Sparsity-aware STAP algorithms for airborne radar based on conjugate gradient techniques

In this paper, new sparsity-aware space-time adaptive processing (STAP) algorithms based on conjugate gradient (CG) techniques are proposed. The idea of sparsity-aware STAP algorithms is based on the incorporation of a sparse regularization (l₁-norm) constraint to the minimum variance (MV) design criterion. To solve this optimization problem, two different l₁-based algorithms based on the conventional CG and the modified CG are derived. An analysis of the computational complexity shows that the proposed algorithms have nearly the same cost as the conventional algorithms. It is also demonstrated that the proposed STAP algorithms outperform the conventional algorithms using the simulated airborne radar data.