摘要
文中提出了一种基于高阶累积量和支持向量机的数字调制信号识别新算法,即从信号的四阶和六阶累积量中提取的参数作为分类特征向量,利用基于二叉树的支持向量机作为分类器的方法实现了2ASK、4ASK、8ASK、4PSK、8PSK等五种数字调制信号的识别。仿真结果表明,在信噪比为10dB情况下,该算法的正确识别率达到了95.83%以上。
This paper presents a new algorithm for digital modulation signals based on high order cumulants（HOC）and support vector machines（SVM）.The characteristic parameters picked up from the four-order and six-order cumulants of the signals are used as the classification vectors. The method could, by using binary tree-based SVM as the classifier, identify five digital modulation signals, that is, 2ASK, 4ASK, 8ASK, 4PSK and 8PSK digital signals. Computer simulation indicates that the success rate is over 95.83% at SNR-10dB.
出处
《通信技术》
2009年第11期46-47,51,共3页
Communications Technology
关键词
调制识别
高阶累积量
支持向量机
二叉树
modulation identification
higher-order cumulants
support vector machine
binary tree
