Email alert
Latest Articles
Articles in press have been peer-reviewed and accepted, which are not yet assigned to volumes/issues, but are citable by Digital Object Identifier (DOI).
Display Method:
Available online , doi: 10.11999/JEIT231426
Abstract:
In open, dynamic environments where the range of object categories continually expands, the challenge of remote sensing object detection is to detect a known set of object categories while simultaneously identifying unknown objects. To this end, a remote sensing open-set object detection network based on adaptive pre-screening is proposed. Firstly, an adaptive pre-screening module is proposed for object region proposals. Based on the coordinates of the selected region proposals, queries with rich semantic information and spatial features are generated and passed to the decoder. Subsequently, a pseudo-label selection method is devised based on object edge information, and loss functions are constructed with the aim of open set classification to enhance the network’s ability to learn knowledge of unknown classes. Finally, the Military Aircraft Recognition (MAR20) dataset is used to simulate various dynamic environments. Extensive comparative experiments and ablation experiments show that the proposed method can achieve reliable detection of known and unknown objects.
In open, dynamic environments where the range of object categories continually expands, the challenge of remote sensing object detection is to detect a known set of object categories while simultaneously identifying unknown objects. To this end, a remote sensing open-set object detection network based on adaptive pre-screening is proposed. Firstly, an adaptive pre-screening module is proposed for object region proposals. Based on the coordinates of the selected region proposals, queries with rich semantic information and spatial features are generated and passed to the decoder. Subsequently, a pseudo-label selection method is devised based on object edge information, and loss functions are constructed with the aim of open set classification to enhance the network’s ability to learn knowledge of unknown classes. Finally, the Military Aircraft Recognition (MAR20) dataset is used to simulate various dynamic environments. Extensive comparative experiments and ablation experiments show that the proposed method can achieve reliable detection of known and unknown objects.
Available online , doi: 10.11999/JEIT231335
Abstract:
Due to the non-uniform ground clutter in the forward array of airborne weather radar, it is difficult to obtain enough independent and equally distributed samples, which affects the accurate estimation of clutter covariance matrix and wind speed estimation. In this paper, a novel estimation method of low altitude wind shear speed based on convolutional neural network STAP is proposed, which can realize high resolution clutter space-time spectrum estimation with a small number of samples. First, the high-resolution clutter space-time spectrum convolutional neural network is trained based on the convolutional neural network model, and then the clutter covariance matrix is calculated, and then the optimal weight vector of the convolutional neural network STAP is calculated for clutter suppression, so as to accurately estimate the wind shear speed at low altitude. In this paper, the sparse recovery problem is realized by convolutional neural network in the case of small samples, and the space-time spectrum of high-resolution clutter is effectively estimated. The simulation results show that the proposed method can effectively estimate the space-time spectrum and complete the wind speed estimation.
Due to the non-uniform ground clutter in the forward array of airborne weather radar, it is difficult to obtain enough independent and equally distributed samples, which affects the accurate estimation of clutter covariance matrix and wind speed estimation. In this paper, a novel estimation method of low altitude wind shear speed based on convolutional neural network STAP is proposed, which can realize high resolution clutter space-time spectrum estimation with a small number of samples. First, the high-resolution clutter space-time spectrum convolutional neural network is trained based on the convolutional neural network model, and then the clutter covariance matrix is calculated, and then the optimal weight vector of the convolutional neural network STAP is calculated for clutter suppression, so as to accurately estimate the wind shear speed at low altitude. In this paper, the sparse recovery problem is realized by convolutional neural network in the case of small samples, and the space-time spectrum of high-resolution clutter is effectively estimated. The simulation results show that the proposed method can effectively estimate the space-time spectrum and complete the wind speed estimation.
Available online , doi: 10.11999/JEIT231257
Abstract:
Static power consumption dominates the power overhead of Network-on-Chip (NoC) as the technology size shrinks. Power gating, a generalized power saving technique, turns off idle modules in NoCs to reduce static power consumption. However, the conventional power gating technique brings problems such as packet wake-up delay, break-even time, etc. To solve the above problems, the Partition Bypass Transmission Infrastructure (PBTI), which replaces the power gated router for packet transmission, is proposed in this paper, and a low-latency, low-power power gating scheme has been designed based upon this bypass mechanism. PBTI uses mutually independent bypasses to process separately the PBTI uses independent bypasses to handle east-west packets separately, and uses common buffers within the bypasses to improve buffer utilization. PBTI can inject, transmit, and eject packets when the router is powered off. Packets can be transmitted from the source node to the destination node even if all routers in the network are power gated. When the traffic increases beyond the transmission capacity of PBTI, the routers perform a uniform wake-up in columns. Experimental results show that compared to the NoC without power gating, the scheme in this paper reduces 83.4% of static power consumption and 17.2% of packet delay, while adding only 6.2% additional area overhead. Compared to the conventional power gating scheme the power gated design in this paper achieves lower power consumption and delay, which is a significant advantage.
