A multistage STAP CFAR detection technique

Abstract

This paper introduces a new method of partially adaptive CFAR detection. The processor implements a novel sequence of orthogonal subspace projections to decompose the Wiener solution in terms of the cross-correlation observed at each stage. The detection performance is evaluated in the general framework of space-time adaptive processing. It is demonstrated that this new approach to partially adaptive CFAR detection outperforms the more complex eigen-analysis approaches using true pulse-Doppler radar data collected by the multichannel airborne radar measurements (MCARM) radar.