Editor's Note
Guest Editor
Article List
2023, 21(6):713-724.
DOI: 10.11805/TKYDA2022180
Abstract:
In the electromagnetic space,the physical domain and information domain cannot be effectively integrated and the application of electromagnetic information lacks intelligence. A concept of intelligent electromagnetic space is put forward. The intelligent electromagnetic space should bear the ability of representation, understanding, predicting and decision making. Taking the model as the core, taking the software as the carrier, with highly data-driven, it can achieve accurate mapping of the physical world and provide intelligent decision-making services for the combat command and troop operations. According to the organization mechanism of electromagnetic space information, the problem of multi-source heterogeneous physics—information data fusion is further solved, and the technical framework is constructed. Some key technologies and implementation ideas are proposed,including Geographic Coordinate Subdividing Grid with One Dimension Integral Coding on 2n-Tree-N Dimensions(GEOSOT-ND) high-dimensional tensor modeling and computation, multi-domain electromagnetic space sensing and efficient storage, full-dimensional electromagnetic situation reconstruction and intelligent deduction, and grid data-driven application methods.
In the electromagnetic space,the physical domain and information domain cannot be effectively integrated and the application of electromagnetic information lacks intelligence. A concept of intelligent electromagnetic space is put forward. The intelligent electromagnetic space should bear the ability of representation, understanding, predicting and decision making. Taking the model as the core, taking the software as the carrier, with highly data-driven, it can achieve accurate mapping of the physical world and provide intelligent decision-making services for the combat command and troop operations. According to the organization mechanism of electromagnetic space information, the problem of multi-source heterogeneous physics—information data fusion is further solved, and the technical framework is constructed. Some key technologies and implementation ideas are proposed,including Geographic Coordinate Subdividing Grid with One Dimension Integral Coding on 2n-Tree-N Dimensions(GEOSOT-ND) high-dimensional tensor modeling and computation, multi-domain electromagnetic space sensing and efficient storage, full-dimensional electromagnetic situation reconstruction and intelligent deduction, and grid data-driven application methods.
2023, 21(6):751-758.
DOI: 10.11805/TKYDA2023063
Abstract:
With the increasing complexity of the electromagnetic signal environment and the increasing number of communication devices, the interference with electromagnetic signals is gradually increasing. Therefore, the study on signal reception and processing techniques in different noise environments and the use of signal data indicators and the information they carry in complex electromagnetic environments is very critical. In order to understand the performance of noisy signals in different electromagnetic environments and improve the quality and reliability of signal utilization, a time series decomposition-based electromagnetic data processing method is proposed. A noisy signal processing model is established based on additive seasonal time series decomposition, and the model is also employed to analyze and evaluate the performance of signals in noisy environments with regularity, trend, BER, etc., and to data-mine the original information and carrier information. Compared with the traditional methods, the proposed time series decomposition-based signal mining and prediction model is more accurate for signal prediction in noisy environment.
With the increasing complexity of the electromagnetic signal environment and the increasing number of communication devices, the interference with electromagnetic signals is gradually increasing. Therefore, the study on signal reception and processing techniques in different noise environments and the use of signal data indicators and the information they carry in complex electromagnetic environments is very critical. In order to understand the performance of noisy signals in different electromagnetic environments and improve the quality and reliability of signal utilization, a time series decomposition-based electromagnetic data processing method is proposed. A noisy signal processing model is established based on additive seasonal time series decomposition, and the model is also employed to analyze and evaluate the performance of signals in noisy environments with regularity, trend, BER, etc., and to data-mine the original information and carrier information. Compared with the traditional methods, the proposed time series decomposition-based signal mining and prediction model is more accurate for signal prediction in noisy environment.
2023, 21(6):734-744.
DOI: 10.11805/TKYDA2022243
Abstract:
Due to the diversity of RF signals, the complexity of the electromagnetic environment, and the difficulty of feature extraction, the robustness and applicability of the existing artificial features-based RF-specific emitter identification methods cannot meet the application requirements. Although the data-driven deep learning methods can provide a more flexible mode of specific emitter identification, they are less interpretable and lack a general test mode to evaluate their advantages and disadvantages. An evaluation method is explored for the deep learning model on the target individual dataset of the Electromagnetic Big Data Super Contest, and a general testing system architecture is proposed for the specific emitter identification model based on deep neural networks. The framework constructs the simulation test samples through signal feature masking, Generative Adversarial Network (GAN), deception signal collection, channel simulation and other methods, and imports the test samples and original data into the deep model to compare the recognition results. The test results are employed to judge the location of the signal key features extracted by the deep model, to analyze the robustness of the model, and to reveal the impact of the channel environment on the recognition performance, thus the performance of the deep learning model can be interpretable.
Due to the diversity of RF signals, the complexity of the electromagnetic environment, and the difficulty of feature extraction, the robustness and applicability of the existing artificial features-based RF-specific emitter identification methods cannot meet the application requirements. Although the data-driven deep learning methods can provide a more flexible mode of specific emitter identification, they are less interpretable and lack a general test mode to evaluate their advantages and disadvantages. An evaluation method is explored for the deep learning model on the target individual dataset of the Electromagnetic Big Data Super Contest, and a general testing system architecture is proposed for the specific emitter identification model based on deep neural networks. The framework constructs the simulation test samples through signal feature masking, Generative Adversarial Network (GAN), deception signal collection, channel simulation and other methods, and imports the test samples and original data into the deep model to compare the recognition results. The test results are employed to judge the location of the signal key features extracted by the deep model, to analyze the robustness of the model, and to reveal the impact of the channel environment on the recognition performance, thus the performance of the deep learning model can be interpretable.
2023, 21(6):725-733.
