CFAR detector for background noise with two-parameter distribution

We present a double-parameter CFAR with very reasonable losses and low computational complexity. Its basic architecture has been conceived from tail extrapolation theory. The detector uses a detection threshold, set from the measured PFA which is obtained with an auxiliary threshold (pseudothreshold), lower than the final detection threshold. Starting from the basic scheme, a CFAR detector for Weibull clutter has been designed; both, the pseudo threshold control mechanism and the correction of the basic extrapolation equation are described.