To generate high-frequency radio frequency(RF) vector signals, a vector signal generation method by optical frequency sextupling using a dual-parallel modulator is proposed. The method modulates vector signal on +3 rd...To generate high-frequency radio frequency(RF) vector signals, a vector signal generation method by optical frequency sextupling using a dual-parallel modulator is proposed. The method modulates vector signal on +3 rd order optical sideband and local oscillator(LO) signal on-3 rd order sideband using the intermodulation process in the DPMZM. After suppressing of the optical carrier and other sidebands through proper adjustment for modulator biases and modulation index, a frequency sextupled millimeter-wave vector signal can be generated after photodetection. The frequency sextupling will lower the bandwidth of the modulator, the local oscillator and the driving circuits. In addition, the phase of generated signal is not distorted after detection, and the power fading after fiber transmission can be avoided. In the simulation, a 500-MSym/s QPSK signal at 60 GHz is generated by 10-GHz drive signal. After travelling over fiber with length of 20/30/40-km, receiver power penalty keeps below 2.5 dB.展开更多
In this paper, a classification method based on Support Vector Machine (SVM) is given in the digital modulation signal classification. The second, fourth and sixth order cumulants of the received signals are used as c...In this paper, a classification method based on Support Vector Machine (SVM) is given in the digital modulation signal classification. The second, fourth and sixth order cumulants of the received signals are used as classification vectors firstly, then the kernel thought is used to map the feature vector to the high dimensional feature space and the optimum separating hyperplane is constructed in space to realize signal recognition. In order to build an effective and robust SVM classifier, the radial basis kernel function is selected, one against one or one against rest of multi-class classifier is designed, and method of parameter selection using cross- validation grid is adopted. Through the experiments it can be concluded that the classifier based on SVM has high performance and is more robust.展开更多
The existing direction of arrival （DOA） estimation algorithms based on the electromagnetic vector sensors array barely deal with the coexisting of independent and coherent signals. A two-dimensional direction findin...The existing direction of arrival （DOA） estimation algorithms based on the electromagnetic vector sensors array barely deal with the coexisting of independent and coherent signals. A two-dimensional direction finding method using an L-shape electromagnetic vector sensors array is proposed. According to this method, the DOAs of the independent signals and the coherent signals are estimated separately, so that the array aperture can be exploited sufficiently. Firstly, the DOAs of the independent signals are estimated by the estimation of signal parameters via rotational invariance techniques, and the influence of the co- herent signals can be eliminated by utilizing the property of the coherent signals. Then the data covariance matrix containing the information of the coherent signals only is obtained by exploiting the Toeplitz property of the independent signals, and an improved polarimetric angular smoothing technique is proposed to de-correlate the coherent signals. This new method is more practical in actual signal environment than common DOA estimation algorithms and can expand the array aperture. Simulation results are presented to show the estimating performance of the proposed method.展开更多
