Minimal sample support space–time adaptive processing with fast subspace techniques

This is the authors' abstract. We don't add key points for this paper.

The authors investigate finite data support for subspace or projection methods for STAP which are robust against strong clutter returns. A theoretical analysis of the eigenvector projection technique is presented that provides insight into the problem of determining the optimum choice of the projected clutter subspace and matched filter adjustments (with respect to target Doppler frequency). An estimator of the optimum subspace dimension, which is significantly smaller than clutter rank, as a function of the number of samples is presented. This result, combined with recently proposed near-optimal eigenvector-free projection techniques with minimal sample support, reduce the computational burden so drastically that even fully adaptive optimum STAP with large degrees of freedom may become practical for real-time applications.