基于信号精细特征的雷达抗欺骗式干扰技术研究

In Nowadays, radar usually works in the complicated electromagnetic environment and the deception jamming is one of the most important jamming. The equipped radars use waveform agility or sidelobe blanking to excise deception jamming. However, with the development of the ECM facilities, existing methods cannot excise the jamming completely. It is of great importance to develop new anti-deception jamming methods. The key to excise the deception jamming is to discover the difference between true targets and deception jamming. In this project, researches and applications of the deception jamming excision method based on jamming subtle features are conducted. First of all, we begin with the slow time feature and analyze the micro-motion feature between true targets and jamming. Secondly, inspired by the specific emitter identification technique, we will analyze the inter-pulse subtle feature and the corresponding excision method. The reasonable combination of these two methods can develop a complete procedure. All of these work using the unintentional modulation of pulse to excise deception jamming, which lay the foundation to develop a sort of new anti-deception jamming methods and can significantly improve the operational performance in the complicated electromagnetic environment.

现代雷达工作于愈发复杂的电磁环境中，欺骗式干扰是雷达面对的主要干扰之一。现役装备通常采用波形捷变或副瓣匿影技术在波形域或空域对抗欺骗式干扰，但电子对抗装备的发展及抗干扰技术自身的限制大大削弱了雷达装备的生存能力，亟待研究更加稳健的欺骗式干扰对抗方法。欺骗式干扰对抗的关键在于寻找真假目标间的区别，本项目开展真假目标信号精细特征的分析与运用技术研究，首先从真假目标脉间特征入手，分析真假目标在微动特征上的差异及鉴别方法。同时，受个体辐射源识别技术的启发，在时频域中分析脉内精细特征的差异和相应的鉴别方法。两类方法互为补充，可以构成完整的欺骗式干扰对抗流程。以上研究工作用电子对抗设备自身的脉冲无意调制作为鉴别欺骗式干扰的依据，为开拓新的欺骗式干扰对抗技术奠定基础，显著提高雷达在复杂电磁环境下的抗干扰能力。

本项目针对复杂电磁环境下雷达面临的欺骗式干扰鉴别难问题，考虑到欺骗式干扰难以完全模拟微动特征、目标散射特性等细微特征，提出通过区分真假目标的细微特征实现欺骗式干扰鉴别，从根本上提高雷达抗干扰能力。研究成果包括：（1）在慢时间域建立了微多普勒调制模型，分析了不同微动特征、不同雷达观测参数带来的谱形差异。而欺骗式干扰大多只存在干扰吊舱运动带来的平动多普勒频移。因此，通过设计特征参数提取方法及分类器，完成了基于微多普勒特征的欺骗式干扰鉴别方法。.（2）项目深入分析了真实目标与欺骗式干扰的产生机理，根据真实目标物理散射过程导致的回波调制与欺骗式干扰机中存储转发带来的脉冲无意调制间的物理差异，建立了快时间回波模型。.（3）考虑快时间多脉冲叠加的问题，基于建立的快时间无意调制干扰模型，提出基于匹配滤波分离与时频域联合特征提取的鉴别思路。重点考虑难以提前获得干扰样本的困难，没有简单使用神经网络分类器，而以真假目标特征差异性与假目标间特征相似性为核心思想，设计了基于聚类的欺骗式干扰鉴别方法。.上述方法通过仿真数据和雷达实测数据的验证，本项目所研究方法能够有效提升欺骗式干扰的鉴别概率，切实提高雷达对抗能力。