摘要
从特征参数提取的角度对基于决策论的数字信号调制识别算法进行改进,提取五个相对简单的瞬时信息特征参数,并利用决策树方法对信号进行调制识别。改进后的算法除了识别2ask、2fsk、2psk、4ask、4fsk和4psk六种典型数字信号外,还可识别16qam,若进一步增加参数(递归零中心归一化瞬时相位绝对值的平均值),又可识别8psk。实验结果表明:改进算法的复杂度明显降低,且信号识别正确率及适用信噪比范围都有很大程度的提高。
An improved method of digital signals modulation identification based on decision theory was proposed in this paper. The method can be used to carry out modulation identification by extracting five simple instantaneous characteristic parameters of digital signals and decision-tree. Besides the six typical digital signals i. e. 2ask, 2fsk, 2psk, 4ask, 4fsk and 4psk, 16qam can be identified by the improved method. Moreover, if the parameter which was the mean absolute value of reeursive zero-center normalization instantaneous phase was extracted, 8psk can also be identified by this method. The experimental result shows that the complexity of the improved method is reduced greatly. Furthermore, the correct recognition rate and the range of SNR are also increased significantly.
出处
《计算机应用》
CSCD
北大核心
2009年第12期3227-3230,共4页
journal of Computer Applications
基金
贵州省教育厅自然科学基金资助项目(黔科教〔2007〕321号)
关键词
特征提取
调制识别
决策论
识别正确率
信噪比
feature extraction
modulation identification
decision theory
recognition rate
Signal to Noise Rate （SNR）
