Relevance of on-line noise compensated approaches for STAP based on AR-filter

Space Time Adaptive Processing (STAP) is known as one of the key techniques for radar processing, but its computational cost is high. To reduce it, the so-called STAP filters based on adaptive 2D FIR filters consists in modeling the clutter-plus-noise as a multichannel autoregressive (N-AR) process. In this paper, our contribution is twofold. 1/ the secondary data are assumed to be modeled as a N-AR process disturbed by an additive white noise. This has the advantage of distinguishing the clutter and the thermal noise. 2/ We suggest using on-line methods such as EKF and Sigma Point Kalman Filter (UKF, CDKF) to estimate the model parameters. These methods have the advantage of significantly reducing the number of secondary data to be used.