Opportunistic Sensor Resource Management for Extended Operating Conditions

"Toyon Research Corporation proposes to develop algorithms for creating sensor tasks and assigning them to appropriate sensors based on a priori data, information gathered during surveillance, and assumptions about future events. Our approach will entailconstructing an objective function using an information-theoretic definition of entropy. Candidate sensor tasks will be evaluated by their ability to reduce the entropy of the track database. The selection of sensor schedules is equivalent to finding theminimum of the objective function using optimization techniques. For the Phase I effort we plan to adapt the approach that we have developed in previous work to extended operating conditions such as targets that alternate between "hidden" and "open"states. Toyon will implement the algorithm in software and integrate it in our software testbed, SLAMEMT. The benefit of the algorithm over na¿ve approaches to sensor tasking will be demonstrated by running the simulation on an example militarysurveillance mission. The analysis of the results will suggest features of the algorithm which could be improved in future research. If we are selected to continue development of the concept in Phase II, we would work to refine our algorithm as well astest against more sophisticated surveillance missions. Development of algorithms for the intelligent management of multiple sensors will result in considerably more useful information to warfighters in dev