Improved Chirp Parameter Estimation Using Signal Recovery Method

We consider phase-based estimators for chirp parameters due to their high computational efficiency for implementation. Their performances can approach the Cram´{e}r-Rao lower bound for high signal-to-noise ratio (SNR). One challenge for the phase-based estimators is the performance loss at low SNR because the reliability of phase unwrapping and the accuracy of the phase noise model decrease seriously at low SNR. We here propose a signal recovery method to enhance the SNR for phase-based estimators. The method first uses an adaptive filter to recover the chirp signal from the noisy received signals. Then, the reconstructed chirp signal is used as the input to the traditional phase-based estimators for parameter estimation. The significantly enhanced SNR at the output of the filter brings much improved estimation performance. A robust algorithm is also proposed to achieve good performance over a range of values of chirp parameters of interest. Our simulation results show that significantly improved performance can be obtained by using the signal recovery method for the phase-based estimators.