Static power consumption dominates the power overhead of Network-on-Chip (NoC) as the technology size shrinks. Power gating, a generalized power saving technique, turns off idle modules in NoCs to reduce static power consumption. However, the conventional power gating technique brings problems such as packet wake-up delay, break-even time, etc. To solve the above problems, the Partition Bypass Transmission Infrastructure (PBTI), which replaces the power gated router for packet transmission, is proposed in this paper, and a low-latency, low-power power gating scheme has been designed based upon this bypass mechanism. PBTI uses mutually independent bypasses to process separately the PBTI uses independent bypasses to handle east-west packets separately, and uses common buffers within the bypasses to improve buffer utilization. PBTI can inject, transmit, and eject packets when the router is powered off. Packets can be transmitted from the source node to the destination node even if all routers in the network are power gated. When the traffic increases beyond the transmission capacity of PBTI, the routers perform a uniform wake-up in columns. Experimental results show that compared to the NoC without power gating, the scheme in this paper reduces 83.4% of static power consumption and 17.2% of packet delay, while adding only 6.2% additional area overhead. Compared to the conventional power gating scheme the power gated design in this paper achieves lower power consumption and delay, which is a significant advantage.
Available online , doi: 10.11999/JEIT231202
Abstract:
The integration of satellite communication and ground mobile communication in a complementary manner has emerged as a prevailing trend, which means the wireless radio frequency front-end with Power Amplifier (PA) as the core need to tackle the dual challenges of high efficiency and large bandwidth. In this paper, the proposed input harmonic phase control method effectively overcomes the bottleneck of mutual restriction between bandwidth and efficiency. By employing a continuous inverse Class-F operating mode, it enables the reconstruction of transistor drain waveform through precise control of the input second harmonic phase. This approach ensures high efficiency, while significantly enhancing the impedance design space. Based on the expanded design space, a continuous inverse Class-F PA is designed and fabricated over the frequency band of 1.7-3.0 GHz. Experimental results demonstrate an output power of 40.6-42.8 dBm, accompanied by a drain efficiency ranging from 72.2% to 78.6%. Additionally, the gain of the designed PA ranges from 10.6 dB to 14.8 dB.
The integration of satellite communication and ground mobile communication in a complementary manner has emerged as a prevailing trend, which means the wireless radio frequency front-end with Power Amplifier (PA) as the core need to tackle the dual challenges of high efficiency and large bandwidth. In this paper, the proposed input harmonic phase control method effectively overcomes the bottleneck of mutual restriction between bandwidth and efficiency. By employing a continuous inverse Class-F operating mode, it enables the reconstruction of transistor drain waveform through precise control of the input second harmonic phase. This approach ensures high efficiency, while significantly enhancing the impedance design space. Based on the expanded design space, a continuous inverse Class-F PA is designed and fabricated over the frequency band of 1.7-3.0 GHz. Experimental results demonstrate an output power of 40.6-42.8 dBm, accompanied by a drain efficiency ranging from 72.2% to 78.6%. Additionally, the gain of the designed PA ranges from 10.6 dB to 14.8 dB.
Available online , doi: 10.11999/JEIT230826
Abstract:
Cervical cell classification plays a crucial role in assisting the diagnosis of cervical cancer. However, existing methods for cervical cell classification do not enough consider relationships among cells and background information, and fail to effectively simulate the diagnostic approach of pathology doctors. As a result, their classification performance is limited. In this study, a novel approach that integrates cell relationships and background information for cervical cell classification is proposed. The proposed method consists of a Graph Attention Branch for Cell-Cell Relationships (GAB-CCR) and a Background Attention Branching for Whole Slide Images (BAB-WSI). GAB-CCR utilizes cosine similarity of cell features to construct preliminary graphs representing similar and distinct cell relationships. Additionally, GAB-CCR enhances the ability of models in modeling cell relationships through GATv2. BAB-WSI employs multi-head attention to effectively capture crucial information on the slide background and reflect the importance of different regions. Finally, the enhanced cell and background features are fused to improve the classification performance of the network. Experimental results demonstrate that the proposed method achieves significant improvements over the baseline model, Swin Transformer-L, with improvement in accuracy, sensitivity, specificity, and F1-Score by 15.9%, 30.32%, 8.11%, and 31.62% respectively.
