基于EMD和二叉树支持向量机的调制识别算法  

An Algorithm for Digital Signals' Modulation Recognition Based on EMD and Binary Tree of SVM

作　　者：闫朋展[1] 王振宇[1] 杨少华[1] 

机构地区：[1]解放军电子工程学院,合肥230037

出　　处：《火力与指挥控制》2012年第4期134-137,共4页Fire Control & Command Control

摘　　要：针对数字调制模式识别问题,提出了一种基于经验模式分解(EMD)和二叉树支持向量机的识别算法。该算法通过对数字信号进行经验模式分解,利用分解得到的本征模式函数构造分类特征,再运用二叉树支持向量机作为分类器实现了6种信号的分类。仿真结果表明,该算法具有很好的识别性能。Based on EMD and binary tree of SVM an algorithm is proposed for digital signals＇ modulation recognition.Using Intrinsic Mode Functions which are obtained by Empirical Mode Decomposition,this algorithm forms two classification features at first,then identifies QAM and FSK by taking binary tree of SVM as classifier.The simulation results show this algorithm has good performance in the automatic modulation recognition.

关 键 词：调制识别 经验模式分解 二叉树支持向量机 

分 类 号：TN911[电子电信—通信与信息系统]