N.V. Verdenskaya, A.G. Vinogradov, I.A. Ivanova

Propagation of radar signals through Earth-s troposphere aquire an additional phase shift, additional delay, and additional Doppler frequency shift. These distortions are divided into two parts: a regular, defined by a very smooth behavior of dielectric permeability of Earth-s troposphere, and random, dependent on fluctuations of dielectric permeability. In the general case, modeling random distortions assumes modeling the space-time field of random phase that can be considered as the Gaussian random field with structure function of the power type. If only the sequence of the radar measurements of the same object moving along a trajectory is of interest, the task is simplified and reduces to modeling the random process. This process is the statistically homogeneous Gaussian process. Its inhomogeneity is related to the different ways that radar signal passes at different moments of time. The base of random phase simulation algorithm is in modeling the stationary Gaussian random process with zero mean and a given correlation function. Since correlation function in this case is not rational, we use the algorithm of random process modeling based on the method of sliding summation. The results of the simulation show that the random phase of the signal received by a radar can exceed one radian, which will lead to a distinct destruction of signal coherence within the synthetic aperture and, consequently, to a noticeable distortions in radio-images.