Comparison of space-time processing approaches using experimental airborne radar data

Design and performance evaluation of a portable parallel library for space-time adaptive processing

“…When the weight calculation and output vector formation steps are applied to the entire data matrix e, the algorithm is referred to as joint-domain optimum or fully adaptive space-time processing [7], [8], [9]. This method is expensive, and full statistical confidence in the results may require more data than is available.…”

“…DiPietro [8] describes higher-order post-Doppler processing, where after Doppler processing, the computation is broken into smaller problems, the number of which is related to the order of the method. Another possible post-Doppler method, given by Brennan et al [7], and also discussed by Ward and Steinhardt [9], is referred to as PRI-staggered post-Doppler STAP. This method performs FFTs on multiple subsets of the data to obtain the matrices for weight calculation.…”

“…Parameter values were chosen based on examples given by Ward [6], Brennan et al [7], and Di Pietro [8].…”

A study of Reduced-Rank stap

“…3DT-SAP [ 1] and slide DFT cascaded by STAP (SDFT +STAP) [2] are representative examples of this kind which hold an identical degree of freedom(DOF) and share a similar mechanism. Both the two methods perform pre-filtering via Doppler filter banks to each space channel at the first stage where the clutter fully spreading in space-time plane is partitioned into several effective narrow band jammers densely distributing along space axis, which are followed by adaptive processing on several adjacent Doppler filter bank outputs.…”

“…For most practical airborne phased array radar, the dimension of the received space-time data during a coherent processing interval (CPI) is often so large that the implementation of full rank STAP will at least encounter the following three problems: (1) the sreat difficulty in hardware implementation due to the high computation burden; (2) the deficiency in the support of independently identical distribution (lid) reference cells required to estimate the clutter covariance matrix; (3) the high system complexity and cost resulting from direct conversions of the array element's radio frequency (RF) signals to digital I & Q channels which should require large number of receiver channels simultaneously. Therefore, extensive studies of reduced-rank(RR) space-time adaptive processing(STAP) have been carried out with the purpose of avoiding the above obstacles.…”