Cervical cell classification plays a crucial role in assisting the diagnosis of cervical cancer. However, existing methods for cervical cell classification do not enough consider relationships among cells and background information, and fail to effectively simulate the diagnostic approach of pathology doctors. As a result, their classification performance is limited. In this study, a novel approach that integrates cell relationships and background information for cervical cell classification is proposed. The proposed method consists of a Graph Attention Branch for Cell-Cell Relationships (GAB-CCR) and a Background Attention Branching for Whole Slide Images (BAB-WSI). GAB-CCR utilizes cosine similarity of cell features to construct preliminary graphs representing similar and distinct cell relationships. Additionally, GAB-CCR enhances the ability of models in modeling cell relationships through GATv2. BAB-WSI employs multi-head attention to effectively capture crucial information on the slide background and reflect the importance of different regions. Finally, the enhanced cell and background features are fused to improve the classification performance of the network. Experimental results demonstrate that the proposed method achieves significant improvements over the baseline model, Swin Transformer-L, with improvement in accuracy, sensitivity, specificity, and F1-Score by 15.9%, 30.32%, 8.11%, and 31.62% respectively.
Available online , doi: 10.11999/JEIT231417
Abstract:
Empathic dialogue aims to provide mental health support for anxious users, thus chatbots with empathic capabilities is a noteworthy issue. The existing methods can only identify users' sentiment states, but can not effectively generate empathetic responses according to users' different sentiment states and let alone effectively relieve users' bad emotions. Therefore, in the research of building sentiment support chatbots, how to dynamically capture users' fine-grained sentiment features and provide corresponding psychological support according to sentiment features needs to be further explored. This paper proposes an empathic dialogue generation method based on the fusion of emotion and strategy. Firstly, the sentiment classification network is used to dynamically perceive the user's sentiment state. Then the support strategy is used to accurately model the strategy matching network which is introduced according to the context of the conversation to generate the conversation. Finally, by comparing the experimental results of the proposed method and the current advanced methods on the corresponding datasets, the effectiveness of the proposed method and the importance of sentiment support are verified.
Empathic dialogue aims to provide mental health support for anxious users, thus chatbots with empathic capabilities is a noteworthy issue. The existing methods can only identify users' sentiment states, but can not effectively generate empathetic responses according to users' different sentiment states and let alone effectively relieve users' bad emotions. Therefore, in the research of building sentiment support chatbots, how to dynamically capture users' fine-grained sentiment features and provide corresponding psychological support according to sentiment features needs to be further explored. This paper proposes an empathic dialogue generation method based on the fusion of emotion and strategy. Firstly, the sentiment classification network is used to dynamically perceive the user's sentiment state. Then the support strategy is used to accurately model the strategy matching network which is introduced according to the context of the conversation to generate the conversation. Finally, by comparing the experimental results of the proposed method and the current advanced methods on the corresponding datasets, the effectiveness of the proposed method and the importance of sentiment support are verified.
Available online , doi: 10.11999/JEIT240050
Abstract:
The Radio Environment Map (REM) is one of the effective ways to represent the electromagnetic situation. Considering the issue that the actual observed incomplete spectrum map is corrupted by the impulses and the noises, the incomplete radio environment map is reconstructed and the specific emitter identification is performed based on the reconstructed maps. First, the spectrum map in the complex electromagnetic environment is modeled as the high-dimensional spectrum tensor, and the incomplete spectrum tensor is initially completed by the linear interpolation in preprocessing. Then, the vision transformer model is employed to solve the semantic segmentation problem in order to identify the spectrum semantic regions, in which the power of only one emitter dominates and the low-rank property of each semantic tensor is further preserved. To reconstruct the REM, the compressed tensor decomposition algorithm is proposed, and the expected signal spectrum and impulses are recovered utilizing the Alternating Direction Method of Multipliers (ADMM) in the semantic regions. Finally, the locations of the unknown emitters are detected on the reconstructed spectrum map. The proposed approach leverages the low-rank property of spectrum data and works well in wide-area electromagnetic scenarios involving multiple emitters. The simulation results demonstrate that the proposed approach outperforms the comparative approach in terms of reconstruction performance. It requires fewer observation samples to achieve the same spectrum map recovery accuracy and can accurately detect emitters.
