Correlation estimators based on simple nonlinear transformations
M.C. Sullivan, Edward J. Wegman · IEEE Transactions on Signal Processing · 1995
The computational cost of estimating correlations may be reduced by employing sums of simple nonlinear functions of the data. A quadruplex transformation is presented and the performance of the associated estimator is analyzed for real and complex Gaussian processes. With independent observations, the variance of the estimator is approximately 14% higher than that obtained by averaging lag products.>