Cramer-Rao analysis of reduced order STAP processors

Abstract

The Cramer-Rao bounds (CRB) of azimuth and velocity estimates of suboptimum space-time adaptive (STAP) processors are presented. Several spatial and space-time order reducing transforms are analysed and compared with the optimum (ML) processor. Although some of the processors discussed exhibit near-optimum performance in terms of signal-to-clutter ratio some of them produce high standard deviations of target velocity and azimuth. Since estimation errors are a major input of subsequent tracking algorithms the selection and design of an appropriate processor architecture is an important task.