Estimation of signal-detection characteristics by a simulation method using an input-sampling modification

The paper examines the estimation of the operating characteristics of a bialternative detection system by a simulation method that employs an input-sampling modification and subsequent correction of the system response. A series of detection-curve points is estimated in a single computational experiment using a single sequence of random numbers. The gain as compared to the direct-simulation method is calculated. The practical implementation of the proposed method is demonstrated and numerical results are obtained for the example of the effect of a signal with a random Rayleigh-distributed amplitude in a mixture with uncorrelated noise on the detector.