大频偏条件下一种新的调制识别方法

A Novel Modulation Recognition Scheme under Large Frequency Offset

针对存在较大残留频偏的MPSK和MQAM信号,提出了一种基于概率分布匹配的调制识别新方法。该方法先利用幅度方差区分MPSK和MQAM信号。对于MPSK信号,利用瞬时相位的差分运算降低残留频偏的影响,再统计表示相邻符号间相位状态相同的比例的特征量,分级匹配实现阶数识别;对于MQAM信号,则统计与残留频偏无关的幅度聚类分布,然后匹配识别其调制阶数。仿真实验结果表明,该方法在较大的残留频偏范围内具有优良的识别性能。

For MPSK and MQAM signals under large residual frequency offset, a novel scheme of modulation recognition based on matching of probability distribution function was proposed. The scheme classifies MPSK and MQAM by the variance of amplitude. For MPSK, instantaneous phase＇s difference operation reduced the effect of residual frequency offset, then the ratio of adjacent symbols with the same phase state was calculated, to classify the modulation orders by hierarchically matching; For MQAM, the distribution of amplitude clusters was estimated, which was independent to the frequency offset. Then the distribution was compared with the candidate modulation schemes to recognize the modulation order. Simulation experiments indicate that, the proposed method possesses a good performance in the condition of large residual frequency offset.