A novel frequency estimator and its comparative performances for short record lengths

Based on the linear prediction property of sinusoidal signals, a closed form unbiased frequency estimator for a real sinusoid in white noise is devised. The frequency estimate is determined from minimization of a modified least squares cost function. Computer simulations are included to evaluate the performance of the proposed estimator by comparing with existing closed form frequency estimators and the Cramér-Rao lower bound, particularly for short observation intervals.