An efficient architecture for nonhomogeneity detection in space-time adaptive processing airborne early warning radar

Emerging sensor technology has spawned renewed interest in the deployment of space-time adaptive processing (STAP) based airborne early warning (AEW) radar. Numerous theoretical studies and simulation analyses have clearly identified the potential for improved detection performance over classically designed AEW sensors, especially for weak target returns masked by sidelobe clutter and jamming. However, many of the underlying assumptions about the nature of the interference competing with and masking weak target returns are unrealistic. Most notable among these is the requirement for independent and identically distributed (i.i.d.) training data for covariance matrix estimation. The analysis of experimental data collected under the Rome Laboratory Multi-Channel Airborne Radar Measurements (MCARM) program clearly indicates that the nonhomogeneous nature of the clutter environment often results in severely degraded detection performance. The authors have developed a statistical signal processing technique designed to address this fundamental limitation to STAP, which is referred to in the literature as the nonhomogeneity detector (NHD). This algorithm is designed to be compatible with existing STAP techniques, and is used to screen training data for the excision of outliers prior to covariance matrix estimation. Published results indicate a significant improvement (in excess of several dB) in the detection performance for weak targets when NHD is applied in conjunction with STAP. The NHD performance and an efficient STAP architecture with embedded NHD are presented.