A new memory based ordered statistic-CFAR processing for coherent detection

Adaptive radar detection in clutter and noise environment is a key issue for maritime coherent detection. This paper proposes a new version of the so-called ordered statistic constant false alarm rate technique which exploits past and current information of the environment to optimize the detection processing. The effectiveness of the memory-based detection concept is proved on synthetic and real data recorded by an X-band experimental radar using high range and Doppler resolutions waveform at low grazing angle. Furthermore, a comparative study on coherent detection is performed between classical detectors and the proposed approach in terms of probability of false alarm and detection.