The Radio Environment Map (REM) is one of the effective ways to represent the electromagnetic situation. Considering the issue that the actual observed incomplete spectrum map is corrupted by the impulses and the noises, the incomplete radio environment map is reconstructed and the specific emitter identification is performed based on the reconstructed maps. First, the spectrum map in the complex electromagnetic environment is modeled as the high-dimensional spectrum tensor, and the incomplete spectrum tensor is initially completed by the linear interpolation in preprocessing. Then, the vision transformer model is employed to solve the semantic segmentation problem in order to identify the spectrum semantic regions, in which the power of only one emitter dominates and the low-rank property of each semantic tensor is further preserved. To reconstruct the REM, the compressed tensor decomposition algorithm is proposed, and the expected signal spectrum and impulses are recovered utilizing the Alternating Direction Method of Multipliers (ADMM) in the semantic regions. Finally, the locations of the unknown emitters are detected on the reconstructed spectrum map. The proposed approach leverages the low-rank property of spectrum data and works well in wide-area electromagnetic scenarios involving multiple emitters. The simulation results demonstrate that the proposed approach outperforms the comparative approach in terms of reconstruction performance. It requires fewer observation samples to achieve the same spectrum map recovery accuracy and can accurately detect emitters.
Available online , doi: 10.11999/JEIT231255
Abstract:
Forward Scatter Radar (FSR) can obtain high level Radar Cross Section (RCS), so it plays an important role in anti-stealth. The Global Navigation Satellite System (GNSS) has the advantage of all-weather coverage throughout the day as a radiation source and the ground/sea/air target surveillance network can be built by deploying multiple receiving nodes. According to the development status of GNSS-based FSR, the key technologies and the existing problems in target detection, target parameter estimation, Shadow Inverse Synthetic Aperture Radar (SISAR) imaging, and target classification are summarized. What’s more, the development trend of GNSS-based FSR is prospected from the aspects of network detection, multi-target location, station optimization and polarization information acquisition.
Forward Scatter Radar (FSR) can obtain high level Radar Cross Section (RCS), so it plays an important role in anti-stealth. The Global Navigation Satellite System (GNSS) has the advantage of all-weather coverage throughout the day as a radiation source and the ground/sea/air target surveillance network can be built by deploying multiple receiving nodes. According to the development status of GNSS-based FSR, the key technologies and the existing problems in target detection, target parameter estimation, Shadow Inverse Synthetic Aperture Radar (SISAR) imaging, and target classification are summarized. What’s more, the development trend of GNSS-based FSR is prospected from the aspects of network detection, multi-target location, station optimization and polarization information acquisition.
Available online , doi: 10.11999/JEIT231434
Abstract:
The DNA origami nanostructure encapsulates intricate sequence-folding information, presenting a novel avenue for exploiting cryptography with a vast key space. This paper introduces an encryption strategy that fully realizes the structure-based potential of DNA origami. In contrast to previous approach centered on the folding of DNA origami scaffold, an alternative methodology is introduced based on the nonlinear combination characteristics of staple ensembles. This approach aims to achieve a larger key space by exploring the inherent extensive folding diversity of staple. The key space computational model is delineated into three factors: the binding domain mode, cooperative folding, and independence of staples. These three factors respectively account for the intra-chain distribution, inter-chain arrangement diversity, and sequence specificity of staples. The combination of these factors effectively converts the folding diversity of DNA origami in per unit of geometric space into key space. This strategy represents a cryptography rooted in the principles of biomolecular thermodynamics, offering new possibilities for extending the application scenarios of information security.
The DNA origami nanostructure encapsulates intricate sequence-folding information, presenting a novel avenue for exploiting cryptography with a vast key space. This paper introduces an encryption strategy that fully realizes the structure-based potential of DNA origami. In contrast to previous approach centered on the folding of DNA origami scaffold, an alternative methodology is introduced based on the nonlinear combination characteristics of staple ensembles. This approach aims to achieve a larger key space by exploring the inherent extensive folding diversity of staple. The key space computational model is delineated into three factors: the binding domain mode, cooperative folding, and independence of staples. These three factors respectively account for the intra-chain distribution, inter-chain arrangement diversity, and sequence specificity of staples. The combination of these factors effectively converts the folding diversity of DNA origami in per unit of geometric space into key space. This strategy represents a cryptography rooted in the principles of biomolecular thermodynamics, offering new possibilities for extending the application scenarios of information security.
