OPTIMUM DETECTION OF A CLASS OF RANDOM GAUSSIAN PULSED SIGNALS IN THE PRESENCE OF NOISE WITH A PRIORI PARAMETRIC UNCERTAINTY.

Recursion likelihood ratio and filtering algorithms are derived within the framework of the adaptive Bayes approach by using the method of synthesis of optimum receivers of Markov RF pulses. It is assumed that signals and noise belong to the class of periodically nonstationary Gaussian processes that are described by the general Gaussian model with a priori parametric uncertainty.