the number of bits needed to represent the data, the proposed bit-allocation approach will in turn reduce the number of bits needed to transmit the signals from SURTASS vessels to shore for processing and analysis. Preliminary testing on real, unclassified data already achieves significant compression, and thus provides a promising foundation for continued research.
