Parameter estimation of LFM signal via compressive sensing

A novel approach for parameter estimation of LFM signal via compressive sensing is proposed. Considering that LFM signal is sparse in time-frequency domain, compressive sensing technique is used to reconstruct short time Fourier transform of LFM signal. In the reconstruction process, practical sampling rate is considerably lower than Nyquist rate and hence the implementation complexity can be reduced significantly. The Radon transform is then used to search for line in time-frequency plane and estimate parameters. Simulation results show that this approach is capable of estimating parameters of LFM signal using limited numbers of measurements acquiring at sub-Nyquist rate and such a sampling saving at high Nyquist rates (several GHz) would be immense.