Frequency Estimation/Tracking for Short Data Windows

In low signal to noise ratio conditions it becomes difficult to estimate the parameters of a time-varying sinusoidal waveform where only a short data record is available at any one time instant. We show how the Discrete Fourier Transform (DFT) may be exploited to obtain a simple frequency estimator, which has a variance in the range one to three times the Cramer-Rao Bound (CRB) for frequency estimation of a sinusoid, even for small data windows. Computer simulations verify the performance of the new estimator and an example demonstrates its utility in tracking the parameters of a modulated waveform in noise.