Radio Frequency Tomography for Reduction of Improvised Explosive Device Threats

This proposal focuses on developing innovative algorithms which exploit a constellation of narrowband RF sensors to detect IED placement activities. The algorithms include techniques for (i) detecting, localizing, and analyzing behavior of vehicles, (ii) imaging to determine if an area has been disturbed, and (iii) optimally locating sensors to enable items (i) and (ii). In Phase I, we developed a suite of baseline algorithms. These included an new nonlinear filtering approach to detection and localization and a novel CRB method of optimal sensor placement. The algorithms were illustrated in simulations involving both sensor placement and target tracking. In Phase II, we propose to refine and extend the algorithms from Phase I, develop additional algorithms, and evaluate the system on real data. Refinements include improved modeling and characterization of urban phenomena and modeling enhancements. Additional algorithms include development of kinematic profiling and compressed sensing coherent change detection. Real data will be used to validate the algorithms. The success of the Phase I algorithms warrants continued R&D in Phase II. The expected result of the Phase II effort is a software prototype for automated selection of sensor placement, detection and tracking of vehicles, kinematic profiling, and CCD imaging. BENEFIT: At the completion of this effort, the proposed technologies will provide an algorithm suite for automatic target detection, kinematic classification, high-resolution imaging, and tracking. These algorithms apply in many commercial applications, including border protection, airport security, and even traffic analysis. In particular, DHS, the FAA, and other agencies routinely need to secure a border or perimeter and analyze all traffic passing through. Furthermore, policing applications often require certain hot-spots to be intensely monitored, particularly in situations poorly equipped for traditional EO techniques (e.g., nighttime and poor weather conditions). IAI and MTRI have engaged AKELA, a leading manufacturer of RF hardware both for commercial and military applications, as a potential vendor for the first product from this technology. No funding for AKELA is costed into the Phase II effort. However, discussions and meetings for transition planning are included. The first product would incorporate the algorithms developed here and narrowband hardware for an intrusion monitoring system that could be sold to police departments, the FAA, and commercial entities requiring perimeter control.