This paper proposes a novel open set recognition method,the Spatial Distribution Feature Extraction Network(SDFEN),to address the problem of electromagnetic signal recognition in an open environment.The spatial distri...This paper proposes a novel open set recognition method,the Spatial Distribution Feature Extraction Network(SDFEN),to address the problem of electromagnetic signal recognition in an open environment.The spatial distribution feature extraction layer in SDFEN replaces convolutional output neural networks with the spatial distribution features that focus more on inter-sample information by incorporating class center vectors.The designed hybrid loss function considers both intra-class distance and inter-class distance,thereby enhancing the similarity among samples of the same class and increasing the dissimilarity between samples of different classes during training.Consequently,this method allows unknown classes to occupy a larger space in the feature space.This reduces the possibility of overlap with known class samples and makes the boundaries between known and unknown samples more distinct.Additionally,the feature comparator threshold can be used to reject unknown samples.For signal open set recognition,seven methods,including the proposed method,are applied to two kinds of electromagnetic signal data:modulation signal and real-world emitter.The experimental results demonstrate that the proposed method outperforms the other six methods overall in a simulated open environment.Specifically,compared to the state-of-the-art Openmax method,the novel method achieves up to 8.87%and 5.25%higher micro-F-measures,respectively.展开更多
Generative adversarial network(GAN)has achieved great success in many fields such as computer vision,speech processing,and natural language processing,because of its powerful capabilities for generating realistic samp...Generative adversarial network(GAN)has achieved great success in many fields such as computer vision,speech processing,and natural language processing,because of its powerful capabilities for generating realistic samples.In this paper,we introduce GAN into the field of electromagnetic signal classification(ESC).ESC plays an important role in both military and civilian domains.However,in many specific scenarios,we can’t obtain enough labeled data,which cause failure of deep learning methods because they are easy to fall into over-fitting.Fortunately,semi-supervised learning(SSL)can leverage the large amount of unlabeled data to enhance the classification performance of classifiers,especially in scenarios with limited amount of labeled data.We present an SSL framework by incorporating GAN,which can directly process the raw in-phase and quadrature(IQ)signal data.According to the characteristics of the electromagnetic signal,we propose a weighted loss function,leading to an effective classifier to realize the end-to-end classification of the electromagnetic signal.We validate the proposed method on both public RML2016.04c dataset and real-world Aircraft Communications Addressing and Reporting System(ACARS)signal dataset.Extensive experimental results show that the proposed framework obtains a significant increase in classification accuracy compared with the state-of-the-art studies.展开更多
Electromagnetic signals may be a promising precursor to seismic activity which has been observed in many case studies in past decades.However,the correlation and causation between the electromagnetic signals and the s...Electromagnetic signals may be a promising precursor to seismic activity which has been observed in many case studies in past decades.However,the correlation and causation between the electromagnetic signals and the seismic activity are still unclear without intensive observation network.In order to find seismoelectromagnetic phenomenon,we deployed AETA(acoustic and electromagnetic testing all-in-one system),a high-density multi-component seismic monitoring system in the China Earthquake Science Experiment site(CESE,in Sichuan Province and Yunnan Province,China)and the capital circle(areas with a distance which is≤200 km from Beijing),to record electromagnetic and geo-acoustic data across 0.1 Hz−10 kHz.In the course of data collection,we discovered an electromagnetic waveform that occurs on a daily basis.Because the signal generally coincides with sunrise and sunset,we named this phenomenon the SRSS(Sunrise-Sunset)waveform.After conducting three statistical tests based on seismicity and SRSS,we determined that the SRSS waveform is roughly correlated with the onset of seismic activity.It generally occurs at the regions where seismicity occurs.This discovery might have significant implications with respect to the future of earthquake prediction.展开更多
The idea of the earthquake (EQ) focus as a coherent electromagnetic (EM) emitter is suggested. This idea elucidates enigmatic properties of the EM voice of the focus: its emission is not continuous, occurring periodic...The idea of the earthquake (EQ) focus as a coherent electromagnetic (EM) emitter is suggested. This idea elucidates enigmatic properties of the EM voice of the focus: its emission is not continuous, occurring periodically in flashes, which are structured as the pulses occurring in bursts;the EM activity increases starting approximately two weeks before the EQ and becomes very weak or completely disappears one day before the EQ (gap of silence). The mechanism of coherency starts with electric discharges of any mini-cracks as a mini-capacitor, which generates EM waves;the latter induces discharges of other cracks, multiplying the amplitude of the wave and creating the pulse of seismic EM signal. It is an avalanche-like mechanism of coherency, which transforms even weak EM signals into intensive EM seismic flashes.展开更多
