摘要
在分析提升小波应用在调制模式自动识别的基础上,提出了一种新的特征提取方法。该方法首先利用最优估计理论获得小波的最佳预测系数,根据最佳预测系数进行分解提取特征值,最后利用支持向量机分类器进行信号的分类识别。在求解支持向量机的参数优化问题中,提出了一种通用的解决方案,利用带惯性权重的粒子群算法来确定其最优系数。新方法提取的特征值经计算机仿真研究证明,该算法具有较好的工程应用性和有效性。
Based on analysis of the digital modulation recognition based on lifting wavelet, a new feature extraction method is proposed. First, the optimal estimation theory is used to obtain the best prediction coefficients, then the feature is extracted according to the distribution of the decomposition with second generation wavelets, and finally by using SVM classification machine, the categorized identification of the signal is done. In order to determine the optimal coefficient, a universal solution using particle swarm optimization algorithm is presented. The computer simulation on the extracted feature values indicates that this new algorithm is feasible and practicable in engineering application.
出处
《通信技术》
2012年第11期11-13,共3页
Communications Technology
关键词
调制识别
粒子群算法
提升小波
支持向量机
modulation recognition
particle swarm optimization＂ lifting wavelet
supportvector machine
