ABSTRACT
In order to get some experience in simulation and testing of random processes, it is helpful to have a simple tool that will generate random processes with specified properties, such as mean and autocorrelation. Such a tool is provided by difference equation models. One well-known method of obtaining approximate numerical solutions to a differential equation is to replace the unknown derivatives therein by their finite-difference approximations. Under the presumption that the sequence X(n) is stationary, the Yule-Walker equations express the relation between its autocorrelations and the coefficients in the ARMA model. The random sine wave certainly does not look very noisy. The properties of a random sine wave buried in noise are a classical problem in detection theory. Most kinds of flow, such as, fluid flow, current flow, etc. are typically viewed as continua, in the sense that the fluid or charge is modeled by a continuous mass or charge density.
