ABSTRACT
The complete characterization of a stochastic process describing variation of an optical signal in time and space requires specification of the joint probability density functions of the field of all orders at all time instances and at all spatial positions. While in some cases, such as calculations involving Gaussian distributions, it can be done, this task becomes unattainable in problems involving random sources and media which are governed by non-Gaussian statistics. Several decades ago the second-order coherence theory for the wide-sense statistically stationary fields has been also formulated in the space-frequency domain. The central quantity of this theory is the cross-spectral density function, defined at two spatial arguments r1 and r2 and at angular frequency. The key importance of the cross-spectral density function follows from the fact that two readily measurable quantities: the spectral density and the spectral degree of coherence of a random field can be readily calculated from it.
