SEASCAPE: System for Efficient Agile Signal Classification of Adversary Produced Emitters Navy SBIR FY2012.2 Sol No.: Navy SBIR FY2012.2 Topic No.: N122-140 Topic Title: SEASCAPE: System for Efficient Agile Signal Classification of Adversary Produced Emitters Proposal No.: N122-140-0097 Firm: Charles River Analytics Inc. 625 Mount Auburn Street Cambridge, Massachusetts 02138-4555 Contact: Daniel Stouch Phone: (617) 491-3474 Web Site: www.cra.com Abstract: Tactical electromagnetic environments have become much more complex over the last two decades with a higher density of more diverse and more sophisticated emitters. Signals of interest are becoming much more frequency agile and have a lower probability of intercept, making it much harder to detect and identify them. To support the Navy in maintaining a combat advantage in dense signal environments, we propose a System for Efficient Agile Signal Classification of Adversary Produced Emitters, or SEASCAPE. SEASCAPE provides an energy-efficient, dynamic, hierarchical ensemble classifier that accurately fingerprints sophisticated emitters while characterizing detection uncertainty and reliability. SEASCAPE is supported by a signal processing workbench and a fusion module to evolve the classifier and its supporting feature sets over time as new agile signals of interest are encountered. Benefits: The SEASCAPE technology will initially be focused on increasing the speed and effectiveness of existing electronic warfare and signals intelligence capabilities for sensor processing military applications. SEASCAPE technologies will be useful in any domain or application that requires adaptive classification of high-dimensional data, such as sensor processing for computer vision and video surveillance, speech recognition, and even laboratory analysis of DNA microarray data. We will also transition specific classification and data fusion algorithms from SEASCAPE to our VisionKitT software suite to increase its commercial viability. Return