Deep learning has been fully verified and accepted in the field of electromagnetic signal classification. However, in many specific scenarios, such as radio resource management for aircraft communications, labeled dat...Deep learning has been fully verified and accepted in the field of electromagnetic signal classification. However, in many specific scenarios, such as radio resource management for aircraft communications, labeled data are difficult to obtain, which makes the best deep learning methods at present seem almost powerless, because these methods need a large amount of labeled data for training. When the training dataset is small, it is highly possible to fall into overfitting, which causes performance degradation of the deep neural network. For few-shot electromagnetic signal classification, data augmentation is one of the most intuitive countermeasures. In this work, a generative adversarial network based on the data augmentation method is proposed to achieve better classification performance for electromagnetic signals. Based on the similarity principle, a screening mechanism is established to obtain high-quality generated signals. Then, a data union augmentation algorithm is designed by introducing spatiotemporally flipped shapes of the signal. To verify the effectiveness of the proposed data augmentation algorithm, experiments are conducted on the RADIOML 2016.04C dataset and real-world ACARS dataset. The experimental results show that the proposed method significantly improves the performance of few-shot electromagnetic signal classification.展开更多
In order to solve the problem of low signal-to-noise ratio（about 15 d B） in magnetic signal acquisition of banknotes, a new method of magnetic signal acquisition and processing is proposed taking RMB as an example. ...In order to solve the problem of low signal-to-noise ratio（about 15 d B） in magnetic signal acquisition of banknotes, a new method of magnetic signal acquisition and processing is proposed taking RMB as an example. In this method, weak signa detection is performed to reduce the noise accompanied with the signal. Seven orders of Chebyshev（Ⅰ） filter and the anti-jamming technology are used in the PCB layout, and grounding modes are introduced to reduce the noise of the amplitude waveform. The proposed method reduce the final output noise by 2/3 and the sig nal-to-noise ratio is increased to 24 d B. The experimental results show that the magnetic signal of RMB banknotes are acquired by the circuit stability, which provides an important guarantee for the improvements of the anti-counterfeit and discrimination for banknotes performance.展开更多
A uniform array of scalar-sensors with intersensor spacings over a large aperture size generally offers enhanced resolution and source localization accuracy,but it may also lead to cyclic ambiguity.By exploiting the p...A uniform array of scalar-sensors with intersensor spacings over a large aperture size generally offers enhanced resolution and source localization accuracy,but it may also lead to cyclic ambiguity.By exploiting the polarization information of impinging waves,an electromagnetic vector-sensor array outperforms the unpolarized scalar-sensor array in resolving this cyclic ambiguity.However,the electromagnetic vector-sensor array usually consists of cocentered orthogonal loops and dipoles（COLD）,which is easily subjected to mutual coupling across these cocentered dipoles/loops.As a result,the source localization performance of the COLD array may substantially degrade rather than being improved.This paper proposes a new source localization method with a non-cocentered orthogonal loop and dipole（NCOLD）array.The NCOLD array contains only one dipole or loop on each array grid,and the intersensor spacings are larger than a half-wavelength.Therefore,unlike the COLD array,these well separated dipoles/loops minimize the mutual coupling effects and extend the spatial aperture as well.With the NCOLD array,the proposed method can effciently exploit the polarization information to offer high localization precision.展开更多
文摘This paper proposes a novel open set recognition method,the Spatial Distribution Feature Extraction Network(SDFEN),to address the problem of electromagnetic signal recognition in an open environment.The spatial distribution feature extraction layer in SDFEN replaces convolutional output neural networks with the spatial distribution features that focus more on inter-sample information by incorporating class center vectors.The designed hybrid loss function considers both intra-class distance and inter-class distance,thereby enhancing the similarity among samples of the same class and increasing the dissimilarity between samples of different classes during training.Consequently,this method allows unknown classes to occupy a larger space in the feature space.This reduces the possibility of overlap with known class samples and makes the boundaries between known and unknown samples more distinct.Additionally,the feature comparator threshold can be used to reject unknown samples.For signal open set recognition,seven methods,including the proposed method,are applied to two kinds of electromagnetic signal data:modulation signal and real-world emitter.The experimental results demonstrate that the proposed method outperforms the other six methods overall in a simulated open environment.Specifically,compared to the state-of-the-art Openmax method,the novel method achieves up to 8.87%and 5.25%higher micro-F-measures,respectively.
基金the National Natural Science Foundation of China(Nos.61771380,U19B2015,U1730109).
文摘Generative adversarial network(GAN)has achieved great success in many fields such as computer vision,speech processing,and natural language processing,because of its powerful capabilities for generating realistic samples.In this paper,we introduce GAN into the field of electromagnetic signal classification(ESC).ESC plays an important role in both military and civilian domains.However,in many specific scenarios,we can’t obtain enough labeled data,which cause failure of deep learning methods because they are easy to fall into over-fitting.Fortunately,semi-supervised learning(SSL)can leverage the large amount of unlabeled data to enhance the classification performance of classifiers,especially in scenarios with limited amount of labeled data.We present an SSL framework by incorporating GAN,which can directly process the raw in-phase and quadrature(IQ)signal data.According to the characteristics of the electromagnetic signal,we propose a weighted loss function,leading to an effective classifier to realize the end-to-end classification of the electromagnetic signal.We validate the proposed method on both public RML2016.04c dataset and real-world Aircraft Communications Addressing and Reporting System(ACARS)signal dataset.Extensive experimental results show that the proposed framework obtains a significant increase in classification accuracy compared with the state-of-the-art studies.
