A Novel Adaptive Radar Target Detector for Non-Gaussian Clutter

Tao Jian,You He,Feng Su,Changwen Qu,Xinfeng Gu

2010-01-01

Abstract:In the clutter modeled as spherically invariant random vector (SIRV), the adaptive coherent radar target detection is addressed. The SIRVs are assumed to possess the same covariance matrix and possibly correlated texture components. A novel covariance matrix estimator is proposed, and then its corresponding adaptive normalized matched filter (ANMF) is obtained. Theoretical analysis shows that, when there exists a match between the estimated clutter group size and the actual one, the proposed ANMF guarantees a constant false alarm rate (CFAR) property to the clutter covariance matrix and the clutter power levels. Moreover, the performance assessments conducted by Monte Carlo simulation show that the proposed ANMF is almost CFAR for mismatch cases. Compared with the existing ANMF, the proposed ANMF performs better with less computational burden. Furthermore, it has an acceptable loss with respect to its nonadaptive normalized matched filter (NMF).