In this paper the Generalized maximum entropy method and its application to the reconstruction of two-dimensional complex functions (namely, far-field antenna pattern distribution) from nonuniformly distributed samples are considered.
            Two approaches to solving the reconstruction problem are proposed. The first one assumes the reconstruction in pattern domain, and the second - the reconstruction in a dual aperture domain.
            Theoretical aspects described are verified by the numerical simulation technique.
            Simulation of these approaches is carried out without and with measurement eitors in amplitude, phase and sample positions. It is shown that the Generalized maximum entropy method is sufficiently stable in the presence of errors.
            Thus, it is show that the Generalized maximum entropy method suitable for the reconstruction of function of any type seems to be effective for solving an interpolation problem, in particular, in antenna technology for the reconstruction of complex antenna pattern distribution from nonuniformly distributed samples.