Detection of a Well-Characterized Signal in Persistent Clutter

In scenarios involving persistent signal and clutter, such as occur in passive radar applications in urban environments containing fixed commercial transmitters, subspaces of clutter and signals of interest may be well characterized. With signal and clutter subspaces known, but not necessarily linearly independent, standard matched-subspace detectors based on generalized likelihood ratio test project measurement data into the orthogonal complement of the clutter subspace, thereby ignoring any signal energy that falls into the clutter subspace. This paper proposes a model on the distribution of the clutter subspace that enables signal energy in the clutter subspace to be exploited by the detector. Therefore, a combined detector making use of full signal subspace is proposed, which improves the detection performance with sensible weight formula. Its effectiveness is verified by the receiving operating characterized curves.