Adaptive Target Detection in Nonhomogeneous Clutter and Multipath Environment

This article addresses the adaptive radar detection within environments characterized by nonhomogeneous clutter and multipath component. The nonhomogeneous clutter is represented as the compound Gaussian model, combining an inverse Gaussian texture with a complex Gaussian distribution for the speckle component. We further consider the multipath component caused by the diffuse multipath phenomena, modeling by the complex circular Gaussian distribution vector. Two detectors are designed by the two-step maximum a posteriori generalized likelihood ratio test (GLRT) and complex valued Rao test. In the derivation for the proposed detectors, the test statistics are derived first, followed by incorporating the estimated covariance matrix to obtain adaptive detectors. In addition, we present a tunable parameter to adapt to the multipath environment in the GLRT-based detector, which may degrade into the conventional detector derived in the inverse Gaussian texture compound Gaussian distribution. Besides, the proposed detection algorithms are constant false alarm rate tests to the speckle covariance matrix. Numerical experiments using both simulated and measured data confirm that the proposed detectors outperform competitors in multipath environments.