A New OS-CFAR Detector Design

For radar systems, target detection procedures need to be conducted with a received signal amplitude that exceeds the CFAR (Constant False Alarm Rate) threshold. An Ordered Statistics CFAR (OS-CFAR), based on a sorting algorithm, is useful for an automotive radar system in a multi-target situation. However, it has a limitation in that it requires a higher computation effort than a typical Cell Averaging CFAR. In this paper, we propose a new OS-CFAR architecture that has a low computational effort. In the proposed approach, a single sorting step is performed as a pre-processing for the decision of the CFAR threshold, and this can reduce the computing time for the entire CFAR process. When the fast sorting algorithm is employed for OS-CFAR, the computing effort is always low compared with normal OS-CFAR architecture regardless of the data size.