Optimal Sensor Scheduling for Ballistic Missile Defense

Our Phase I research involved the development of an innovative sensor scheduling algorithm in support of Ballistic Missile Defense. This algorithm schedules multiple diverse and distributed sensors to perform time-critical tasks relating to multiple moving targets; specifically, to detect, classify, localize, track, support the interception of, and conduct BDA for ballistic missiles throughout their trajectories. Our technique generalizes Brown’s algorithm, which allocates search effort in order to maximize the probability of task success. The first iteration uses a myopic approach to optimally schedule the sensors at each successive time step; subsequent iterations incorporate both future and past information to account for long-term conditions. Each target is modeled by a collection of “sub-targets” (or “particles”), which represent potential target paths. During Phase I, we developed the algorithm’s operating framework, including models for Phased Array Radar and Infrared-like sensor classes. Then, our Matlab demonstration verified algorithm computational tractability and showed that the long-term schedules utilize sensor resources more efficiently, even given the short timeframes involved. Furthermore, when sensors of differing abilities are available, the algorithm appropriately leverages each sensor’s strengths. In Phase II we will develop a more detailed full-scale prototype in a low-level programming language and test it on realistic scenarios.