An Iterative Frequency Estimation Algorithm Using Generalized Fourier Interpolation

Motivated by the iterative estimation approach basedon Fourier interpolation in a recent literature, this paper proposes ageneralized interpolation on Fourier coefficients and an iterativefrequency estimation algorithm based on the generalized interpolation.Both theoretical analysis and simulation tests show that the iterativegeneralized Fourier interpolation algorithm converges in two iterationswith the estimation variance only marginally above the AsymptoticalCramer-Rao bound (ACRB) over the entire frequency estimation range.Moreover, the proposed algorithm allows setting different values forits initial parameter. An approach on how to choose the initialparameter is also presented. By using a suggested initial value, theproposed algorithm is much more efficient than the original algorithmin the literature while maintains totally the same estimation accuracy.