低截获概率(low probability of intercept,LPI)雷达信号凭借其卓越的抗截获能力,在现代电子战中得到了广泛应用。但LPI雷达信号的低峰值功率使其极易被加性白高斯噪声(additive white Gaussian noise,AWGN)淹没,导致信噪比(signal-to-n...低截获概率(low probability of intercept,LPI)雷达信号凭借其卓越的抗截获能力,在现代电子战中得到了广泛应用。但LPI雷达信号的低峰值功率使其极易被加性白高斯噪声(additive white Gaussian noise,AWGN)淹没,导致信噪比(signal-to-noise ratio,SNR)较低,给信号的检测和识别带来了极大的挑战。为了从AWGN背景中提取原始LPI雷达信号,本文提出了一种名为LPI-U-Net的深度神经网络(deep neural network,DNN),用于端到端的时域LPI雷达信号增强。该网络由特征提取模块(feature extract module,FEM)、特征聚焦模块(feature focus module,FFM)和信号恢复模块(signal recover module,SRM)组成。首先FEM通过卷积操作提取信号的特征,然后FFM利用卷积和通道间注意力进一步关注对信号增强任务有利的特征,最后SRM利用反卷积操作从特征中重构信号,从而完成LPI雷达信号增强。仿真实验表明LPI-U-Net在低SNR下的LPI雷达信号增强性能优于传统信号处理中典型的降噪方法,验证了其可行性和有效性。展开更多
In this paper,we present a novel unimodular sequence design algorithm based on the coordinate descent(CD)algorithm,aimed at countering electronic surveillance(ES)systems based on cyclostationary analysis.Our algorithm...In this paper,we present a novel unimodular sequence design algorithm based on the coordinate descent(CD)algorithm,aimed at countering electronic surveillance(ES)systems based on cyclostationary analysis.Our algorithm not only provides resistance against cyclostationary analysis(CSA)but also maintains low integrated sidelobe(ISL)characteristics.Initially,we derive the expression of the cyclostationary feature(CSF)detector and simplify it into an iterative quadratic form.Additionally,we derive a quadratic form to ensure the similarity of the autocorrelation sidelobes.To balance the minimization of the detection probability and the ISL values,we introduce a Pareto scalar that transforms the multiobjective optimization problem into a convex combination of objective functions.This approach allows us to find an optimal trade-off between the two objectives.Finally,we propose a monotonic algorithm based on the CD algorithm to counter CSA analysis.This algorithm efficiently solves the optimization problem mentioned earlier.Numerical experiments are conducted to validate the correctness and effectiveness of our proposed algorithm.展开更多
This paper proposes a desirable method to detect different kinds of low probability of intercept (LPI) radar signals, targeted at the main intra-pulse modulation method of LPI radar signals including the signals of li...This paper proposes a desirable method to detect different kinds of low probability of intercept (LPI) radar signals, targeted at the main intra-pulse modulation method of LPI radar signals including the signals of linear frequency modulation, phase code, and frequency code. Firstly, it improves the coherent integration of LPI radar signals by adding the periodicity of the