摘要
利用通信信号的循环平稳特性,在循环累积量域内构造信号分类特征矢量,采用支持矢量机将分类特征矢量映射到高维空间并构建最优分类超平面,实现对QAM调制信号的自动识别。该算法解决了样本在低维空间中的不可分问题,具有良好的泛化推广性能,并且可在多种调制信号环境下实现对感兴趣信号类型的识别。理论分析和仿真结果均证明了算法的正确性和有效性。
A support vector machines （SVM） based algorithm for the automatic classification of QAM signals is proposed. The algorithm utilizes the cyclostationary property of communication signals and presents classification features in cyclic cumulants domain. SVM maps the classification features nonlinearly into the high dimension space and constructs the optimal separating hyperplane in the space to classify interesting signals. The algorithm resolves the non-separable problem in low dimension space and has high generalization performance. Interesting signals can also be classified under the presence of interference signals. The efficiency of the proposed classification algorithm is verified via theoretical analysis and extensive simulations.
出处
《系统工程与电子技术》
EI
CSCD
北大核心
2007年第4期520-523,共4页
Systems Engineering and Electronics
基金
国家自然科学基金(60572146)
高等学校优秀青年教师教学科研奖励计划
国家自然科学基金重大项目(60496316)
国家"863"重大课题(2005AA123910)
教育部科学技术研究重点项目(107103)
博士点基金资助课题(20050701007)
关键词
调制识别
循环累积量
循环平稳性
支持矢量机
modulation classification
cyclic cumulants
cyclostationary
support vector machine
