Radar Image Generation Using Multiple Compensation Decoupling Control Scheme and FJND-constrained Dual Truncation Compression

The radar image simulation of artificial targets using the existing image integration formula exhibits spectrum errors in comparison with bistatic integration results. These errors hinder the simulation results from having the accurate frequency dependence and azimuth anisotropy. In this manuscript, we propose a novel radar image generation framework based on the image integration formula. The proposed method enables two-dimensional radar image generation at arbitrary look angles with an acceptable error level. The globe turntable model, able to describe imaging planes, is first established. It provides a mathematical model for the proposed image generation. Subsequently, we derive the image primitive as the fundamental image contribution unit in radar image generation. Then, a multiple compensation decoupling control scheme is proposed to calculate the radar complex image for consistency with bistatic integration. The frequency domain just noticeable distortion (FJND) constrained dual truncation compression method guides the truncations of the image primitive and spectrum. This operation maintains precision while reducing the computational load. The effectiveness and accuracy of the proposed method are demonstrated by simulation results. Numerical validations demonstrate that the proposed method achieves errors in the spectrum below the defined 5% FJND threshold, with a mean error of 0.45 dB.