Space-time Adaptive Processing (stap) for Low Sample Support Applications

Airborne radar Space-Time Adaptive Processing (STAP) in a heterogeneous, target-rich environment is addressed. An efficient Kalman Filter implementation of the normalized form of the Parametric Adaptive Matched Filter (NPAMF) is introduced and shown to perform well against a detailed simulation of a site-specific, dense-target environment, Ground Moving Target Indication (GMTI) scenario. The number of secondary data range cells in a Coherent Processing Interval (CPI) required by NPAMF is much smaller than the product of spatial channels and pulses and, thus, NPAMF is attractive for low sample support applications. Other promising methods for low sample support applications are introduced and studied, as well. These include a Generalized Likelihood Ratio Test (GLRT)-based PAMF (ParaGLRT) [shown to perform about as well as the matched filter when used in combination with Multiple Pass Processing (MPP)], Sub-CPI Smoothing and a GLRT variant called Severely Non-homogeneous Interference Processing (SNIP). These methods are shown also to perform much better than conventional STAP methods such as Joint Domain Localized (JDL). An optimized variant of MPP called T-SNIP (the “T” is for “target-rich environment”) is introduced, as well. Beam space variants of the above methods also are evaluated and found to require less processing time than element space counterparts while performing at least as well.