Abstract
This research focuses on estimation of the spectrum and the spectral power density function of non-Gaussian second order autoregressive model AR(2) through performing simulation?s experiments to calculate the values of power spectral density function (PSD) for different sample sizes and various frequencies.The random errors for the used model follow Non-Gaussian distribution (discrete, and continuous), and also the comparison between simulation results is made by using the criteria of Mean Square Error (MSE) and Mean Absolute Percentage Error (MAPE).The main conclusions of the research are: There is no essential differences appeared for the spectrum values, or spectral power density function, and mean square errors and mean absolute percentage errors between discrete and continuous distributions for all frequencies.In case of Cauchy distribution, whenever the sample size increases the value of p(w) and f(w) goes up, while the value of MSE, MAPE decreases.