Greedy Integration Based Multi-Frame Detection Algorithm in Radar Systems

In this paper, we address the problem of detection and tracking of dim targets in radar systems through the use of multi-frame detection (MFD) techniques. We first study the energy expansion problem during multi-frame integration for classical MFD methods, which shows that the extended energy peak envelope poses a great challenge for target detection in nonhomogeneous backgrounds and multi-target cases. Next, a greedy integration based MFD algorithm is proposed, which separates the overall multi-frame joint maximization into two processes of confidence building and extended energy suppression. The proposed algorithm can eliminate the energy expansion intrinsically and achieve better detection performance with a lower implementation complexity. In addition, we extend the proposed algorithm to radar systems by deriving the accurate nonlinear conversion relationships between target states of Cartesian coordinates and echo measurements of polar coordinates. In radar scenarios, the proposed algorithm is capable to make detection adaptively after multi-frame integration through the use of traditional constant false alarm rate (CFAR) procedures. Finally, numerical results and tests with real radar data are presented to demonstrate the effectiveness of the proposed algorithm.