Applications of the new generalized form of maximum entropy method to solving inverse problems

A generalized form of the maximum entropy method is discussed. It is suitable for the reconstruction of functions of any type (real nonnegative and real with alternating signs, as well as complex ones). This generalized method seems to be promising for solving inverse problems in many areas of physics and technology. It is shown that the maximum entropy method can be successfully applied to imaging coherent sources, to interpolation of functions from nonuniformly distributed samples, to estimation of measurement errors, and to retrieving mission information using only magnitude or phase incomplete information for minimum phase equivalent signals. The applications considered can be effectively used for digital signal processing in optics, radio astronomy, radio holography, communications, and antenna technology. Simulation results show the high quality of reconstruction provided by this method as well as high stability to errors.<>