Available online , doi: 10.11999/JEIT240120
Abstract:
The Residual Self-Interference (RSI) caused by Co-frequency and Co-time Full Duplex Device-to-Device (CCFD-D2D) and the interference introduced by spectrum sharing between D2D User (DU) and Cellular User (CU) lead to a degradation in the quality of experience for CUs. Therefore, the CCFD-D2D underlaying cellular system is considered and two algorithms are proposed, that is Maximizing Sum-rate of CU (MaxSumCU) and Maximizing Minimum-rate of CU (MaxMinCU) algorithm, to enhance the experience for CUs while spectral efficiency of the system is improved. For the MaxSumCU algorithm, an optimization problem is investigated to maximize the sum rate of CUs in the system, and formulate it as a Mixed Integer NonLinear Programming problem (MINLP) which is NP-hard in mathematics. MaxSumCU is designed to decompose it into two sub-problems as power control and spectral resource allocation. The power control is solved by geometric programming, and the resource allocation is achieved by employing Kuhn-Munkres algorithm to determine the spectrum sharing pairs of CUs and DUs. To provide a more uniform rate performance across all CUs, the MaxMinCU algorithm is designed to maximize the minimum rate among the CUs. The novel spectrum resource allocation algorithm based on bisection searching and Kuhn-Munkres minimum-weight algorithm is proposed to solve this optimization problem. Numerical results show that, compared with Maximizing Sum-rate of Cell (MaxSumCell) design, our proposed algorithm effectively optimize the CU’s experience while improve the spectral efficiency of system in CCFD-D2D underlaying cellular networks.
The Residual Self-Interference (RSI) caused by Co-frequency and Co-time Full Duplex Device-to-Device (CCFD-D2D) and the interference introduced by spectrum sharing between D2D User (DU) and Cellular User (CU) lead to a degradation in the quality of experience for CUs. Therefore, the CCFD-D2D underlaying cellular system is considered and two algorithms are proposed, that is Maximizing Sum-rate of CU (MaxSumCU) and Maximizing Minimum-rate of CU (MaxMinCU) algorithm, to enhance the experience for CUs while spectral efficiency of the system is improved. For the MaxSumCU algorithm, an optimization problem is investigated to maximize the sum rate of CUs in the system, and formulate it as a Mixed Integer NonLinear Programming problem (MINLP) which is NP-hard in mathematics. MaxSumCU is designed to decompose it into two sub-problems as power control and spectral resource allocation. The power control is solved by geometric programming, and the resource allocation is achieved by employing Kuhn-Munkres algorithm to determine the spectrum sharing pairs of CUs and DUs. To provide a more uniform rate performance across all CUs, the MaxMinCU algorithm is designed to maximize the minimum rate among the CUs. The novel spectrum resource allocation algorithm based on bisection searching and Kuhn-Munkres minimum-weight algorithm is proposed to solve this optimization problem. Numerical results show that, compared with Maximizing Sum-rate of Cell (MaxSumCell) design, our proposed algorithm effectively optimize the CU’s experience while improve the spectral efficiency of system in CCFD-D2D underlaying cellular networks.
Available online , doi: 10.11999/JEIT240171
Abstract:
Due to the coupling effect of emitter distortion and receiver distortion, the actual received signal contains the information of the current emitter system and the receiving system, which makes the Radio Frequency Fingerprinting (RFF) technology unable to be generalized in cross-receiving system scenarios. In order to eliminate the effect of receiver, in this paper, a universal RFF method across receiving systems based on receiving domain separation is proposed which considers the influence of the receiver as a separate scope. Through the dual-label multi-channel fusion feature and domain separation adversarial reconstruction method, after trained with multi-receiver data in the source domain, the proposed method can separate domains of transmitting and receiving, extract emitter fingerprint information, which improves the generalization of RFF in scenarios such as cross-receiving system and cross-platform. Compared with the existing cross-receiver RFF methods, the proposed method can truly adapt to the actual unsupervised scenario. And the more the number of source domain receivers participating in the training, the better the domain adaptation effect. It can be directly applied to the new receiving system without repeated training, which has high practical application value.
