Phase noise characterization of oscillators through Ito calculus

Recent phase noise analysis techniques of oscillators mainly rely on solving a stochastic differential equation governing the phase noise process. This equation has been solved in the literature using a number of mathematical tools from probability theory like deriving the Fokker–Planck equation governing the phase noise probability density function. Here, a completely different approach for solving this equation in presence of white noise sources is introduced that is based on the Ito calculus for stochastic differential equations. Time‐domain analytical expressions for the correlation of the noisy variables of the oscillator are derived that in asymptotically large times give the steady‐state stochastic correlations as well as the power spectral densities of the variables. The validity of the new approach is verified by comparing its results against extensive Monte‐Carlo simulations. This approach is applied to an oscillator with a dielectric resonator at 4.127 GHz, and a very good agreement between its results with those of the Monte‐Carlo simulations and the previous approaches is observed. Copyright © 2014 John Wiley & Sons, Ltd.