This paper presents a novel method for radar emitter signal recognition. First, wavelet packet transform （WPT） is introduced to extract features from radar emitter signals. Then, rough set theory is used to select t...This paper presents a novel method for radar emitter signal recognition. First, wavelet packet transform （WPT） is introduced to extract features from radar emitter signals. Then, rough set theory is used to select the optimal feature subset with good discriminability from original feature set, and support vector machines （SVMs） are employed to design classifiers. A large number of experimental results show that the proposed method achieves very high recognition rates for 9 radar emitter signals in a wide range of signal-to-noise rates, and proves a feasible and valid method.展开更多
将不同长度线性调频(Linear Frequency Modulation, LFM)信号装入矢量信号源,其载波频率设置为中频,分别采样其输出的不同长度的LFM信号,对输出的LFM信号进行脉冲压缩Matlab仿真处理。实验表明,在采样频率及输出幅度相同的条件下,102.00...将不同长度线性调频(Linear Frequency Modulation, LFM)信号装入矢量信号源,其载波频率设置为中频,分别采样其输出的不同长度的LFM信号,对输出的LFM信号进行脉冲压缩Matlab仿真处理。实验表明,在采样频率及输出幅度相同的条件下,102.00μs脉冲内LFM信号仿真处理后的信噪比和旁瓣抑制比,相较3.00μs的指标提高了13.35 dB和4.61dB。进一步说明了由组合脉冲组成的LFM发射波形信号,其脉压体制雷达存在近距离回波比远距离回波差问题。提出了一种复杂海洋环境下多目标探测方法,对海用脉压体制雷达真实回波性能进行了改善和提升。展开更多
目前的脑电(EEG)情感识别模型忽略了不同时段情感状态的差异性,未能强化关键的情感信息。针对上述问题,提出一种多上下文向量优化的卷积递归神经网络(CR-MCV)。首先构造脑电信号的特征矩阵序列,通过卷积神经网络(CNN)学习多通道脑电的...目前的脑电(EEG)情感识别模型忽略了不同时段情感状态的差异性,未能强化关键的情感信息。针对上述问题,提出一种多上下文向量优化的卷积递归神经网络(CR-MCV)。首先构造脑电信号的特征矩阵序列,通过卷积神经网络(CNN)学习多通道脑电的空间特征;然后利用基于多头注意力的递归神经网络生成多上下文向量进行高层抽象特征提取;最后利用全连接层进行情感分类。在DEAP(Database for Emotion Analysis using Physiological signals)数据集上进行实验,CR-MCV在唤醒和效价维度上分类准确率分别为88.09%和89.30%。实验结果表明,CR-MCV在利用电极空间位置信息和不同时段情感状态显著性特征基础上,能够自适应地分配特征的注意力并强化情感状态显著性信息。展开更多
为了减小低快拍数和低信噪比下采样协方差矩阵误差,并降低其运算复杂度,提出了一种基于实数化的均匀圆阵采样协方差矩阵重构方法。针对均匀圆阵的特点,通过组建特殊的基向量,构成特殊的重构矩阵。通过将采样协方差矩阵实数化,进一步降...为了减小低快拍数和低信噪比下采样协方差矩阵误差,并降低其运算复杂度,提出了一种基于实数化的均匀圆阵采样协方差矩阵重构方法。针对均匀圆阵的特点,通过组建特殊的基向量,构成特殊的重构矩阵。通过将采样协方差矩阵实数化,进一步降低了重构矩阵的复杂度。考虑到多通道不一致性对重构矩阵的影响,引入0位校正算法,提高了重构方法的稳健性。最后应用重构后的协方差矩阵进行子空间类波达方向估计(direction of arrival,DOA)。实验仿真证明,该特殊重构矩阵在实数化下与原矩阵重构能力相同;当快拍数为100、信噪比为0 dB时,双信源分辨力较重构前由74%提高到95%以上;理论重构运算复杂度降低到原来的53.99%。展开更多
基金Sponsored by the Programme of Introducing Talents of Discipline to Universities(Grant No.B08038)
文摘To generate high-frequency radio frequency(RF) vector signals, a vector signal generation method by optical frequency sextupling using a dual-parallel modulator is proposed. The method modulates vector signal on +3 rd order optical sideband and local oscillator(LO) signal on-3 rd order sideband using the intermodulation process in the DPMZM. After suppressing of the optical carrier and other sidebands through proper adjustment for modulator biases and modulation index, a frequency sextupled millimeter-wave vector signal can be generated after photodetection. The frequency sextupling will lower the bandwidth of the modulator, the local oscillator and the driving circuits. In addition, the phase of generated signal is not distorted after detection, and the power fading after fiber transmission can be avoided. In the simulation, a 500-MSym/s QPSK signal at 60 GHz is generated by 10-GHz drive signal. After travelling over fiber with length of 20/30/40-km, receiver power penalty keeps below 2.5 dB.
文摘In this paper, a classification method based on Support Vector Machine (SVM) is given in the digital modulation signal classification. The second, fourth and sixth order cumulants of the received signals are used as classification vectors firstly, then the kernel thought is used to map the feature vector to the high dimensional feature space and the optimum separating hyperplane is constructed in space to realize signal recognition. In order to build an effective and robust SVM classifier, the radial basis kernel function is selected, one against one or one against rest of multi-class classifier is designed, and method of parameter selection using cross- validation grid is adopted. Through the experiments it can be concluded that the classifier based on SVM has high performance and is more robust.