ambiguity function. Then, it develops a frequency domain detection method based on fast Fourier transform (FFT) and segmented autocorrelation function to detect signals without features of linear frequency modulation by virtue of the distribution characteristics of noise signals in the frequency domain. Finally, this paper gives a verification of the performance of the method for different signal-to-noise ratios by conducting simulation experiments, and compares the method with existing ones. Additionally, this method is characterized by the straightforward calculation and high real-time performance, which is conducive to better detecting all kinds of LPI radar signals.展开更多
低截获概率(low probability of intercept,LPI)雷达已成为新时代雷达装备中关键的技术体制或工作模式,针对LPI雷达信号调制识别及参数估计方法的研究是当前雷达对抗侦察领域的热点。首先,分析了几种典型LPI雷达信号的脉内特征,梳理了LP...低截获概率(low probability of intercept,LPI)雷达已成为新时代雷达装备中关键的技术体制或工作模式,针对LPI雷达信号调制识别及参数估计方法的研究是当前雷达对抗侦察领域的热点。首先,分析了几种典型LPI雷达信号的脉内特征,梳理了LPI雷达信号调制识别及参数估计的传统和主流方法,并说明其原理、优缺点和研究现状。最后,总结了现有LPI雷达信号调制识别及参数估计方法尚存的问题,并指出其未来发展趋势,旨在为今后的研究提供参考。展开更多
针对先验信息残缺的非合作电子对抗背景下的低截获概率雷达信号识别问题,提出一种基于改进的半监督朴素贝叶斯的识别算法。该算法首先提取出4种低截获概率(low probability of intercept,LPI)雷达信号的双谱对角切片作为识别特征;针对...针对先验信息残缺的非合作电子对抗背景下的低截获概率雷达信号识别问题,提出一种基于改进的半监督朴素贝叶斯的识别算法。该算法首先提取出4种低截获概率(low probability of intercept,LPI)雷达信号的双谱对角切片作为识别特征;针对传统的半监督朴素贝叶斯(semi-supervised Na?ve Bayes,SNB)在更新训练样本集过程中会产生迭代错误的不足,利用改进的SNB(Revised SNB,RSNB)算法构建分类器,完成对测试样本的识别。该方法通过在无标记样本集生成的置信度列表中选取置信度较高的样本添加到有标记样本集中,再利用预测后的分类结果对分类器参数(即特征期望向量珡mi和方差向量σi)进行改进,有效解决了传统算法分类精度低且分类性能不稳定等缺点。理论分析和仿真结果表明,在LPI雷达信号识别问题,相比于SNB算法和传统的主成分分析加支持向量机法(principal component analysis-support vector machine,PCA-SVM),该算法具有更高的分类识别率和更好的分类性能。展开更多
合理的雷达低截获(low probability of interception,LPI)性能评估方法是提高其隐身性能的基础。针对雷达LPI性能难以有效实时评估的问题,提出一种群广义直觉模糊软集(group-generalized intuitionistic fuzzy soft sets,G-GIFSS)算法...合理的雷达低截获(low probability of interception,LPI)性能评估方法是提高其隐身性能的基础。针对雷达LPI性能难以有效实时评估的问题,提出一种群广义直觉模糊软集(group-generalized intuitionistic fuzzy soft sets,G-GIFSS)算法与主客观权重相结合的雷达LPI性能评估方法。首先从反映雷达低截获性能的3个准则层信号层、功率层以及天线层确定6个目标属性指标层,选择直觉模糊集熵法确定客观权重、层次分析法((analytic hierarchy process,AHP))确定主观权重,并线性合成主客观权重。结合G-GIFSS算法利用多专家参量集的优势,对雷达LPI性能进行综合评判。通过案例分析并与经典评估方法对比,验证了该方法的优越性。展开更多
针对LPI信号分类识别问题中,时频图像受噪声干扰严重的问题,提出了一种基于二维快速经验模式分解(FBEMD)的图像降噪算法,并利用该算法实现对LPI信号的分类。首先利用时频分析方法,获得待分类信号的时频分布图像;使用二维EMD分解算法对...针对LPI信号分类识别问题中,时频图像受噪声干扰严重的问题,提出了一种基于二维快速经验模式分解(FBEMD)的图像降噪算法,并利用该算法实现对LPI信号的分类。首先利用时频分析方法,获得待分类信号的时频分布图像;使用二维EMD分解算法对图像降噪;截取包含时频信息的图像部分,通过主分量分析法提取特征矢量;最后采用RBF神经网络完成信号的分类识别任务。对常见的LPI雷达信号进行仿真,结果表明较低信噪比情况下,该方法仍能获得较好的分类结果。当信噪比为-2 d B时,采用二维EMD降噪算法,平均正确识别率能够达到93%。展开更多
