摘要
雷达脉内特征分析是信号分选领域一个重要手段,当脉内信号有多个雷达信号交叠在一起时,很难分析出其各自的调制方式。经验模态分解(EMD)在分析非平稳混合信号时效果显著,但其存在2个明显弊端。针对端点效应问题,首先采用小波分解将信号分解成各分量,随后对除噪声外的各分量利用ARMA模型对信号进行预测,接着对预测后的各分量进行小波重构,从而消除了端点效应,针对虚假分量的问题,结合雷达信号的特点改进了其终止条件,提高了EMD分解的性能。最后,对EMD分解后的各分量进行时频分析,得出雷达脉内信号的调制特征。仿真验证了算法的有效性。
In the area of signal sorting, it is an important method to analyze radar signal in pulse. While the signal in one pulse is built up by more than one radar emitters, it＇ s hard to analyze the modulation methods for each signal. Empirical mode decomposition （EMD） is so efficient in processing non-stationary combined signal, but there are also two obvious disadvantages. Aimed at end effect, firstly wavelet decomposition has been used to decompose the combined signal into some series, and then the ARMA model has been used to forecast the series except the yawp, at last wavelet reconstruction has been used to build a new signal according to the series before to eliminate the end effect. Aimed at false modes, a new stopping condition has been proposed combined with the characteristic of radar signal, so the effect of EMD has been enhanced. Subsequently, each mode has been analyzed by time-frequency analysis method, and the modulation methods for each signal can be obtained. And this new method is verified efficient by the simulation.
出处
《现代防御技术》
北大核心
2011年第4期138-143,共6页
Modern Defence Technology
关键词
脉内调制
经验模态分解
端点效应
终止条件
时频分析
pulse modulation
empirical mode decomposition （EMD）
end effect
stopping condition
time-frequency analysis
