一种改进的BPSK/QPSK信号调制识别算法

摘要

A modulation recognition algorithm based on the feature of order statistics is proposed for binary phase shift keying/quadrature phase shift keying( BPSK/QPSK) signals. The spectrum of squaring trans-formation of the observed signal is calculated by using discrete Fourier transform( DFT) . Consequently,the modified spectrum is obtained by removing the maximum value of the magnitude spectrum. The maximum value of the modified spectrum is defined as the recognition feature. By comparing the recognition feature with a threshold which is set based on the constant false alarm rate( CFAR) criterion,BPSK or QPSK signal can be identified. Computer simulation results show that the proposed algorithm has a good performance to discrminate the BPSK or QPSK signal under moderate signal-to-noise ratio( SNR) . The average recogni-tion accuracy is greater than 90% when the SNR is above 1 dB.%提出了一种基于顺序统计量特征的二进制相移键控/正交相移键控(BPSK/QPSK)信号调制识别算法.对观测信号进行平方并做离散傅里叶变换运算,将变换结果取模后去除最大值得到修正频谱,并取修正频谱的最大值作为识别特征量,利用恒虚警准则确定判决门限,将识别特征量与门限比较来完成对BPSK和QPSK两种调制信号的识别.计算机仿真表明,当信噪比适度时,所提算法可对BPSK及QPSK两类信号进行有效识别.当信噪比大于1 dB时,算法的平均识别正确率达到90%以上.