摘要
针对低信噪比下传统方法识别雷达信号脉内调制类型准确率低的问题,提出了一种有效的脉内调制类型自动识别方法。该方法首先计算雷达信号的模糊函数,然后利用图像处理算法提取其奇异值特征,将奇异值特征矢量作为神经网络的输入对脉内调制类型自动识别。仿真表明该方法在0dB信噪比下,对常见脉内调制信号识别率均大于84%。该方法需要的特征维数少、分类器结构简单、识别率高、抗噪能力强。
Aiming at the problem of low correct recognition rate to the modulation types of intra-pulse radar signal with common methods, an effective automatic recognition method of intra-pulse modulation types is put forward. Firstly, it calculates Ambiguity Function （AF） of radar signal, and then uses image processing algorithm to extract its singular value. The singular value vector is used as the in-put eigenvector of ANN to identify the intra-pulse modulation types automatically. The simulation re- suits show that the correct recognition rate is more than 84 % to common intra-pulse radar signal in the 0dB SNR. It is of fewer dimension of the character vectors, simple structure of classifiers, high correct recognition rate and isn＇t sensitive to noise.
出处
《电子信息对抗技术》
2011年第2期30-35,共6页
Electronic Information Warfare Technology
关键词
自动识别
脉内调制
模糊函数
奇异值分解
神经网络
automatic recognition
intra-pulse modulation
ambiguity function
singular value distribution
ANN
