一种非协作通信信号特征提取方法

A method of non-cooperative communication signals feature extraction

在非协作通信中,如何在低信噪比的衰落信道下,从截获的信号中提取特征参数,从而准确的判别其调制方式是研究的重点。特征参数的高阶累积量可以去除噪声干扰,针对当前非协作通信领域中低信噪比下调制识别率不高的问题,结合高阶累积量的自身特性,提出新的信号特征提取方法。以Nakagami信道为例,以信号的四阶累积量作为特征参量,并对四阶累积量进行分类门限的划分,主要讨论BPSK,8PSK和16QAM这3种信号在不同符号数下的准确识别率。仿真表明,8PSK在符号数为500时,在信噪比为-5 dB到0 dB的高斯白噪声信道下,识别率在90%以上。

In non-cooperative communication,channel quality is poor and signal fading is serious.In the low SNR fading channel,how to extract the characteristic parameters from the intercepted signal so as to accurately distinguish its modulation mode,is the focus of this paper.The high-order cumulant of the feature parameters can remove the noise interference.Aiming at the problem that the modulation recognition rate is not high under the low SNR in the current non-cooperative communication field,a new signal feature extraction method is proposed with the characteristics of the high-order cumulant in view.In this paper,taking Nakagami channel as an example,the fourth-order cumulant of signal is taken as the characteristic parameter,and the classification threshold of the fourth-order cumulant is divided.Simulation results show that the recognition rate of 8PSK is more than 90%when the number of symbols is 500 and the SNR is-5 dB to 0 dB.When the signal-to-noise ratio is greater than 10 dB,the modulation recognition rate is 100%.