Discrimination Via Phased Derived Range

The Phased Derived Range (PDR) method generates higher resolution object features than current radar processing methods, making available more physics based features for discrimination. These new physics based PDR features must be combined with currently existing features to determine if new information is being brought to the table. This determination is performed with adaptive feature selection using Data Modeling (a method for extracting [O(3n)] multi-variable polynomial solutions to the Kolmogorov-Gabor polynomials describing the relationships between data sets). Data Modeling also provides a method for enabling physics based target discrimination algorithms to be embedded into existing system firmware for real-time in-theater use. To date, these algorithms cannot be embedded into the limited computational resources of the currently fielded and proposed systems because they use features that are signal and data processing intensive. Data Modeling extracts an equivalent transfer function that captures the functionality within the discrimination algorithm that can be embedded in firmware using resources available in current fielded systems and future proposed systems. Transfer functions of this type can also be derived for algorithms such as radar slewing, antenna pointing, and for populating the probability tables of Bayesian Belief networks used in Battle Management Command and Control (BMC2) algorithms.