LAUSR

The inverse synthetic aperture radar (ISAR) imaging for nonuniformly rotating target has always been a challenging task due to the time-varying Doppler parameter, especially in the low signal-to-noise ratio (SNR) environment. In this paper, a novel ISAR imaging algorithm for nonuniformly rotating targets in the low SNR environment based on the parameter estimation approach is presented. First, the received signal in this work in a range bin is modeled as a multicomponent cubic phase signal (CPS) after motion compensation. Two approaches, namely coherently integrated modified cubic phase function (CIMCPF) and coherently integrated modified high-order ambiguous function (CIMHAF), are proposed to, respectively, estimate the second-order and the third-order coefficients in the CPS. Thanks to the coherent integrations developed in both CIMCPF and CIMHAF, they demonstrate excellent low SNR performance. Moreover, to efficiently implement the proposed approach, the nonuniform fast Fourier transform (NUFFT) is utilized in this work. Due to the usage of the NUFFT, the computational cost is reduced, and the search procedure also dispenses with the nonuniformly spaced signal. Finally, CIMCPF and CIMHAF are applied to produce ISAR image for a maneuvering target based on the CPS model. Several numerical examples, analyses of the noise tolerance performance for the proposed approaches, and ISAR imaging results demonstrate the superior performance of the proposed method.