This paper deals with the problem of estimating the autocorrelation coefficients for a stationary Gaussian process with known mean and variance. Two unbiased estimates using random clipping are proposed. The variances of the proposed estimates are numerically compared against those of competitors.