DOI: 10.11805/TKYDA2022210
Abstract:
There exist the following problems in the available methods for Direct Position Determination(DPD) of Non-Circular(NC) sources with multiple base stations: high computational complexity of spectral peak search, sensitivity to the location of base stations, and lack of consideration of the difference in signal loss when propagating in space, which leads to unstable performance. A DPD method is proposed for NC sources: Joint Reduced Dimension Propagator Method and Taylor Compensation(JRT-PM). First, the array aperture is expanded according to the elliptic covariance information of NC signals. Then, the NC phase search dimension is eliminated by the dimension reduction method for rough estimation to reduce the computational complexity. Next, the Taylor compensation is combined with the information of all base stations to improve the estimation performance of the algorithm. Simulation experiments show that compared with the traditional two-step localization algorithm by Angle Of Arrival K-means clustering(AOA-clustering), Minimum Variance Distortionless Response(MVDR) DPD method and Subspace Data Fusion(SDF) DPD method, the proposed algorithm can estimate more targets while improving the localization accuracy. Compared with Non-Circular Propagator Method(NC-PM) DPD method, the proposed algorithm significantly reduces the computational complexity while ensuring the estimation performance.
There exist the following problems in the available methods for Direct Position Determination(DPD) of Non-Circular(NC) sources with multiple base stations: high computational complexity of spectral peak search, sensitivity to the location of base stations, and lack of consideration of the difference in signal loss when propagating in space, which leads to unstable performance. A DPD method is proposed for NC sources: Joint Reduced Dimension Propagator Method and Taylor Compensation(JRT-PM). First, the array aperture is expanded according to the elliptic covariance information of NC signals. Then, the NC phase search dimension is eliminated by the dimension reduction method for rough estimation to reduce the computational complexity. Next, the Taylor compensation is combined with the information of all base stations to improve the estimation performance of the algorithm. Simulation experiments show that compared with the traditional two-step localization algorithm by Angle Of Arrival K-means clustering(AOA-clustering), Minimum Variance Distortionless Response(MVDR) DPD method and Subspace Data Fusion(SDF) DPD method, the proposed algorithm can estimate more targets while improving the localization accuracy. Compared with Non-Circular Propagator Method(NC-PM) DPD method, the proposed algorithm significantly reduces the computational complexity while ensuring the estimation performance.
2023, 21(6):703-712.
DOI: 10.11805/TKYDA2022178
Abstract:
The electromagnetic spectrum is the sixth dimensional battle space after land, sea, air, space and cyber, and runs through the other five dimensional battles. In order to effectively deal with the challenges and threats of the electromagnetic spectrum, the US Department of Defense puts forward the concept of electromagnetic spectrum operations, which has a profound impact on future combat patterns and even war forms. From the perspective of electromagnetic spectrum application, the application status and representative work of the US Department of Defense on electromagnetic spectrum are systematically introduced. Firstly, the traditional military applications of the electromagnetic spectrum are systematically sorted out, such as communications, radar, signals intelligence, infrared sensors, electronic warfare, navigation warfare. Secondly, the emerging military applications of the electromagnetic spectrum are emphatically introduced, such as 5G communications, artificial intelligence applications, laser communications, directed-energy weapons, anti-UAV systems and emerging concepts. Finally, the strategies and policies of the electromagnetic spectrum made by the US Department of Defense are summarized, and some suggestions are put forward to strengthen the operational capability of electromagnetic spectrum.
The electromagnetic spectrum is the sixth dimensional battle space after land, sea, air, space and cyber, and runs through the other five dimensional battles. In order to effectively deal with the challenges and threats of the electromagnetic spectrum, the US Department of Defense puts forward the concept of electromagnetic spectrum operations, which has a profound impact on future combat patterns and even war forms. From the perspective of electromagnetic spectrum application, the application status and representative work of the US Department of Defense on electromagnetic spectrum are systematically introduced. Firstly, the traditional military applications of the electromagnetic spectrum are systematically sorted out, such as communications, radar, signals intelligence, infrared sensors, electronic warfare, navigation warfare. Secondly, the emerging military applications of the electromagnetic spectrum are emphatically introduced, such as 5G communications, artificial intelligence applications, laser communications, directed-energy weapons, anti-UAV systems and emerging concepts. Finally, the strategies and policies of the electromagnetic spectrum made by the US Department of Defense are summarized, and some suggestions are put forward to strengthen the operational capability of electromagnetic spectrum.
6 Communication signal modulation recognition method based on complex network and attention mechanism
2023, 21(6):745-750.
DOI: 10.11805/TKYDA2022231
Abstract:
As an important means of managing and monitoring the electromagnetic spectrum, communication signal modulation recognition shows important research value and application prospects. A signal modulation recognition method based on complex neural network is proposed by using the frequency domain information of modulated signals for modulation recognition. Firstly, the I and Q signals are combined into complex signals, and the real and imaginary parts obtained are combined as the data set of the input network after Fast Fourier Transform(FFT). Secondly, a complex neural network structure is designed, and an attention mechanism is introduced to improve the network structure. Finally, the simulation results show that the proposed method can effectively identify nine modulation modes, and the average correct recognition rate reaches 96.33% when the signal-to-noise ratio is 6 dB.
As an important means of managing and monitoring the electromagnetic spectrum, communication signal modulation recognition shows important research value and application prospects. A signal modulation recognition method based on complex neural network is proposed by using the frequency domain information of modulated signals for modulation recognition. Firstly, the I and Q signals are combined into complex signals, and the real and imaginary parts obtained are combined as the data set of the input network after Fast Fourier Transform(FFT). Secondly, a complex neural network structure is designed, and an attention mechanism is introduced to improve the network structure. Finally, the simulation results show that the proposed method can effectively identify nine modulation modes, and the average correct recognition rate reaches 96.33% when the signal-to-noise ratio is 6 dB.
