This paper presents a novel adaptive algorithm for multicomponent signal decomposition from the time–frequency (TF) plane using the short-time Fourier transform (STFT). The approach is inspired by a common technique used within radar detection called constant false alarm rate (CFAR). The areas with the strongest magnitude are detected and clustered, allowing for TF mask creation and filtering only those signal modes that contribute the most. As a result, one can extract a particular component void of noise and interference regardless of the signal character. The superiority understood as an improved reconstructed waveform quality of the proposed method is shown using both simulated and real-life radar signals.

中文翻译：

---

短时傅里叶变换的雷达检测启发式信号检索

本文提出了一种新的自适应算法，用于使用短时傅里叶变换 (STFT) 从时频 (TF) 平面进行多分量信号分解。该方法的灵感来自雷达检测中使用的一种常见技术，即恒定误报率 (CFAR)。检测和聚类具有最强幅度的区域，允许创建 TF 掩模并仅过滤那些贡献最大的信号模式。因此，无论信号特征如何，都可以提取没有噪声和干扰的特定组件。使用模拟和现实生活中的雷达信号显示了被理解为所提出方法的改进的重建波形质量的优越性。