基金Projects(KJYY20170721151955849,JCYJ20190808161401653)supported by Fundamental Research Grant from Shenzhen Science&Technology,China。
文摘Electromagnetic signals may be a promising precursor to seismic activity which has been observed in many case studies in past decades.However,the correlation and causation between the electromagnetic signals and the seismic activity are still unclear without intensive observation network.In order to find seismoelectromagnetic phenomenon,we deployed AETA(acoustic and electromagnetic testing all-in-one system),a high-density multi-component seismic monitoring system in the China Earthquake Science Experiment site(CESE,in Sichuan Province and Yunnan Province,China)and the capital circle(areas with a distance which is≤200 km from Beijing),to record electromagnetic and geo-acoustic data across 0.1 Hz−10 kHz.In the course of data collection,we discovered an electromagnetic waveform that occurs on a daily basis.Because the signal generally coincides with sunrise and sunset,we named this phenomenon the SRSS(Sunrise-Sunset)waveform.After conducting three statistical tests based on seismicity and SRSS,we determined that the SRSS waveform is roughly correlated with the onset of seismic activity.It generally occurs at the regions where seismicity occurs.This discovery might have significant implications with respect to the future of earthquake prediction.
文摘The idea of the earthquake (EQ) focus as a coherent electromagnetic (EM) emitter is suggested. This idea elucidates enigmatic properties of the EM voice of the focus: its emission is not continuous, occurring periodically in flashes, which are structured as the pulses occurring in bursts;the EM activity increases starting approximately two weeks before the EQ and becomes very weak or completely disappears one day before the EQ (gap of silence). The mechanism of coherency starts with electric discharges of any mini-cracks as a mini-capacitor, which generates EM waves;the latter induces discharges of other cracks, multiplying the amplitude of the wave and creating the pulse of seismic EM signal. It is an avalanche-like mechanism of coherency, which transforms even weak EM signals into intensive EM seismic flashes.
基金supported by the National Natural Science Foundation of China(No.61772401)the Fundamental Research Funds for the Central Universities(No.RW180177)supported by the Science and Technology on Communication Information Security Control Laboratory。
文摘Deep learning has been fully verified and accepted in the field of electromagnetic signal classification. However, in many specific scenarios, such as radio resource management for aircraft communications, labeled data are difficult to obtain, which makes the best deep learning methods at present seem almost powerless, because these methods need a large amount of labeled data for training. When the training dataset is small, it is highly possible to fall into overfitting, which causes performance degradation of the deep neural network. For few-shot electromagnetic signal classification, data augmentation is one of the most intuitive countermeasures. In this work, a generative adversarial network based on the data augmentation method is proposed to achieve better classification performance for electromagnetic signals. Based on the similarity principle, a screening mechanism is established to obtain high-quality generated signals. Then, a data union augmentation algorithm is designed by introducing spatiotemporally flipped shapes of the signal. To verify the effectiveness of the proposed data augmentation algorithm, experiments are conducted on the RADIOML 2016.04C dataset and real-world ACARS dataset. The experimental results show that the proposed method significantly improves the performance of few-shot electromagnetic signal classification.
基金Supported by the project of image recognition and control system in class A machine(HT201403)
文摘In order to solve the problem of low signal-to-noise ratio（about 15 d B） in magnetic signal acquisition of banknotes, a new method of magnetic signal acquisition and processing is proposed taking RMB as an example. In this method, weak signa detection is performed to reduce the noise accompanied with the signal. Seven orders of Chebyshev（Ⅰ） filter and the anti-jamming technology are used in the PCB layout, and grounding modes are introduced to reduce the noise of the amplitude waveform. The proposed method reduce the final output noise by 2/3 and the sig nal-to-noise ratio is increased to 24 d B. The experimental results show that the magnetic signal of RMB banknotes are acquired by the circuit stability, which provides an important guarantee for the improvements of the anti-counterfeit and discrimination for banknotes performance.
基金supported by the Scientifc Research Fund of Zhejiang Provincial Education Department(No.Y201225848)the Scientifc and Technological Innovation Programs of Higher Education Institutions in Shanxi(No.2013124)
文摘A uniform array of scalar-sensors with intersensor spacings over a large aperture size generally offers enhanced resolution and source localization accuracy,but it may also lead to cyclic ambiguity.By exploiting the polarization information of impinging waves,an electromagnetic vector-sensor array outperforms the unpolarized scalar-sensor array in resolving this cyclic ambiguity.However,the electromagnetic vector-sensor array usually consists of cocentered orthogonal loops and dipoles（COLD）,which is easily subjected to mutual coupling across these cocentered dipoles/loops.As a result,the source localization performance of the COLD array may substantially degrade rather than being improved.This paper proposes a new source localization method with a non-cocentered orthogonal loop and dipole（NCOLD）array.The NCOLD array contains only one dipole or loop on each array grid,and the intersensor spacings are larger than a half-wavelength.Therefore,unlike the COLD array,these well separated dipoles/loops minimize the mutual coupling effects and extend the spatial aperture as well.With the NCOLD array,the proposed method can effciently exploit the polarization information to offer high localization precision.
