Wide-band Physics-based Radar Sensor Model for Signature Phenomenology Prediction

The Air Force Research Laboratory (AFRL) desires a method to map Automatic Target Recognition (ATR) performance to signature performance in order to aid database optimization.  To better understand the physical mechanisms relating the sensing platform, candidate ATR algorithms, and the classifier database, advanced radar signal processing models are required.  Analytic Designs, Incorporated (ADI) proposes a ``Wide-band Physics-based Radar Sensor Model for Signature Phenomenology Prediction'''''''' to model the radar sensor to target signature relationship.  The technical goals of this Phase I program are as follows: (1) development of a wide-band physics-based radar sensor model for signature phenomenology prediction, and (2) validation of simulated results against suitable measured data.  ADI proposes to leverage initial wide-band radar modeling capabilities developed under a previous Air Force program to provide a high-fidelity system level model for a Linear Frequency Modulated (LFM) waveform which includes sensor non-linearities and differential velocity effects over an extended target with full six degree-of-freedom (6DOF) motion.  Measured data from the Automatic Target Radar Identification (ARTI) program will be used for validation.  Development of these physics-based models may lend insight not only into the optimization of the database, but perhaps classifier and sensor design as well. BENEFIT: Commercialization of this modeling effort follows naturally as several of the major air framers including Northrop Grumann and Lockheed-Martin are concerned with database optimization for CID.  The lack of measured data on non-cooperative targets makes simulation crucial to the comprehensive testing required to field ATR algorithms.  Although the large air framers likely have in-house component level modeling capabilities, it is unlikely that they have explored simulations in terms of database portability, and may therefore benefit greatly from our approach to system level modeling.