Constraint optimization for partially adaptive subspace STAP algorithms

Space-time adaptive processing (STAP) algorithms have been widely studied for clutter nulling in airborne radar systems. Many STAP algorithms exploit subspace processing to reduce the complexity of the problem. The algorithm performance obtained by conventional projections of the constraint vector into the appropriate subspace can be improved upon by optimizing the constraint vector projection. This is demonstrated on pre-Doppler (or adapt-then-filter) algorithms, which adaptively process small subsets of radar pulses and non-adaptively Doppler filtering the results. This paper presents a new pre-Doppler algorithm which offers significant performance improvements by allowing adaptivity within the constraint vector to replace the non-adaptive Doppler processing used in conventional pre-Doppler approaches. The new approach provides improves SINR loss and provides narrower clutter notches.