The purpose of this Phase I of fort is to develop an advanced IRST discrimination algorithm by applying a newly developed mathematical technique to tracking filters and thus exploiting the reduced 'tracking error in the predicted position to better reject false detections. 'the new filter concept is founded on penetrating implicit (and presumably heretofore unrecognized) assumptions upon which classical minimum Mean square error type filters are based allowing for the exploitation of known information about target a previously considered unusable. An example Bug potential 80% rejection of false targets in th-4 Op predicted position compared to classical filters. The generalized filter formulation is simple, flexible, and Comprehensive. It can be tailored io alfty target, sensor, and scenario parameters, A bonus is that it obviates the need for a maneuver detector in the case' of a maneuvering target.