V. Clarkson et al., ANALYSIS OF THE VARIANCE THRESHOLD OF KAY WEIGHTED LINEAR PREDICTOR FREQUENCY ESTIMATOR, IEEE transactions on signal processing, 42(9), 1994, pp. 2370-2379
A theoretical approximation for the variance of Kay's weighted linear
predictor frequency estimator is derived. From this expression, an ine
quality describing the variance threshold of the estimator is found. T
he window weights are then optimized to improve the variance. Numerica
l simulations demonstrate that the variance approximations are valid f
or medium to high signal-to-noise ratios or for large numbers of sampl
es.