基于时频二维的雷达信号脉内调制识别方法

本文提出一种基于对时频矩阵分布图像的边缘信息与瞬时频率序列中的细节信息的提取来实现雷达信号脉内调制类型识别的算法。一方面对雷达信号时频矩阵进行图像处理强化边缘信息,并对其进行奇异值分解,另一方面得到信号的瞬时频率序列,基于AR模型得到其模型参数,得到信号的特征参数,构建相应的分类器进行调制类型的识别。通过一系列蒙特卡洛仿真表明,这种算法有较强的抗噪声能力,对信号的脉内参数变化不敏感,在低信噪比、信号参数变化与小样本条件下可以以较高的识别率得到稳定的识别结果。

Radar Signal Intra-Pulse Modulation Identification Based On Time-Frequency Distribution

An algorithm based on edge information extracted from time-frequency matrix distribution image and detailed information extracted from instantaneous frequency sequence is introduced for radar signal intra-pulse modulation identification.To enhance edge information by way of image processing to radar signal time-frequency matrix and extract the features using singular value decomposition;and to obtain instantaneous frequency sequence of the signal and AR modeling parameters and characteristic parameters of signal,then build relevant classifier for identification of modulation.A series of Monte Carlo simulations show that this algorithm features high anti-noise capability and it is insensitive to intra-pulse parameter variation,stable identification result can be obtained even in the event of low signal-to-noise ratio,signal parameter varying or small simples.