NON-LINEAR METHODS IN INFORMATION PROCESSING: MODELLING, NUMERICAL SIMULATION AND APPLICATION

摘要

A method of modelling of non-Gaussian random processes is suggested. The approach is based on representing a process under investigation as a response of a non-linear systems, excited by a white Gaussian noise or a fractional Brownian motion. Systems with random structure are used to model processes with are stationary only locally. Fractional stochastic differential equations are used to model processes with long term dependence.