DFT-based multiple frequency estimation of real sinusoids by analytic signal generating

Frequency estimation of multiple real sinusoids is widely concerned in a wide range of areas. In this paper, a DFT-based multiple-frequency estimation method for real signal is proposed. It contains a coarse estimation stage and a fine estimation stage. This method, which is performed in the frequency domain, focuses on generating the corresponding analytic signal of each component and reducing the spectrum leakage of other sinusoids to obtain an accurate estimation of each frequency. The proposed method is efficient and has a lower calculation cost than other recently proposed frequency domain methods for real sinusoids. Simulation results show that the proposed method is capable of achieving the Cramer-Rao lower bound under a wide range of signal-to-noise ratios.