Skip to Main Content
To avoid target signal cancellation, which decreases the output signal-to-interference-plus-noise ratio (SINK) of space-time adaptive processing ( STAP ) significantly, it is necessary to eliminate the data vectors contaminated by interference-targets from the training data vectors set. There often exist errors between the actual steering vector and the assumed desired one for interference-targets (outliers) signals in nonhomogeneous clutter environments. The errors can cause distortion of main beam pattern, which leads the conventional reiterative censoring adaptive power residue (RAPR) method to degrade significantly in performance or even completely fail. An enhanced robust effective methodology is presented here, which first suppresses the mainlobe distortion to censor the strong interference-targets by diagonal loading the covariance matrix of training data vectors with a large value; and then eliminates the weak interference-targets coaligned with the desired signal vector. In addition, the new method achieves low computational complexity due to the recursive updating for inverse of covariance matrix. The computer simulation results show that the enhanced method is superior to the conventional RAPR method.