Bistatic SAR/InSAR Echo Simulation Based On Electromagnetic Scattering Model And Engineering Implementation

As an active Earth observation technology,Synthetic Aperture Radar(SAR)offers the advantage of all-weather,all-day operation.The SAR imaging process is essentially a process of reconstructing the reflectivity function of the target point,which is determined by parameters such as the material of the target,the slope and the observation geometry.SAR images can reflect the rich textural and physical characteristics of targets,making them widely used in military and civil applications such as high-precision terrain mapping,marine monitoring,forestry resource management,precision guidance,target identification,and military reconnaissance.Since the 21st century,more and more new institutional SAR systems have been proposed and SAR systems are evolving towards high resolution,large bandwidth and multi-mode.For different modes of operation,the echo acquisition method and the characteristics of the echo signal vary greatly.Different working modes need to design different systems which results in a complex engineering verification process that requires a lot of money and time.Echo simulation software for unified multi-mode processing is developed to significantly reduce the design cycle of data processing systems.This paper investigates the key problems in multi-mode satellite-based bistatic SAR/InSAR echo simulation,aiming to improve the efficiency of echo simulation and the accuracy of simulation results and improve the processing flow of multi-mode bistatic SAR/InSAR echo simulation.The main work of the thesis is as follows.(1)Research on distributed target refinement modelingIn Chapter 2,a modeling method for terrain and ground environment information is proposed for the distributed target refinement modeling problem.First,an R-D localization algorithm based on an auxiliary DEM achieves real-time localization of the target area with the aid of a publicly available DEM library.We use the DEM as echo simulation scenarios.Second,given that the publicly available DEM grid spacing does not meet the echo simulation accuracy requirements.This section proposes a geometric terrain modelling algorithm based on adaptive fractal parameter estimation,which achieves refined terrain interpolation for coarse DEM.Finally,since the interpolated results only have the slope information of the terrain but not the ground environment information,a refined modelling of common ground environments such as bare soil,asphalt and concrete is achieved based on the Monte-Carlo method.This work provides a valid input to the observation scene for subsequent echo simulations.(2)Research on the electromagnetic scattering characteristic of the Distributed Target The SAR imaging process is essentially a reconstruction of the reflectivity function at the target point,and the Bi SAR scattering coefficient is an important parameter of the reflectivity function.First,the conditions of applicability and the scope of application of the Kirchhoff approximation(KA)model,the advanced integral equation approximation model(AIEM),are explored.The results show that the KA model is only suitable for incidence angles less than 60°,while the AIEM model is more suitable for incidence angles less than 80°.Then,the shadow shading function is introduced into the AIEM model.The validity of the AIEM model for the calculation of the two-station scattering coefficient is verified by comparing it with the measured data.At the same time,sensitivity analysis is carried out on parameters such as observed geometric angle,root mean square height,correlation length,dielectric constant and frequency.The results show that the Bi SAR scattering coefficients are very sensitive to the root-mean-square height,the correlation length and the random rough surface power spectral function.The selection of valid simulation parameters is crucial for subsequent experiments.(3)Bi SAR echo simulation technique based on electromagnetic scattering model In Chapter 4,a Bi SAR echo simulation technique based on the electromagnetic scattering model is proposed in response to the few studies on distributed target Bistatic InSAR echo simulation at domestic and international.First,a concrete form of a frequency domain echo simulation model based on the definition of the reflectivity function is given by defining it in the form of a two-station scattering coefficient modulated by coherent speckle noise.It also combines the platform’s motion to derive an accurate bistatic delay for the propagation of the dual-base signal.It breaks away from the limitations of the application of the traditional "go-stop" assumption for deriving echo forms.Then,a fast shadow determination algorithm based on observed geometric projection is proposed for the problem of inefficient shadow determination of distributed targets.It solves the problem that traditional ray tracing algorithms are too computationally intensive to meet the demands of fast simulations.The SAR images simulated using the methods in this section can effectively reflect physical features such as overlay masking,shadows,and coherent speckle noise in the images.(4)Engineering implementation and multi-mode echo simulation verificationThe bistatic SAR/InSAR echo simulation technology is engineered in C++ using multi-threaded parallelism.The software can simulate echoes from multi-platform,multi-pass,multi-mode,fully polarised and complex background targets.Firstly,the DEM of the area of the accuracy detection field in Hebei Chicheng obtained by TH-2 and the DEM of the SRTM library were used as input parameters.The orbit parameters corresponding to this region are designed with the help of STK software.The spatial baseline InSAR echoes and multi-channel InSAR echoes obtained from the simulation are processed by the InSAR process,point target imaging accuracy evaluation and DEM product accuracy evaluation.It proves the validity and correctness of the echo simulation algorithm.In this paper,a multi-information fusion-based superpixel detection algorithm is proposed to mask layover and shadow regions during the phase unwrapping process.Finally,the above application examples demonstrate the important application value of research related to echo simulation techniques.