用高阶累积量实现数字调相信号的分级识别  

Hierarchical Modulation Recognition of MPSK Signals Using Higher Order Cumulants

作　　者：蔡占辉[1] 姚远程[1] 秦明伟[1] 

机构地区：[1]西南科技大学信息工程学院,四川绵阳621010

出　　处：《西南科技大学学报》2011年第4期71-74,共4页Journal of Southwest University of Science and Technology

基　　金：国家自然科学基金委员会与中国工程物理研究院联合基金资助项目(10876035)

摘　　要：针对BPSK,QPSK,OQPSK和8PSK信号的调制模式自动识别,传统的基于高阶累积量算法无法区分QPSK和OQPSK,因此提出了一种基于差分高阶累积量的识别算法。该算法首先用四阶累积量提取待识别信号和其差分序列的特征参数,然后用决策树分类法实现信号的分级识别。理论分析和计算机仿真结果表明该算法有较强的抗噪声和抗相位抖动能力,在信噪比>3 dB时识别率达95%以上,更适用于较低信噪比下信号的识别。A new method based on feature extraction was proposed for the recognition of M-ary Phase Shift Keying （MPSK） signals. As features, fourth order cumulants of the input samples and phase differential sequences were applied, and classifier based on decision tree was used. Theoretical arguments were verified via extensive simulations. It is shown that the cumulant-based features have robust anti-noise and an- ti-phase jitter ability. Simulation results demonstrate that the correct classification probability （Pcc） with the proposed algorithm is above 95% when signal-to-noise （SNR） 〉 3 dB, thus this method is more suitable for the identification of MPSK signals at low SNR situation.

关 键 词：调制识别 高阶累积量 相位差分 决策分类 

分 类 号：TN911.3[电子电信—通信与信息系统]