摘要
为提高雷达辐射源信号的识别准确率和抗噪性能以满足现代复杂电子对抗的需要,文中提出一种基于PLDE和小波熵的雷达辐射源信号识别方法。先进行小波包分解,提取各个子频带的小波能量谱熵特征;再基于PLDE算法对特征的相像系数进行选择,选择分选能力强的特征重构特征向量;最后采用有向无环图支持向量机(DAGSVM)对不同调制参数的六类信号进行识别。仿真结果表明:该方法具有较优的抗噪性能,在不低于0d B信噪比情况下,用3维特征能获得高于95%的平均识别率。
To improve the recognition accuracy of radar emitter signals（ RES） and anti-noise property to meet to the requirements of modern complicated electronic warfare,a method for radar emitter signal recognition（ RESR） based on PLDE and wavelet entropy of sub-band is proposed. Wavelet packet decomposition is used to extract wavelet entropy feature of every sub-band. Then,the feature selection approach of feature ＇s resemblance coefficient based on PLDE is applied to select the features with excellent classification abilities wavelet entropy features to reconstruct feature vector. Finally,the radar emitter signals of different modulation types and different parameters modulated are recognized based on directed acyclic graph SVM（ DAGSVM）. The simulation results show that the time complexity of the method is very low and it has excellent anti-noise property,when the SNR is greater than 0 d B,the overall average recognition rate is higher than 95% only by 3 dimensional features.
作者
王尚平
胡杰
WANG Shang-ping;HU Jie(Key Electronic Information Control Lab.,Chengdu 610036,China;Unit 78111 of PLA,Chengdu 610000,China)
出处
《信息技术》
2018年第8期94-96,101,共4页
Information Technology
