The incoming radar signals to a constant false alarm rate (CFAR) processor, having some known cumulative distribution function F(X) or probability density function f(X), are transformed into new signals Y according to the equationY=.sigma.[-1n {1-F.sub.x (X)}].sup.1/Kwhere .sigma. and K are a scale parameter and shape parameter, respectively, and can be arbitrarily set. The threshold to which the processed signal is compared in the CFAR processing is variable in accordance with the selected value of K.