Parameter Estimation for Multi-component LFM Signals with Alpha-Stable Noise

Cognitive radio (CR) has been widely used in complex electromagnetic environment, and there is a large number of linear frequency modulation (LFM) signals in the environment, they often form multi-component LFM signals which CR will receive, and the multi-component LFM signals are accompanied by non-Gaussian noise. So the parameter estimation of multi-component LFM signals with non-gaussian noise in CR is necessary. In this paper, a neoteric parameter estimation method of multi-component LFM signals based on golden section fractional Fourier transform (GSFRFT) and synchroextracting short-time fractional Fourier transform-Hough (SSFT-Hough) with alpha-stable noise in CR is proposed. First, we use a nonlinear transformation to restrain alpha-stable noise. Second, GSFRFT is used to obtain the optimal orders, which can be used to estimate the frequency modulation rate of multi-component LFM signals and get the number of signals in them. Finally, we use SSFT-Hough to estimate the initial frequency of multi-component LFM signals. In addition, simulation results show that the properties of the proposed method are good in low mixed signal-to-noise ratio (MSNR), and the proposed method is better than existing methods.