A unified model for non-stationary and/or non-Gaussian random processes

An analytical model for non-stationary and/or non-Gaussian random processes described in the paper is based on a normal stationary random process. The non-stationarity is introduced as a deterministic dependence of the parameters of the marginal distribution function or those of the correlation function upon the argument t. Consideration that the mentioned parameters are random variables or stationary random processes results in generating non-Gaussian distributions of the unconditioned process. By combining deterministic and random components of the parameters' dependencies, non-stationary and simultaneously non-Gaussian random processes may be easily specified. The model described may be useful for analytical treatment, for identification of experimentally obtained realizations of random processes and for simulation of random processes on computers as well as in the laboratory.