Comparison of the coherent and the component methods for estimating the characteristics of the periodically correlated random processes

The coherent and the component methods for estimating the periodically correlated random process probabilistic characteristics are considered, and theorems on asymptotically unbiased and consistent estimates of the mean, the correlation function and their Fourier coefficients are proved. The asymptotic formulae for the bias and the variance of these estimates are investigated and the comparison results for both methods are provided.