Abstract:
© 2002 IEEE. In work it is offered the resulting probability density of set of signal (interference) parameters to describe through a mixture of private probability densities. For definition of unknown parameters of private probability densities it is offered to use iterative method of maximum likelihood. This method is new; it is essential by best in relation to classical method of maximum likelihood. In the report the proofs of optimality of method and it reducibility in the classical method are reduced. The considered iterative procedure is applied for the identification of classes of one-dimensional stochastic signals.