低截获概率雷达信号特征分析技术研究

发布时间：2018-06-08 00:42

本文选题：雷达侦察 + LPI信号　；参考：《吉林大学》2016年博士论文

【摘要】：在现代电子战中,通过雷达侦察精确地获取敌方雷达的战术技术参数是进行干扰和打击的前提。低截获概率(LPI)雷达的出现给雷达侦察信号的截获和分析处理带来了新的挑战,有效地截获LPI雷达信号并准确分析其脉内特征成为当前雷达侦察的核心问题。受制于研制时期的技术限制,我军现役雷达侦察装备无法有效截获分析LPI雷达信号。为此,论文以实际科研项目为依托,针对LPI雷达信号的截获与脉内特征分析的关键技术开展了深入研究。对LPI雷达的有效截获是实施LPI雷达信号特征分析的基础,而现代LPI雷达普遍采用极化调制技术。首先,针对传统单极化侦察参数测量准确性不高的问题,提出了雷达侦察全极化接收技术,给出了全极化接收机的实现方案。其次,为解决民用通信信号对雷达侦察设备的影响问题,提出了一种基于极化滤波的民用通信信号抑制方法,该方法依据雷达侦察信号与民用通信信号的极化差异,在极化域滤除民用通信信号,提取出雷达侦察信号,理论分析和仿真实验均验证了该方法的有效性。针对低信噪比下LPI雷达信号脉内调制类型识别准确性不高的问题,提出了一种基于层次决策的脉内调制类型自动识别算法。该算法首先利用截获信号的时宽带宽积识别非调制信号,然后基于相位差分和改进短时傅里叶变换(STFT)提取特征参数,最后依据各调制类型的不同特性采用层次决策方法实现调制类型识别。仿真实验结果表明,该识别算法的分类特征明显,识别类型完整,抗噪性强,具有较高的工程应用价值。在频率调制信号参数估计的研究中,针对低信噪比下正弦调频连续波信号的参数估计问题,提出了基于信号自相关和重排累积谱图的调制参数估计方法,提高了低信噪比下正弦调频连续波信号的参数估计性能。针对频率编码信号的参数估计问题,首先引入多尺度路径追踪算法自适应地完成信号的调制参数估计,该方法具有较高的估计精度,然后提出利用小波变换和平滑功率谱进行部分参数预估,降低多尺度路径追踪算法的计算复杂度。仿真结果表明,相比其它方法,本文提出的两种频率调制信号参数估计方法均具有较高的估计性能。在多相编码信号参数估计与类型识别的研究中,针对多相编码信号在低信噪比下估计性能不佳的问题,提出了基于二次相位函数(QPF)和分数阶傅里叶变换(Fr FT)的多相编码信号参数估计方法。仿真实验结果表明,该方法能够同时累积信号的多条时频脊线,较好地解决了低信噪比下的多相编码信号参数估计的问题,且参数估计性不受制于编码类型。其次,针对目前多相编码信号调制类型识别准确率低的问题,提出了基于多特征融合支持向量机(SVM)的调制类型识别方法,实现了多相编码信号调制类型的高准确率识别。
[Abstract]:In modern electronic warfare, accurate acquisition of tactical and technical parameters of enemy radar by radar reconnaissance is the prerequisite for jamming and striking. The emergence of low probability of interception (LPI) radar brings a new challenge to the interception and analysis of radar reconnaissance signal. It is the core problem of current radar reconnaissance to intercept LPI radar signal effectively and accurately analyze its in-pulse characteristics. Limited by the technical constraints during the development period, our active radar reconnaissance equipment can not effectively intercept and analyze LPI radar signals. Therefore, based on the actual scientific research projects, the key technologies of LPI radar signal interception and intra pulse feature analysis are studied in this paper. The effective interception of LPI radar is the basis of signal characteristic analysis of LPI radar, and polarization modulation technology is widely used in modern LPI radar. Firstly, aiming at the problem that the accuracy of the traditional single polarization reconnaissance parameter measurement is not high, a radar reconnaissance full polarization receiver technology is proposed, and the realization scheme of the full polarization receiver is given. Secondly, in order to solve the problem of the influence of civil communication signal on radar reconnaissance equipment, a method of civil communication signal suppression based on polarization filtering is proposed, which is based on the polarization difference between radar reconnaissance signal and civil communication signal. The radar reconnaissance signal is extracted by filtering the civil communication signal in polarization domain. The effectiveness of the method is verified by theoretical analysis and simulation experiments. Aiming at the low SNR of LPI radar signal the accuracy of intra-pulse modulation type recognition is not high. A hierarchical decision based automatic recognition algorithm for intra-pulse modulation type is proposed. Firstly, the unmodulated signal is identified by the time-wideband wide product of the intercepted signal, and then the characteristic parameters are extracted based on the phase difference division and the improved STFT-based short time Fourier transform (STFT). Finally, according to the different characteristics of each modulation type, the hierarchical decision method is used to realize modulation type recognition. The simulation results show that the classification feature of the algorithm is obvious, the recognition type is complete, the noise resistance is strong, and the algorithm has high engineering application value. In the research of frequency modulation signal parameter estimation, a modulation parameter estimation method based on autocorrelation and rearrangement cumulant spectrum is proposed to estimate the parameters of sinusoidal frequency modulated continuous wave signal under low SNR. The parameter estimation performance of sinusoidal FM continuous wave signal with low SNR is improved. To solve the parameter estimation problem of frequency-coded signal, a multi-scale path tracing algorithm is introduced to estimate the modulation parameters of the signal adaptively. The proposed method has a high estimation accuracy. Then wavelet transform and smoothing power spectrum are used to estimate some parameters to reduce the computational complexity of multi-scale path tracing algorithm. The simulation results show that the two methods proposed in this paper have higher estimation performance than other methods. In the research of parameter estimation and type recognition of multiphase coded signal, the problem of poor estimation performance of multiphase coded signal under low SNR is discussed. A method for parameter estimation of multiphase coded signals based on quadratic phase function (QPF) and fractional Fourier transform (Fr FT) is proposed. The simulation results show that the proposed method can accumulate multiple time-frequency ridges at the same time and solve the problem of parameter estimation of polyphase coded signals at low SNR, and the estimation of parameters is not restricted by coding type. Secondly, aiming at the problem of low accuracy of modulation type recognition of polyphase coded signals, a modulation type recognition method based on multi-feature fusion support vector machine (SVM) is proposed, which realizes the high accuracy recognition of modulation types of multi-phase coded signals.
【学位授予单位】：吉林大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：TN957.51

【相似文献】