基于高阶累积量和谱分析的数字调制信号识别

Digital modulation signals recognition based on high order cumulant and spectral analysis

提出了一种基于高阶累积量和谱分析识别多种数字调制信号的算法。首先根据各调制信号四阶和八阶累积量的不同,定义一个特征参数实现信号的类间识别;其次根据不同调制信号二次方谱与四次方谱的不同,提取出相应的特征参数,从而实现信号的类内识别。仿真实验结果表明,该方法在较低信噪比条件下可以对2/4/8PSK、2/4/8FSK信号实现识别,且识别率较高,具有很强的实用性。

This paper proposes a digital modulation signal recognition technique based on higher order cumulant and spectrum analysis. Firstly,a feature based on the difference of signal＇s four order and eight order cumulant is defined to realize the inter class recognition. Then another characteristic parameters based on the difference of signals＇ square spectrum and quartic spectrum are extracted to realize the within class identification. The simulation results show that the proposed technique can recognize 2 /4 /8PSK,2 /4 /8FSK at low SNR,achieves high recognition rate,and is very practical.