Phase noise in coherent systems: basic model for simulations

For modelling filtered phase noise more efficiently, the number of numerical samples can be reduced by the method of Foschini et al., making use of the Radon-Nikodym theorem. In contrast to this abstract formalism, the author reviews the underlying principles using only very basic techniques of probability theory. As a consequence, a transparent, more intuitive understanding is developed, which allows a simple error estimation. The deduced formulas depend only on the modelling of phase noise as Brownian motion, but not on the details of the filtering process.