Parameter estimation of a sinusoidal FM signal using CAF-MR algorithm

In this paper, we propose an original strategy for estimating sinusoidal frequency-modulated (FM) signal. This paper derives the cyclic autocorrelation function of sinusoidal FM signal using its character of cyclostationary, and estimates the parameters of modulation frequency, modulation index and amplitude. Compared with conventional time-frequency analysis methods, the proposed parameter estimation method based on cyclic autocorrelation can suppress stationary noise more effectively and perform well when the sinusoidal FM signal has a low modulation index. Simulation confirms the validity and high accuracy of the proposed algorithm when the SNR is low.