Continuous Feature-Aided Intelligent and Dynamic Sensor Tasking

Toyon proposes to develop a feature-aided dynamic sensor tasking (FA-DST) algorithm for high fidelity radars so that the continuous location and identification (ID) of time-critical ground targets can be supported. Our approach is to treat the problem as aclosed-loop feedback system where a fusion-aided continuous ID process works with a FA-DST module to maximize the information contained in a track database. The amount of information is quantified using an information-theoretic definition of the expectedentropy in the track database given a candidate sensor schedule. The domain of sensor tasks involves a range of different radar modes and waveforms so that the feature database and the fusion and tracking module may best be serviced for different values ofthe expected signal-to-interference ratio, resolution, background traffic, vehicle classification and radar cross-section. Toyon will develop a feature-aware radar model within the SLAMEM(TM) simulation with the assistance CAE Soft Incorporated. We willthen implement the FA-DST algorithm to control the new radar model and evaluate its performance. In Phase II, Toyon will improve the algorithm and radar models and will demonstrate the real-time operation of the concept for an Air Force approved scenarioor at a future military exercise like Millennium Challenge 2002. The successful completion of this research will result in the development of a feature-aided sensor manager that supports the improved acquisition, location and identification of groundtargets. This will be useful for many military sensor tasking and fusion and tracking applications in addition to civilian uses such as border patrol, counter drug and counter-terrorist operations, and environmental law enforcement.