In side-looking Ground Moving Target Indication in the clutter returns over the range swath of interest [3], (GMTI) radar, the 2-Dimensional (2D) Space Time (azimuth-[4]. This results in elevation-dependent clutter statistics which Doppler) domain can adequately define a clutter spectrum which has a deleterious impact on classical STAP performance. This is accurate for all range gates. However, in applications where the array boresight is not perpendicular to the velocity vector (e.g. problem lsparticularly acute for close-in sensig geometries forward-looking radar), the azimuth-Doppler clutter spectrum due to the elevation diversity that exists across short sensing exhibits a dependence on elevation angle-of-arrival, creating baselines. range-varying (but elevation-dependent) clutter statistics, or nonRejecting interference using STAP typically involves "averstationary clutter. Classical Space Time Adaptive Processing aging" many presumably target-free range samples to estimate (STAP) algorithms suffer substantial performance losses in nonstationary clutter since classical STAP assumes clutter station-clutter statistics, (i.e. spectral shape and location). The resultarity along the range (training) dimension. Planar arrays are ing interference covariance estimate is then used to synthesize inherently able to observe the azimuth-Doppler clutter spectrum an adaptive (via the covariance estimate) space-time (2D) filter as a function of the elevation angle, a capability which linear to remove interference from a particular Rangecell Under Test arrays lack. The incorporation of the planar array's vertical (RUT). Thus, having elevation-dependent (i.e., non-stationary) dimension into the joint azimuth-Doppler (2D) STAP domain . '. . in a has previously resulted in 3D STAP. This paper demonstrates the clutter statistc s in a "blurred" 2D clutter covariance ability of 3D STAP to solve the non-stationary clutter problem estimate which is inaccurate for any one particular range by accounting for the elevation-dependent clutter statistics in gate. This poor estimate of the space-time (2D) interference a 3D covariance matrix. A forward-looking array is used to covariance matrix is correspondingly responsible for poor provide non-stationary clutter, and the performance of 2D and STAP performance. Therefore, methods that compensate for 3D versions of the Adaptive Matched Filter (AMF) and Joint non-stationarity Training Data (TD) are needed to regain Domain Localized (JDL) are used in a close-in sensing paradigm. pormancee The results show a >55 dB improvement in output SINR near performance the clutter null using 3D STAP algorithms in lieu of 2D STAP In the STAP literature, elevation-dependent clutter is ofalgorithms applied to the same (subarrayed) data. ten described as heterogeneous clutter [5], range-dependent clutter [3], [4], or non-stationary clutter [6], [7]. For the
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.