The effects of complex electromagnetic environment on a radar system involve many factors and their interrelationships are also complicated, so it is difficult to quantitatively describe the impact of jamming on the radar. In this study, the authors start from the signal models of several jamming types and the basic signal processing flow of pulse-Doppler radar and propose a unified evaluation framework based on expert knowledge and data mining. On the one hand, the evaluation index from expert knowledge is calculated by the two-dimensional image entropy of the time–frequency distributions of echoes. On the other hand, the evaluation index from data mining is constructed from the principle patterns extracted from samples by robust principal component analysis. In addition, the fused evaluation index is achieved by taking a weighted average of the indices from expert knowledge and data mining, where the weights are designed from the samples. Finally, the proposed evaluation approach is validated by analysing its feasibility, consistency and robustness in experiments and the results show that it can stably and quantitatively measure the impact of jamming on the radar.

中文翻译：

---

基于专家知识和数据挖掘的干扰对雷达影响评估方法

复杂的电磁环境对雷达系统的影响涉及很多因素，而且它们之间的相互关系也很复杂，因此很难定量描述干扰对雷达的影响。在这项研究中，作者从几种干扰类型的信号模型和脉冲多普勒雷达的基本信号处理流程入手，并提出了基于专家知识和数据挖掘的统一评估框架。一方面，专家知识的评估指数是通过回波的时频分布的二维图像熵来计算的。另一方面，通过稳健的主成分分析从样本中提取的主要模式构建数据挖掘的评价指标。此外，通过从专家知识和数据挖掘中获取指标的加权平均值来获得融合的评估指标，其中权重是根据样本进行设计的。最后，通过分析其可行性，一致性和鲁棒性，对所提出的评估方法进行了验证，结果表明，该方法可以稳定，定量地测量干扰对雷达的影响。