Computationally efficient suboptimum ARMA spectral estimation

There is a great deal of interest in data-adaptive spectral estimation. The general appeal of these methods stems from the fact that their resolution is not determined by a spectral window bandwidth (which depends on the data length) as in the traditional Blackman and Tukey estimators. The method most commonly used, is the autoregressive (AR) or the maximum entropy spectral estimator.