MCARM/STAP Data Analysis. Volume II.

Abstract : The purpose of this report is to assess the quality of the Multi-Channel Airborne Radar Measurements (MCARM) airborne data collected, evaluate several reduced rank Space-Time Adaptive Processing (STAP) algorithms, and provide guidance to researchers. The goal of MCARM is the development of STAP algorithms for enhanced target detection in severe environments through the use of real world multi-channel airborne data. The Rome Laboratory's MCARM program provides real data, monostatic and bistatic, for performance evaluation of existing and future STAP approaches. This report evaluates the performance of several state of the art STAP algorithms using monostatic MCARM data. The algorithm list includes Factored Time Space (FTS), Extended FTS Approach (EFA), Cross Spectral Metric (CSM), Joint Domain Localized (JDL), Principal Components, Parametric Adaptive Matched Filter (PAMP), and others. As expected, adaptive processing results in significant detection performance improvements over conventional beamforming techniques. N single algorithm can be deemed "best" as several factors affect performance. Two primary factors are the sample support requirement for interference estimation and computational complexity. A significant finding from the work is the non-homogeneous nature of the clutter returns. This nature violates the assumption of independent, identically distributed data for interference estimation illustrating the need for Non-Homogeneity Detectors (NHDs). NHDs offer improved detection and false alarm performance through better selection of secondary data.