一种新的MPSK信号调制分类算法

摘要

A new algorithm based on two-order and four-order cyclic cumulants is proposed for modulation classification of Mary Phase Shift Keying（MPSK） signals.MPSK signals are classified to BPSK and QPSK or more than 4PSK modulation subclasses by the feature of their two-order and four-order cyclic cumulants,then a feature parameter is picked up from signals＇ two-order and four order cyclic cumulants at the cyclic frequency which equals the frequency of signals,and finally BPSK and QPSK are classified by the feature parameter.Even if the probabilities of codes are unequal,the algorithm is still effective.Because the cyclic cumulants of Gaussian noise which are higher than two-order are zero,to some extent,the algorithm can depress Gaussian noise.The result of simulation shows that the algorithm is efficient.%针对多进制相移键控（MPSK）信号的调制识别问题,提出了一种新的基于2、4阶循环累积量的调制识别算法。先利用信号的2阶和4阶循环累积量的特征,将MPSK信号分为BPSK和QPSK与8PSK以上两大调制子类,再由信号循环频率等于信号载频处的2阶和4阶循环累积量,构成BPSK和QPSK的分类特征量,实现对BPSK和QPSK信号的调制识别。该算法在MPSK信号码元不等概率出现时依然有效。由于高斯噪声的循环累计量均为0,因此此算法具有一定的抗高斯噪声能力。实验仿真结果验证了该算法的有效性。