Research on RF Signal Reconstruction and Parameter Estimation Algorithm Based on Compressed Sensing
Chunjie Zhang,Dongbin Hao,Shanshuang Li
DOI: https://doi.org/10.1109/piers-fall.2017.8293218
2017-01-01
Abstract:In the field of radar electromagnetic countermeasure (ECM), it's vital to estimate the parameters of enemy radar and communication signals. In this paper, the carrier frequency (CF) and modulation parameters of linear frequency modulation (LFM) of low probability of interception are estimated. Shannon-Nyquist sampling theorem is the theoretical basis for converting analog signals to digital signals. The ever-growing demand for data, as well as advances in RF technology, have promoted the use of instantaneous high-bandwidth signals, for which the rates dictated by the Shannon-Nyquist theorem impose severe challenges both on the acquisition hardware and on the storage. But now emerging compressed sensing (CS) technology breakup the limits of the Shannon-Nyquist sampling theorem. In the theory of CS frame, sampling rate no longer depends on the bandwidth of the signal, but rely on the signal structure and content in the information that can be far less than Shannon-Nyquist sampling of the signal discrete data. In this paper, CS technology to implement the low sampling rate of LFM signals with non-uniform sampling (NUS) are linear measurement on the LFM signal, the reconstructed signal was obtained by Stagewise Orthogonal Matching Pursuit (StOMP) algorithm from low dimensional discrete signal and parameter estimation of the LFM signal carrier frequency and modulation rate though reconstructed signal. The reconstruction of LFM signal waveform can be realized, and the CF and modulation slope can be estimated effectively by MATLAB simulation. Compared with the traditional algorithm for LFM signal parameter estimation, CS technology can improve the signal parameter estimation accuracy under the certain SNR.