Using the Blind Two Real Source Separation Algorithm in the Task of Detection a Useful Signal with an Additive white Gaussian Noise

This article discusses the effectiveness of the application of the algorithm for blind separation of two real signals in the problem of detecting a reference signal with an additive white Gaussian noise. As a detection criterion, the ratio of the maxima of the posterior probability densities is used: if the reference signal is in the sample and in its absence. The article contains a graph of the dependence of the detection criterion of the reference signal on the signal to noise ratio. A narrowband signal with constant parameters is used as a reference signal.