An approach for utilizing known terrain and land feature data in estimation of the clutter covariance matrix

Abstract

The secondary data for estimation of the clutter covariance matrix in space-time adaptive processing (STAP) is normally obtained from range rings in the vicinity of the test range ring. The assumption is that near-by range rings are representative. However, this is not always true. A clutter model is developed and the condition necessary for obtaining a good estimate of the clutter covariance matrix are discussed. A theoretical basis for choosing reference rings which contain clutter patches that are equivalent to one or two patches within the test range ring is provided. A proposal for using a priori map data to classify clutter patches based upon the type of features contained within them is presented. It is conjectured that patches with equivalent classifications will have representative radar returns. An application to post Doppler radar processing is presented.