Multistage partially adaptive STAP CFAR detection algorithm

A new method of partially adaptive constant false-alarm rate (CFAR) detecti on is introduced. The processor implements a novel sequence of orthogonal s ubspace projections to decompose the Wiener solution in terms of the cross- correlation observed at each stage. The performance is evaluated using the general framework of space-time adaptive processing (STAP) for the cases of both known and unknown covariance. It is demonstrated that this new approa ch to partially adaptive STAP outperforms the more complex eigen-analysis a pproaches using both simulated DARPA Mountain Top data and true pulse-Doppl er radar data collected by the MCARM radar.