Toyon(R) proposes to develop a system for optimized resource allocation of networked sensors that will dynamically task sensors to support data fusion for improved surveillance, detection, tracking and discrimination of time-critical ballistic targets. We treat the problem as a closed-loop feedback system where a fusion-aided continuous identification (ID) process works with a dynamic sensor tasking (DST) module to maximize the information contained in the track database. The amount of information is quantified using an information-theoretic definition of the expected entropy in the track database given a candidate sensor schedule. The domain of sensor tasks involves a range of different sensor modes/waveforms so that the fusion and tracking module may best be serviced for different values of the expected signal-to-interference ratio, resolution, background clutter, drag coefficient, object classification and radar cross-section. The networked sensors need not be co-located. In Phase I, Toyon will develop radar, EO/IR and Ladar sensor models for our SLAMEM(R) simulation. We will then implement the DST module to control the new sensor models and evaluate the resulting performance. In Phase II, Toyon will improve the algorithm and the fidelity of the sensor models and will demonstrate the operation of the algorithm in real time for a scenario approved by the MDA and will develop hardware-in-the-loop simulation capability.