Practical estimation of correlation functions of nonstationary Gaussian processes (Corresp.)

The preponderance of Gaussian processes among random signals encountered in practical situations has provided the motivation for developing a practical method for the estimation of correlation functions of nonstationary Gaussian processes. The practicality of the method lies in the fact that only a single data-record is required to obtain unbiased and consistent estimates of the correlation functions, thus effectively waiving the strict requirements of the ergodic hypothesis.