Deceptive jamming template synthesis for SAR based on generative adversarial nets

摘要

High-quality deceptive jamming template generation is critical to high-fidelity deceptive jamming for synthetic aperture radar (SAR). However, SAR imaging results of a target vary greatly with different observation scenarios and performing parameters. To achieve higher-fidelity deceptive jamming performance, the jammer must have the capacity of intelligently generating a deceptive jamming template matched with a practical SAR scenario. In this paper, we propose a deceptive jamming template generative adversarial network (DJTGAN). The DJTGAN consists of a deceptive jamming template generative network and a discriminative network. The generative network combines low-frequency contents and high-frequency details of a target to boost the performance of the deceptive jamming template generation. Meanwhile, the discriminative network adopts the PatchGAN architecture to capture local texture statistics to improve the fidelity of the deceptive jamming template. Several real datasets are utilized to verify the effectiveness and superiority of the proposed DJTGAN. Moreover, the strip SAR and TOPSAR deceptive jamming results utilizing the deceptive jamming templates generated by DJTGAN further validate the effectiveness of the DJTGAN.