Scattering Centers Extraction Of Radar Target Based On Spread Spectrum And Compressed Sensing

As the passive detection technique developing, various broadband, high-performance passive detection systems become the threats to radar today. Reducing radar radiation power and improving radar waveforms uncertainty in time domain, spatial domain, frequency domain are effective methods to improve the radar RF (Radio Frequency) stealth performance. Based on the requirement of radar RF stealth, in this paper, target scattering center extraction is researched for low-power radar based on spread spectrum signal, and frequency domain interval uncertainty radiation waveform based on compressed sensing (CS) theory is studied to be applied in radar target scattering centers extraction. First, the development of radar target scattering center extraction is introduced, including the type of radar target scattering center, mathematical model, parameters extraction algorithm, low-interceptive signal and scattered signal calculation method.Secondly, detailed description of the theory and algorithms applied in this paper are presented, including two-dimensional signal parameter estimation algorithms, such as two-dimensional rotational invariance techniques (2D-ESPRIT), two-dimensional matrix enhancement and matrix pencil method (2D-MEMP) and two-dimensional modified matrix enhancement and matrix pencil method (2D-MMEMP), pseudo-random sequence characteristics and construction method, CS theory, and support vector machine (SVM) used for pattern recognition.Then, based on exponential sum target scattering center model, novel one-dimensional and two-dimensional spread spectrum radar target scattering center models are presented respectively. In one-dimensional scattering centers extraction simulation, fast fourier transform (FFT) is used to estimate scattering center parameters. According to the experimental results, when SNR is-10dB, the spread spectrum target scattering waveforms could be effectively used to estimate the target scattering center parameters, but the target scattering waveforms without spread spectrum could not extract any meaningful information. In two-dimensional scattering centers extraction simulation, FFT and 2D-MEMP are adopted to estimate the scattering center parameters, and the experimental conclusion is the same as in one-dimensional simulation.Finally, based on CS theory, novel one-dimensional target range profile reconstruction algorithm is presented. Frequency domain interval uncertainty radiation waveform is designed according to the CS random sampling theorem, convex optimization is used to estimated target scattering center parameters from the uncertainty scattering signal, and the target range profile is reconstructed. The simulation results show that when the target range profile is sparser, the CS-based reconstruction results will be better. At the same time, the sparsity of the range profile changes acutely when the target azimuth changes, which make the reconstruction method's performance changes accordingly. According to the SVM target recognition simulation results, CS-based range profile reconstruction is only suitable to target identify in limited database.