Density-Based STAP for Detection of Range-Spread Targets

The detection performance of range-spread targets with the conventional space-time adaptive processing (STAP) may deteriorate due to the target self-nulling effect. To tackle this issue, in this paper, a new STAP detector for range-spread targets based on sample density, referred to as den-STAP, is proposed. In the new den-STAP, Capon spectrums are used to measure the distance between clutter samples and target samples. Then, two features of density and low-density distances are designed as the differentiation features of clutter samples and target samples. Finally, den-STAP selects the target-free samples to estimate the clutter covariance matrix and to reduce the target self-nulling effect. The detection stage is implemented via the adaptive matched filter. Simulation results show that the proposed den-STAP achieves superior detection performance on range-spread targets.