Parametric adaptive matched filter for airborne radar applications

Statistical analysis of the non-homogeneity detector for STAP applications

“…This fact is illustrated with simulated as well as measured data from the MCARM program. Future work will undertake extensive performance comparisons between covariance based STAP methods such as the AMF and the normalized adaptive matched filter (NAMF) [11] (with NHD pre-processing) and model-based parametric STAP tests such as the parametric adaptive matched filter (PAMF) [23], normalized parametric adaptive matched filter (N-PAMF) [18] and fast adaptive processors [25] in non-homogeneous interference backgrounds. Finally, the normalized GIP is expressed as p,= IIYI1 2 Sii.…”

Space-Time Adaptive Processing (STAP) for Low Sample Support Applications

“…The filter is determined by solving for the coefficient matrices of an appropriate fixed order prediction error filter (PEF) and the corresponding error covariance matrix. Excellent performance had been demonstrated with the number of secondary data range cells much less than the number of available Degrees of Freedom (DOF) [3]. Various implementations of the PAMF (and likewise for the NPAMF) are obtained using specified coefficient estimators.…”

“…-1 is lower-block triangular and can be expressed as [3] [ Several methods had been investigated for estimating the and, of these, the Least Mean Square (LMS) algorithm had been found to be particularly effective [3]. A Kalman Filter implementation is considered that appears to offer the fast run times associated with Kalman Filter recursive algorithms.…”

“…The PAMF is based on a fixed-model-order representation of the upper triangular matrix form of the interference covariance matrix inverse [1]- [3]. The filter is determined by solving for the coefficient matrices of an appropriate fixed order prediction error filter (PEF) and the corresponding error covariance matrix.…”

“…This is because they can be applied in a sliding window architecture whereby the relatively few required secondary data cells are always in the neighborhood of the range cell under test. Variants of the Parametric Adaptive Matched Filter (PAMF) [1]- [3] and of Sub-CPI Smoothing are particularly promising such methods. This report describes a comprehensive effort with the goal of developing and evaluating such methods, as well as introducing new methods.…”