Context-Based Predictive Track Type Prediction Algorithms

MSI proposes a new approach to develop an algorithm that uses context frames and Bayesian inference to anticipate and predict track types of emerging, potential dynamic targets. Adaptive Identification (AID) will use probabilistic approximation to filter and process information that is arriving from multiple sensors and integrate sensor information according to situation specific track models. The models will generate accurate Positive Identification (PID) assessments based on the information they receive. Using context frame interpretations for PID has the potential to eliminate or greatly reduce delays in the Air Operations Center (AOC) associated with the current PID process because it will parallelize the way PID determines the intent and target type of an emerging target allowing more time for identifying and prosecuting time sensitive targets. In use, the algorithm we describe will simultaneously gather intelligence on the track report of a potential target, analyzing the intelligence from the multiple sensors, and determine if the target is a valid target. Probabilistic approximation methods operate in linear time and will potentially reduce PID to a handful of minutes. This will make more time available for planning and more strike options, resulting in more Time Sensitive Target opportunities taken.