Further investigation on time-domain maximum likelihood estimation of chirp signal parameters

The time-domain maximum likelihood estimation of chirp signal parameters is investigated in this study. Three amplitude weighted phase-based estimators and two phase-unwrapping algorithms, that is, phase prediction and unwrapping algorithm (PP-UA) and differenced-phase prediction and unwrapping algorithm (DPP-UA) were proposed. The DPP-UA and PP-UA have their merits and drawbacks. The authors combine the merits of these two methods and propose a new phaseunwrapping algorithm, which outperforms PP-UA and DPP-UA under low signal-to-noise ratio (SNR) conditions. The authors´ further study on the optimal number of initial data sample show that it is SNR-related and can be determined according to the ´3σ´ rule and the Cramer–Rao lower bound.