A Glrt Based Stap for the Range Dependent Problem

We consider in this paper a likelihood principle based approach for the range dependent problem in space time adaptive processing. The proposed generalized likelihood ratio test (GLRT) addresses the range dependent issue by directly applying the likelihood principle to the range dependent signal model. Using the knowledge of platform geometry, we develop maximum likelihood estimators that facilitate the GLRT. This differs from existing methods that rely on data transformations in dealing with the range dependence issue. Numerical examples show that the new GLRT approach exhibits significant performance gain over existing approaches.