一种基于支持向量机的数字调制识别方法  

Digital modulation recognition based on support vector machine classifier

作　　者：李侃[1] 孙进平[2] 成功[2] 毛士艺[2] 高飞[2] 

机构地区：[1]西南电子设备研究所产品部,四川成都610036 [2]北京航空航天大学电子信息工程学院,北京100083

出　　处：《电路与系统学报》2010年第3期7-12,共6页Journal of Circuits and Systems

摘　　要：自动调制方式识别是非协作通信研究中的重要内容。本文提出了一种基于支持向量机的数字通信信号调制方式识别方法。从信号的瞬时幅度,相位,频率和频谱等特性中提取了7个特征参数,用于训练支持向量机。与已有算法相比,本文算法结构简单、计算量小,同时考虑了符号成型的影响。仿真结果证明,当信噪比大于10dB时,算法的识别率可以达到95%以上。This paper proposes a new modulation classifier for digital communication signal based on support vector machine （SVM）. Automatic modulation classification is an important research area in the non-cooperative communication systems. Several feature parameters which are extracted from the instantaneous amplitude, phase frequency and frequency spectrum of signal are used to train the SVM classifier. Compared to other algorithms, the proposed approach has the traits of simple structure and high computation efficiency; and pulse shaping is also taken into account. The simulation results indicate that the scheme can achieve 95% recognition accuracy when SNR is above 10dB.

关 键 词：调制方式识别 特征提取 数字调制信号 支持向量机 

分 类 号：TN914[电子电信—通信与信息系统]