为了优化单发多收协同雷达(single-transmitter multi-receiver cooperative radar,SMCR)探测系统的低截获概率(low probability of interception,LPI),利用SMCR目标探测的截获因子构造优化目标函数。首先,在二维平面上描述SMCR目标探...为了优化单发多收协同雷达(single-transmitter multi-receiver cooperative radar,SMCR)探测系统的低截获概率(low probability of interception,LPI),利用SMCR目标探测的截获因子构造优化目标函数。首先,在二维平面上描述SMCR目标探测场景,分析探测区域内接收机队列的接收增益及其近似估计方法。然后,针对目标位置先验已知情况,建立SMCR系统的接收机队列优化模型,分析模型解集。最后,针对目标搜索区域先验已知情况,从多个维度仿真分析接收机队列的LPI特性。仿真结果表明,针对目标位置或目标搜索区域先验已知的SMCR探测场景,接收机队列的队形设计有利于改善系统的LPI性能。针对目标位置已知的实测数据定性说明了所提方法仿真结果的合理性。展开更多
针对低信噪比(signal to noise ratio,SNR)低截获概率(low probability of intercept,LPI)雷达脉内波形识别准确率低的问题,提出一种基于时频分析、压缩激励(squeeze excitation,SE)和ResNeXt网络的雷达辐射源信号识别方法。首先通过Cho...针对低信噪比(signal to noise ratio,SNR)低截获概率(low probability of intercept,LPI)雷达脉内波形识别准确率低的问题,提出一种基于时频分析、压缩激励(squeeze excitation,SE)和ResNeXt网络的雷达辐射源信号识别方法。首先通过Choi-Williams分布(Choi-Williams distribution,CWD)获得雷达时域信号的二维时频图像(time-frequency image,TFI);然后进行TFI预处理降低噪声干扰和频率维的位置分布差异,以适应深度学习网络输入;最后在ResNeXt基础上加入扩张卷积和SE结构提取TFI特征,实现雷达辐射源分类。实验结果表明,SNR低至-8 dB时,该方法对12类常见LPI雷达波形的整体识别准确率依然能达到98.08%。展开更多
低截获概率(low probability of intercept,LPI)雷达作为一种具有强抗干扰能力及低截获特性的新型雷达,对其精准高效识别已成为雷达对抗一方波形识别的难点。针对该方向主流分类器卷积神经网络(convolution neural network,CNN)的结构...低截获概率(low probability of intercept,LPI)雷达作为一种具有强抗干扰能力及低截获特性的新型雷达,对其精准高效识别已成为雷达对抗一方波形识别的难点。针对该方向主流分类器卷积神经网络(convolution neural network,CNN)的结构智能寻优问题,提出一种基于粒子群优化(particle swarm optimization,PSO)算法-CNN的波形识别算法。该算法利用PSO的寻优特性,可实现较大范围内自动搭建不定层数、不定层类别及层内参数的CNN结构并进行迭代寻优;采用识别精度及网络复杂度相结合的衡量指标,可根据需求调整两者比重以实现对精度与轻量性的选择。该算法获取的CNN结构实现了比9种经典CNN结构更好的LPI雷达波形识别效果,同时避免了波形识别时人工选定CNN超参数缺乏智能性、客观性的问题,提高了选用CNN结构的适配性及高效性。展开更多
针对现有采用时频图结合深度学习模型对低截获概率(low probability of intercept,LPI)雷达信号识别的方法在开集场景下会失效的问题,提出一种基于互易点学习(reciprocal point learning,RPL)和阈值判断的雷达信号开集识别方法。通过RP...针对现有采用时频图结合深度学习模型对低截获概率(low probability of intercept,LPI)雷达信号识别的方法在开集场景下会失效的问题,提出一种基于互易点学习(reciprocal point learning,RPL)和阈值判断的雷达信号开集识别方法。通过RPL对特征空间进行优化,使已知类和未知类信号样本在特征空间中分布不同,最后确定合适的阈值进行开集识别。根据时频图的特点,在特征提取网络中加入注意力机制使网络更关注图像能量聚集的有效部分。实验结果表明,该方法在开放的电磁环境条件下具有良好的适应性。展开更多
为解决低截获概率(low probability of intercept,LPI)雷达信号中二相相移键控(binary phase shiftkeying,BPSK)信号、线性调频(linear frequency modulation,LFM)信号参数的估计问题,提出了基于信号积分包络的快速参数估计方法。应用...为解决低截获概率(low probability of intercept,LPI)雷达信号中二相相移键控(binary phase shiftkeying,BPSK)信号、线性调频(linear frequency modulation,LFM)信号参数的估计问题,提出了基于信号积分包络的快速参数估计方法。应用快速傅里叶变换(fast Fourier transform,FFT)进行雷达信号粗识别,对信号进行积分包络处理。对两类信号的积分包络值分别设置一定的门限,通过搜索BPSK信号积分包络的峰值即可迅速计算出BPSK信号的码速率。搜索LFM信号积分包络的零值点,按照给出的计算规则即可计算信号的调频斜率以及起始、终止频率。该算法具有快速、准确、容易实现的特点。仿真实验证明了该算法的有效性。同时,该算法具有一定的抗噪性。展开更多
基金support of the National Natural Science Foundation of China under grant numbers 62101570 and 61901494financial support has played a crucial role in the successful completion of this research.
