LAUSR

Sinusoidal frequency-modulated (SFM) signal is widely applied in radar, and wireless communication. Therefore, there is significant importance in studying the estimation of SFM signal parameters. However, traditional estimation algorithms have high computational complexity while the corresponding estimation performance degrades when either modulation index or modulation frequency of SFM signals increases. To address these problems of computational complexity as well as the performance degradation, a two-branch two-stage estimation algorithm is proposed in this paper. In the first stage, received signals are passed through two different branches: one for the modulation index not greater than $\pi /2$, and the other for that larger than $\pi /2$. This two-branch strategy is derived from the spectrum properties of SFM signals with respect to modulation index, where the first branch utilizes the Jacobi-Anger expansion, and the second one employs an iterative estimation algorithm. In the second stage, precisions of the parameter estimates obtained from the two branches are compared, and the estimates with higher precision correspond to the final estimates of SFM signal parameters. Both theoretical analysis, and numerical results show that the proposed algorithm outperforms the current state-of-the-art algorithms, while yielding significantly lower computational complexities.