摘要
针对复杂体制雷达辐射源信号分类识别问题,提出了一种基于时频分析的雷达脉内调制识别算法。首先对时频矩阵在时间域进行等间隔分区,然后通过检测区间内信号时频能量峰值提取其时频特征,最后用支持向量机实现了分类识别。该方法以信号时频能量峰值分布的差异区别不同的雷达脉内调制方式,有效降低了特征维数。仿真结果表明,该方法对雷达脉内调制具有较好高识别正确率,而且具有较强的抗干扰能力。
Aiming at the problem of emitter identification caused by parameter complexity and agility of muhi-function radars, a new method for feature extraction of intra pulse modulated radar signals based on Time-Frequency （T-F） analysis is proposed. Firstly, the time-frequency matrix is divided into regions with the same interval; and then the intra-pulse features are extracted by detecting the peak time-frequency energy in each region; lastly, the support vector machines （SVM） is introduced to classify and recognize the pulse modulation of radar signals. The energy distribution characteristics of signals in T-F domains are used to discriminate the pulse modulation of different radar signals. Via this operation the feature dimension is decreased efficiently. Simulation result shows that the proposed algorithm can acquire a high recognition precision and has a good anti-noise performance.
出处
《电子技术应用》
北大核心
2012年第8期136-139,共4页
Application of Electronic Technique
关键词
脉内调制
时频分析
时频能量
支持向量机
pulse modulation
time-frequency analysis
time-frequency energy
support vector machines
