Robust Space-Time Adaptive Processing for Nonhomogeneous Clutter in the Presence of Model Errors

In this paper, we first develop a novel array self-calibration method for estimating array gain-phase errors by computing the clutter subspace from the radar system parameters and using the clutter data in space-time adaptive processing (STAP). The proposed algorithm is shown to perform well even in nonhomogeneous clutter, and it can improve the performance of existing STAP algorithms, such as the clutter subspace-based method in the presence of array gain-phase errors. We also develop a two-stage STAP approach for suppressing nonhomogeneous clutter in the presence of model errors in addition to array gain-phase errors. In our two-stage STAP approach, the first stage explores the clutter subspace calculated from the radar system parameters to suppress the main clutter. The second stage employs the conventional method, such as the partially adaptive sample matrix inversion STAP method, to remove any residual clutter. Numerical results illustrate the benefits of the array self-calibration method and the effectiveness of the two-stage STAP method. Finally, the performance of the three STAP methods is compared via the well-known MCARM data set. The results further confirm that there is an improvement in performance when using array self-calibration together with the two-stage STAP method.