Recent studies have been carried out, in which the statistical properties of the received signal of free space
communication systems are assessed by means of simulations. The obtained results, however, have not brought
information about the temporal behavior of the generated photocurrent. This work describes how this can be
accomplished through stochastic processes simulations. Although most of the theory on stochastic field simulation has
been developed for application in mechanics and fluids engineering, it is well suited for signal analysis in wireless
optical systems. Basically, if it is possible to artificially create a signal with realistic properties that is transmitted over
the atmospheric channel, it can be used for system analysis and design without the need of mounting a real transceiver.
This dramatically reduces development costs and time. A technique that has been widely applied in engineering is the
spectral representation method. The concept of spectral representation of Gaussian random processes was introduced in
1944. Its use, however, in generating simulations of random processes was only proposed in 1972. This method deals
with the summation of a large number of weighted trigonometric functions. This work shows how to apply this technique
in order to assess the generation of turbulence-corrupted signals using the above mentioned method and its application in simulating data transmission over the atmospheric channel. As figure of merit, system Bit Error Rate and the eye-figure
of the received signal are obtained.