Exploitation Algorithms in the Compressive Sensing (Sparse Measurement) Space

ABSTRACT: In this proposal, the MaXentric team outlines its strategy for creating a Radar-based Compressive Sensing and Target Classification system, code-named Rad-CSC. In this Phase I, the team will implement and optimize compressive sensing and classification algorithms for use with conventional and synthetic aperture radar data. Data will be chosen from publicly available datasets, such as the MSTAR X-band SAR imagery from AFRLs Sensor Data Management System (SDMS). These data will be analyzed and pre-processed, with ground truths generated where necessary in order to identify targets (such as vehicles) and clutter (such as foliage). Compressive sampling methods will be applied in Matlab to simulate the compressive sensing process, with such algorithms being refined in order to optimally emphasize the features of the targets. Compressive classification techniques will then be implemented to detect and identify multiple categories of targets from the reduced-dimensionality data. These methods will then be extended to an existing platform by Dr. Fathy, providing a design for Phase II hardware which will implement compressive sensing & classification methods on radar imaging data. This design will be presented in a Phase II proposal, and the results from the Phase I study will be presented in a final demonstration.; BENEFIT: MaXentric plans to extend Rad-CSC, and its derivatives, to both military and private commercial markets. Primary military applications are the focus of this greater proposal; diverse secondary applications have been identified for examination during the course of SBIR research, including DARPA ISR (Intelligence, Surveillance, and Reconnaissance) & TRACE (Target Recognition and Adaptation in Contested Environments), lightweight target detection, warfighter awareness/targeting in contested environments, and rapid reconnaissance. Additional applications in the non-consumer private sector include fire/rescue C3, man-portable sensors emergency personnel, and low-energy surveillance (passive & active). MaXentric will also seek to explore any other identified commercialization and secondary applications of the technology developed under this STTR contract.