February Fourier Talks 2008

Space-Time Adaptive Processing (STAP) for Heterogeneous Clutter Scenarios

This talk will provide an overview of STAP starting from the early work of Howells, Applebaum and Widrow on adaptive arrays. The sample matrix inversion (SMI) method and its variants will be discussed in some detail from the standpoint of constant false alarm rate (CFAR) and training data support for covariance matrix estimation. Candidate reduced-dimension methods will also be discussed. Problems encountered in covariance estimation on account of heterogeneous training data will be discussed from a phenomenological, systems, and statistical perspective. The resulting impact on STAP algorithm performance will be addressed. Statistical and ad hoc techniques for characterizing heterogeneous training data will be discussed. Intelligent training data selection schemes will be presented and analyzed. The performance of candidate STAP methods employing intelligent training data selection will be presented using simulated and measured data.