| The rapid and reliable detection of sea targets has significant implications in both military and civil applications.In the face of complex and changeable environments,however,besides effective target signals,echo signals received by sea surface radars include various types of interference,noise,sea clutter,etc.Among them,the non-Gaussian and non-stationary characteristics of sea clutter are obvious,and the high power level makes it one of the most important reasons for limiting the detection performance.Therefore,taking the target detection as the fundamental starting point,sea clutter model of dynamic sea surface could be established based on the special spectral characteristics of sea clutter,thereby enriching the environmental background of radar signal simulator,and providing a basis for effective research on sea surface compound target detection.Due to the increasing cost of radar experiments,the real-time and reliability requirements for subsequent data processing become more stringent,either.Traditional radar signal simulators are subject to poor performance in their own processing and single environmental simulation forms.It is difficult to achieve high computational accuracy and real-time processing speed expected by researchers.Consequently,this paper proposes the use of(NURBS)curve-surface modeling method to establish a space-time-varying rough sea surface model accurately,and then uses electromagnetic simulation to calculate electromagnetic scattering coefficient of the time-varying sea surface;After that,the time-varying coarse sea echo and the sea target signal echo are constructed based on the GPU’s fast computing capability;Finallly,relevant signal processing algorithm should be utilized to verify the correctness of echo structure and simulation algorithm.This paper is mainly composed of the following three modules.First of all,for modeling dynamic sea surface,this paper makes use of linear filtering method by adding time variation factors into static sea surface spectrum for the purpose of obtaining dynamic rough sea surface.The classical PM wave spectrum is selected to generate wave model in the form of point clouds by taking advantage of Monte-Carlo method,and the NURBS curve and surface modeling method is used to inversely reconstruct the oversized ocean wave model into the NURBS grid form sea surface.One of the advantages of NURBS modeling is to improve the modeling accuracy of large-size models,and to pave the way for the subsequent calculation of electromagnetic scattering coefficients on the rough sea surface which changing with time.Then,in regard to the method of calculating large-size electromagnetic scattering coefficient,this paper uses three-dimensional full-wave electromagnetic simulation software FEKO to convert the above-mentioned NURBS sea surface model,and uses its various algorithms to calculate the space-time transformed sea surface electromagnetic scattering coefficient.Using physical optics algorithm can save a lot of computing resources and the computation time is greatly reduced,so that this method is more suitable for large context model calculation.This further advances the construction of subsequent composite sea echo and the implementation of related signal processing algorithms.At last,considering the problem about how to construct complex target echo signals,the sea clutter time series can be implemented by quasi-static methods.Since it is difficult for the CPU to satisfy the large scene and the real-time calculation requirements,the echo structure of the targets and space-time-varied sea clutter of the entire composition scene can be yielded based on the multiple parallel sets of GPU,which greatly improve the speed of echo simulation and practical performance of radar simulation systems.In addition,Range-Dopplor and beamforming algorithms are used to verify the correctness of the echo structure. |
