Study On Raw Signal Simulation Of Synthetic Aperture Radar

Synthetic Aperture Radar (SAR) is a kind of microwave imaging radar on movingplatform, which acquires high resolution in both range and azimuth by pulsecompression technique and synthetic aperture principle. SAR raw signal simulation isa critical study topic for SAR's research and manufacture. The simulator is capable ofthe examination of SAR's general performance, validation of SAR imaging algorithmbased on different models, evaluation of the influence on SAR's imaging by addingnoise and coefficient errors into raw signal. Moreover, it can also help us to explainthe complicated echo mechanism due to the nonlinearity of electromagneticscattering.In this paper, SAR raw signal simulation is described in detail, the main contentscontain: raw signal simulation of natural ground in a large area by Kirchhoff approach,simulation of complex scene in a small region by Finite Difference Time Domain(FDTD) Method, the united simulation by both methods above, design andrealization of radio frequency broad bandwidth noise generator. The mainachievements are following:First, a SAR raw signal simulation method which integrates Kirchhoff approach andFDTD method is presented for the first time. Raw signal simulation of natural groundin a large area by Kirchhoff approach has the advantage of high time efficiency. Onthe contrary, FDTD is capable of calculation of complex Electromagnetic scatteringphenomena such as volume scattering, multiple reflections and so on, but is lowefficient. Comparison between the simulation results by these two methods is given.The simulation results validate the integration method.Second, a SAR raw signal simulation method which is capable of calculation of thedistributed suburban real scene described by sparse DEM (Digital Elevation Model) ispresented. This method features the integration of the fractal interpolation algorithmbased on Brown function and the Kirchhoff approach. The simulation results of realRoma region are given. The results prove that the presented method correctly handledthe shadow effect, the foreshortening phenomenon, the cross-shaped star effect, the correlative speckle phenomenon, the influence of different polarization and differentsurface's coarseness on the final image and so on.Third, based on Gedney's anisotropic lossy PML (Perfect Matched Layer) theory, anovel PML configuration method is presented, which takes the whole computationalspace into uniform consideration without any more treatment to the PML intersectionregion. Moreover through mapping table technology, extra field componentsintroduced by anisotropic PML are stored effectively. So the algorithm's space andtime complexity is greatly reduced. A 3-dimension numerical verifying case shows,the local error brought by the novel 8 layers PML is about -200dB, much smaller thanthe original 8 layers PML's.The last, a novel board bandwidth microwave noise generator, which is composed of adigital algorithm part and an analog modulation part, is presented. This noisegenerator features broad adjustable bandwidth and wide central frequency range. Theexcellent in-band amplitude flatness and broad bandwidth of the digital part areachieved by means of a novel anti-sine square function FIR(Finite Impulse Response)digital filter, which is discussed in detail. Analysis of the algorithms used by thegenerator in terms of their time and frequency domain character is also presented.Based on the analysis the reason why the wide bandwidth and good statisticaldistribution cannot be obtained at the same time, was given.