文摘In this paper,we present a novel unimodular sequence design algorithm based on the coordinate descent(CD)algorithm,aimed at countering electronic surveillance(ES)systems based on cyclostationary analysis.Our algorithm not only provides resistance against cyclostationary analysis(CSA)but also maintains low integrated sidelobe(ISL)characteristics.Initially,we derive the expression of the cyclostationary feature(CSF)detector and simplify it into an iterative quadratic form.Additionally,we derive a quadratic form to ensure the similarity of the autocorrelation sidelobes.To balance the minimization of the detection probability and the ISL values,we introduce a Pareto scalar that transforms the multiobjective optimization problem into a convex combination of objective functions.This approach allows us to find an optimal trade-off between the two objectives.Finally,we propose a monotonic algorithm based on the CD algorithm to counter CSA analysis.This algorithm efficiently solves the optimization problem mentioned earlier.Numerical experiments are conducted to validate the correctness and effectiveness of our proposed algorithm.
基金supported by the National Natural Science Foundation of China(61571462)Weapons and Equipment Exploration Research Project(7131464)
文摘This paper proposes a desirable method to detect different kinds of low probability of intercept (LPI) radar signals, targeted at the main intra-pulse modulation method of LPI radar signals including the signals of linear frequency modulation, phase code, and frequency code. Firstly, it improves the coherent integration of LPI radar signals by adding the periodicity of the ambiguity function. Then, it develops a frequency domain detection method based on fast Fourier transform (FFT) and segmented autocorrelation function to detect signals without features of linear frequency modulation by virtue of the distribution characteristics of noise signals in the frequency domain. Finally, this paper gives a verification of the performance of the method for different signal-to-noise ratios by conducting simulation experiments, and compares the method with existing ones. Additionally, this method is characterized by the straightforward calculation and high real-time performance, which is conducive to better detecting all kinds of LPI radar signals.
文摘低截获概率(low probability of intercept,LPI)雷达已成为新时代雷达装备中关键的技术体制或工作模式,针对LPI雷达信号调制识别及参数估计方法的研究是当前雷达对抗侦察领域的热点。首先,分析了几种典型LPI雷达信号的脉内特征,梳理了LPI雷达信号调制识别及参数估计的传统和主流方法,并说明其原理、优缺点和研究现状。最后,总结了现有LPI雷达信号调制识别及参数估计方法尚存的问题,并指出其未来发展趋势,旨在为今后的研究提供参考。
