Title :
Correlation estimators based on simple nonlinear transformations
Author :
Sullivan, Mark C. ; Wegman, Edward J.
Author_Institution :
Eng. Res. Associates, E-Systems Inc., Vienna, VA, USA
fDate :
6/1/1995 12:00:00 AM
Abstract :
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
Keywords :
Gaussian processes; correlation methods; estimation theory; functions; signal processing; transforms; computational cost; correlation estimators; lag products; nonlinear functions; quadruplex transformation; real complex Gaussian processes; simple nonlinear transformations; variance; Application specific integrated circuits; Computational efficiency; Costs; Field programmable gate arrays; Gaussian processes; Performance analysis; Programmable logic arrays; Random variables; Signal processing; Signal processing algorithms;
Journal_Title :
Signal Processing, IEEE Transactions on
