Poly-Gaussian receivers of arbitrarily fluctuating signals

Formulas are presented for calculating the noise immunity of an optimal receiver constructed by approximating arbitrary multidimensional probability density functions with poly-Gaussian density functions. As an example, consideration is given to a polycorrelation receiver for the identification of random signals and a poly-Rice receiver for the identification of quasi-determinate signals. The theorem concerning the possibility of poly-Gaussian density functions is proven.