基于相位聚类的MPSK信号调制分类算法  

A new classification algorithm of MPSK signals based on phase clustering

作　　者：徐健飞[1] 汪芙平[1] 王赞基[1] 

机构地区：[1]清华大学电机系电力系统国家重点实验室,北京100084

出　　处：《电路与系统学报》2011年第5期55-59,共5页Journal of Circuits and Systems

摘　　要：针对非协作通信中的MPSK信号调制识别问题,提出基于相位聚类的调制识别算法。该算法以MPSK信号相位聚类的周期性为分类特征,应用傅里叶变换作为特征提取的方法,实现了对MPSK信号的类内识别。同时本文还对该算法进行了改进,减少了该算法的计算量,并使该方法可对含有载波频偏的MPSK调制信号进行识别。该算法在信噪比较低的情况下识别准确率高,且可对高阶MPSK信号进行分类识别。仿真实验表明,当码元长度为500,信噪比为3dB时,算法对BPSK、QPSK、8PSK的分类识别率达到了98.7%;当信噪比提高到9dB时,对16PSK也能进行较好的分类识别。In the area of noncooperation communication,a classification algorithm of MPSK signals based on the phase clustering is proposed.To classify the modulation type of the signal,the periodicity of phase clustering of MPSK signal is used as classification feature which can be extracted by FFT.Moreover,the method is improved in this paper that it is extended to MPSK baseband signals with frequency offset,and the calculation complexity is reduced.This algorithm can provide a high probability of successful classification of MPSK at low SNR and even the higher-order MPSK can be classified.Simulation result shows that the accuracy rate of the classification of BPSK,QPSK and 8PSK signals can reach 98.7% when the symbol length is 500,SNR is 3dB,and 16PSK signals can also be well classified when the SNR is improved to 9dB.

关 键 词：MPSK 相位聚类 调制分类 载波频偏 

分 类 号：TN911.72[电子电信—通信与信息系统]