Detection of Gaussian Processes Observed Through Memoryless Nonlinearities

This work deals with the detection of nonGaussian processes which can be modelled as the nonlinear outputs of linear dynamical Systems driven by White Gaussian noise. Our approach is based on the idea to find some auxiliary process theta_t and a nonlinear transformation F(t, theta_t, Y_y) = X_t so that X_t to be Gaussian.