Application of the significant-sample method to the study of signal-detection systems on the basis of simulation

A significant-sample method, general for a class of probabilistic input excitations, is developed for the estimation of the detection probability of signals on the basis of simulation using only a posteriori information about the system response to a given input excitation. The application of this method is illustrated on the example of the estimation of false-alarm probability in the case of the excitation of a two-alternative detection system by a random vector whose components are uncorrelated and distributed according to the Rayleigh law.