A Novel Non-Uniform Rotational Motion Estimation and Compensation Method for Maneuvering Targets ISAR Imaging Utilizing Particle Swarm Optimization

Non-uniform rotational motion of maneuvering targets can result in nonlinear migration through range cells (MTRC) and high order phase term, making final inverse synthetic aperture radar image severely defocused. Therefore, a novel non-uniform rotational motion estimation and compensation method is proposed to achieve the elimination of high order phase term and nonlinear MTRC. Under the constant acceleration model, the estimation of rotational motion parameters is first modeled as an unconstrained optimization problem with the image entropy as the cost function. Particle swarm optimization is utilized to search the global optimal solution making use of its simple structure and high efficiency. Then, resampling is performed to transform the non-uniform rotational motion into linear movement and then residual high order phase term is eliminated by constructing the corresponding compensation term. On this condition, first-order Keystone transform can be applied to correct the MTRC. Focused image can finally be obtained by performing traditional Fourier transform along the cross-range dimension. Experimental results on simulated data, electromagnetic data, and real data are presented to verify the effectiveness and robustness of our proposed algorithm.