On averaging burg spectral estimators for segments

The Burg spectral estimator (BSE) exhibits better peak resolution than conventional linear spectral estimators, particularly for short data records. Based on this property, the quality of the BSE is investigated with the available data record segmented and the relevant parameters or functions associated with each segment averaged. Averaging of autoregressive coefficients, reflection coefficients, or spectral density functions is used with the BSE, and the corresponding performances are studied. Approximate expressions for the mean and variance of these modified Burg spectral estimators are derived. The variance of the estimation errors associated with the modified power spectral density estimators is compared to the theoretical Cramer-Rao lower bound. It is observed from the results that averaging of reflection or autoregressive coefficients has almost no effect on bias and variance of the corresponding estimators. Averaging of reflection coefficients is most robust to segmenting, and is therefore recommended for applications using fixed hardware implementations of the Burg algorithm.