文摘合理的雷达低截获(low probability of interception,LPI)性能评估方法是提高其隐身性能的基础。针对雷达LPI性能难以有效实时评估的问题,提出一种群广义直觉模糊软集(group-generalized intuitionistic fuzzy soft sets,G-GIFSS)算法与主客观权重相结合的雷达LPI性能评估方法。首先从反映雷达低截获性能的3个准则层信号层、功率层以及天线层确定6个目标属性指标层,选择直觉模糊集熵法确定客观权重、层次分析法((analytic hierarchy process,AHP))确定主观权重,并线性合成主客观权重。结合G-GIFSS算法利用多专家参量集的优势,对雷达LPI性能进行综合评判。通过案例分析并与经典评估方法对比,验证了该方法的优越性。
文摘针对LPI信号分类识别问题中,时频图像受噪声干扰严重的问题,提出了一种基于二维快速经验模式分解(FBEMD)的图像降噪算法,并利用该算法实现对LPI信号的分类。首先利用时频分析方法,获得待分类信号的时频分布图像;使用二维EMD分解算法对图像降噪;截取包含时频信息的图像部分,通过主分量分析法提取特征矢量;最后采用RBF神经网络完成信号的分类识别任务。对常见的LPI雷达信号进行仿真,结果表明较低信噪比情况下,该方法仍能获得较好的分类结果。当信噪比为-2 d B时,采用二维EMD降噪算法,平均正确识别率能够达到93%。
文摘为了优化单发多收协同雷达(single-transmitter multi-receiver cooperative radar,SMCR)探测系统的低截获概率(low probability of interception,LPI),利用SMCR目标探测的截获因子构造优化目标函数。首先,在二维平面上描述SMCR目标探测场景,分析探测区域内接收机队列的接收增益及其近似估计方法。然后,针对目标位置先验已知情况,建立SMCR系统的接收机队列优化模型,分析模型解集。最后,针对目标搜索区域先验已知情况,从多个维度仿真分析接收机队列的LPI特性。仿真结果表明,针对目标位置或目标搜索区域先验已知的SMCR探测场景,接收机队列的队形设计有利于改善系统的LPI性能。针对目标位置已知的实测数据定性说明了所提方法仿真结果的合理性。
文摘以二进制正交键控(binary orthogonal keying,BOK)为传统方法调制Chirp信号的检测手段日益丰富,针对常用时频分析手段分数阶傅里叶变换和短时傅里叶变换对Chirp信号的高破译性问题,提出了一种将信息映射到Chirp信号时域的新型调制方式,即时变信息映射(time varying-information mapping,TVIM)调制,通过构建时域携带信息的调制模式,解决了周期能量聚敛特性,增加了以BOK为检测思想的信息干扰,加强了波形的低截获概率(low probability of intercept,LPI)。以数学推导和仿真分析的方法,探究了新型调制方式的误码特性、时频分析下LPI特性及先验信息抗破译性。理论分析和仿真验证表明,TVIM调制架构下,可保证比特信噪比在13 dB前误码率达到10-4,并提高了波形LPI性能。
文摘针对低信噪比(signal to noise ratio,SNR)低截获概率(low probability of intercept,LPI)雷达脉内波形识别准确率低的问题,提出一种基于时频分析、压缩激励(squeeze excitation,SE)和ResNeXt网络的雷达辐射源信号识别方法。首先通过Choi-Williams分布(Choi-Williams distribution,CWD)获得雷达时域信号的二维时频图像(time-frequency image,TFI);然后进行TFI预处理降低噪声干扰和频率维的位置分布差异,以适应深度学习网络输入;最后在ResNeXt基础上加入扩张卷积和SE结构提取TFI特征,实现雷达辐射源分类。实验结果表明,SNR低至-8 dB时,该方法对12类常见LPI雷达波形的整体识别准确率依然能达到98.08%。
文摘低截获概率(low probability of intercept,LPI)雷达作为一种具有强抗干扰能力及低截获特性的新型雷达,对其精准高效识别已成为雷达对抗一方波形识别的难点。针对该方向主流分类器卷积神经网络(convolution neural network,CNN)的结构智能寻优问题,提出一种基于粒子群优化(particle swarm optimization,PSO)算法-CNN的波形识别算法。该算法利用PSO的寻优特性,可实现较大范围内自动搭建不定层数、不定层类别及层内参数的CNN结构并进行迭代寻优;采用识别精度及网络复杂度相结合的衡量指标,可根据需求调整两者比重以实现对精度与轻量性的选择。该算法获取的CNN结构实现了比9种经典CNN结构更好的LPI雷达波形识别效果,同时避免了波形识别时人工选定CNN超参数缺乏智能性、客观性的问题,提高了选用CNN结构的适配性及高效性。
文摘针对现有采用时频图结合深度学习模型对低截获概率(low probability of intercept,LPI)雷达信号识别的方法在开集场景下会失效的问题,提出一种基于互易点学习(reciprocal point learning,RPL)和阈值判断的雷达信号开集识别方法。通过RPL对特征空间进行优化,使已知类和未知类信号样本在特征空间中分布不同,最后确定合适的阈值进行开集识别。根据时频图的特点,在特征提取网络中加入注意力机制使网络更关注图像能量聚集的有效部分。实验结果表明,该方法在开放的电磁环境条件下具有良好的适应性。
文摘为解决低截获概率(low probability of intercept,LPI)雷达信号中二相相移键控(binary phase shiftkeying,BPSK)信号、线性调频(linear frequency modulation,LFM)信号参数的估计问题,提出了基于信号积分包络的快速参数估计方法。应用快速傅里叶变换(fast Fourier transform,FFT)进行雷达信号粗识别,对信号进行积分包络处理。对两类信号的积分包络值分别设置一定的门限,通过搜索BPSK信号积分包络的峰值即可迅速计算出BPSK信号的码速率。搜索LFM信号积分包络的零值点,按照给出的计算规则即可计算信号的调频斜率以及起始、终止频率。该算法具有快速、准确、容易实现的特点。仿真实验证明了该算法的有效性。同时,该算法具有一定的抗噪性。
