Asymptotic Behavior of Sample Autocorrelation Function and Sample-based Power Spectral Density

Demonstrates ACF and PSD convergence to their true values as a function of data record length used.

The sample autocorrelation function (SACF) is repeatedly calculated using an increasing number of data samples from different realizations of the same random process. Then, the PSD of the process is estimated using the SACF. The SACF is compared to it's true value and the same happens for the PSD. The rms error for both the ACF and PSD is calculated and ploted as a function of the number of samples used for their estimations. One demo is about a MA(2) random process and the other concerns an AR(2) process. This demo requires a powerful computer.