基金supported by the National Natural Science Foundation of China (61102106)the Fundamental Research Funds for the Central Universities (HEUCF1208 HEUCF100801)
文摘The existing direction of arrival （DOA） estimation algorithms based on the electromagnetic vector sensors array barely deal with the coexisting of independent and coherent signals. A two-dimensional direction finding method using an L-shape electromagnetic vector sensors array is proposed. According to this method, the DOAs of the independent signals and the coherent signals are estimated separately, so that the array aperture can be exploited sufficiently. Firstly, the DOAs of the independent signals are estimated by the estimation of signal parameters via rotational invariance techniques, and the influence of the co- herent signals can be eliminated by utilizing the property of the coherent signals. Then the data covariance matrix containing the information of the coherent signals only is obtained by exploiting the Toeplitz property of the independent signals, and an improved polarimetric angular smoothing technique is proposed to de-correlate the coherent signals. This new method is more practical in actual signal environment than common DOA estimation algorithms and can expand the array aperture. Simulation results are presented to show the estimating performance of the proposed method.
文摘This paper presents a novel method for radar emitter signal recognition. First, wavelet packet transform （WPT） is introduced to extract features from radar emitter signals. Then, rough set theory is used to select the optimal feature subset with good discriminability from original feature set, and support vector machines （SVMs） are employed to design classifiers. A large number of experimental results show that the proposed method achieves very high recognition rates for 9 radar emitter signals in a wide range of signal-to-noise rates, and proves a feasible and valid method.
文摘将不同长度线性调频(Linear Frequency Modulation, LFM)信号装入矢量信号源,其载波频率设置为中频,分别采样其输出的不同长度的LFM信号,对输出的LFM信号进行脉冲压缩Matlab仿真处理。实验表明,在采样频率及输出幅度相同的条件下,102.00μs脉冲内LFM信号仿真处理后的信噪比和旁瓣抑制比,相较3.00μs的指标提高了13.35 dB和4.61dB。进一步说明了由组合脉冲组成的LFM发射波形信号,其脉压体制雷达存在近距离回波比远距离回波差问题。提出了一种复杂海洋环境下多目标探测方法,对海用脉压体制雷达真实回波性能进行了改善和提升。
文摘目前的脑电(EEG)情感识别模型忽略了不同时段情感状态的差异性,未能强化关键的情感信息。针对上述问题,提出一种多上下文向量优化的卷积递归神经网络(CR-MCV)。首先构造脑电信号的特征矩阵序列,通过卷积神经网络(CNN)学习多通道脑电的空间特征;然后利用基于多头注意力的递归神经网络生成多上下文向量进行高层抽象特征提取;最后利用全连接层进行情感分类。在DEAP(Database for Emotion Analysis using Physiological signals)数据集上进行实验,CR-MCV在唤醒和效价维度上分类准确率分别为88.09%和89.30%。实验结果表明,CR-MCV在利用电极空间位置信息和不同时段情感状态显著性特征基础上,能够自适应地分配特征的注意力并强化情感状态显著性信息。
文摘为了减小低快拍数和低信噪比下采样协方差矩阵误差,并降低其运算复杂度,提出了一种基于实数化的均匀圆阵采样协方差矩阵重构方法。针对均匀圆阵的特点,通过组建特殊的基向量,构成特殊的重构矩阵。通过将采样协方差矩阵实数化,进一步降低了重构矩阵的复杂度。考虑到多通道不一致性对重构矩阵的影响,引入0位校正算法,提高了重构方法的稳健性。最后应用重构后的协方差矩阵进行子空间类波达方向估计(direction of arrival,DOA)。实验仿真证明,该特殊重构矩阵在实数化下与原矩阵重构能力相同;当快拍数为100、信噪比为0 dB时,双信源分辨力较重构前由74%提高到95%以上;理论重构运算复杂度降低到原来的53.99%。