Due to the coupling effect of emitter distortion and receiver distortion, the actual received signal contains the information of the current emitter system and the receiving system, which makes the Radio Frequency Fingerprinting (RFF) technology unable to be generalized in cross-receiving system scenarios. In order to eliminate the effect of receiver, in this paper, a universal RFF method across receiving systems based on receiving domain separation is proposed which considers the influence of the receiver as a separate scope. Through the dual-label multi-channel fusion feature and domain separation adversarial reconstruction method, after trained with multi-receiver data in the source domain, the proposed method can separate domains of transmitting and receiving, extract emitter fingerprint information, which improves the generalization of RFF in scenarios such as cross-receiving system and cross-platform. Compared with the existing cross-receiver RFF methods, the proposed method can truly adapt to the actual unsupervised scenario. And the more the number of source domain receivers participating in the training, the better the domain adaptation effect. It can be directly applied to the new receiving system without repeated training, which has high practical application value.
Available online , doi: 10.11999/JEIT240178
Abstract:
The multimodal fusion method can effectively improve the ground object classification accuracy by using the complementary characteristics of different modalities, which has become a research hotspot in the various fields in recent years. The existing multimodal fusion methods have been successfully applied to multi-source remote sensing classification tasks oriented to HyperSpectral Image (HSI) and Light Detection And Ranging (LiDAR). However, existing research still faces many challenges, including difficulty in capturing spatial dependencies among irregular ground objects and obtaining discriminative information in multimodal data. To address the above challenges, a Scale Adaptive Fusion Network (SAFN)is proposed in this paper, by integrating the fusion of multimodal, multiscale, and multiview features into a unified framework. First, a dynamic multiscale graph module is proposed to capture the complex spatial dependencies of ground object, enhancing the model’s adaptability to irregular and scale-dissimilar ground object. Second, the complementary properties of LiDAR and HSI are utilized to constrain ground object within the same spatial neighborhood to have similar feature representations, thereby acquiring discriminative remote sensing features. Then, a multimodal spatial-spectral graph fusion module is proposed to establish feature interactions among multimodal, multiscale, and multiview features, providing discriminative fusion features for classification tasks by capturing class-recognition information that can be shared among features. Finally, the fusion features are fed into a classifier to obtain class probability scores for predicting the ground object class. To verify the effectiveness of SAFN, experiments are conducted on three datasets (i.e., Houston, Trento, and MUUFL). The experimental results show that, SAFN achieved state-of-the-art performance in multi-source remote sensing data classification tasks when compared with existing mainstream methods.
The multimodal fusion method can effectively improve the ground object classification accuracy by using the complementary characteristics of different modalities, which has become a research hotspot in the various fields in recent years. The existing multimodal fusion methods have been successfully applied to multi-source remote sensing classification tasks oriented to HyperSpectral Image (HSI) and Light Detection And Ranging (LiDAR). However, existing research still faces many challenges, including difficulty in capturing spatial dependencies among irregular ground objects and obtaining discriminative information in multimodal data. To address the above challenges, a Scale Adaptive Fusion Network (SAFN)is proposed in this paper, by integrating the fusion of multimodal, multiscale, and multiview features into a unified framework. First, a dynamic multiscale graph module is proposed to capture the complex spatial dependencies of ground object, enhancing the model’s adaptability to irregular and scale-dissimilar ground object. Second, the complementary properties of LiDAR and HSI are utilized to constrain ground object within the same spatial neighborhood to have similar feature representations, thereby acquiring discriminative remote sensing features. Then, a multimodal spatial-spectral graph fusion module is proposed to establish feature interactions among multimodal, multiscale, and multiview features, providing discriminative fusion features for classification tasks by capturing class-recognition information that can be shared among features. Finally, the fusion features are fed into a classifier to obtain class probability scores for predicting the ground object class. To verify the effectiveness of SAFN, experiments are conducted on three datasets (i.e., Houston, Trento, and MUUFL). The experimental results show that, SAFN achieved state-of-the-art performance in multi-source remote sensing data classification tasks when compared with existing mainstream methods.
Available online , doi: 10.11999/JEIT240189
Abstract:
To address challenges such as high complexity in monocular depth estimation networks and low accuracy in regions with weak textures, a Direction-Aware Enhancement-based lightweight self-supervised monocular depth estimation Network (DAEN) is proposed in this paper. Firstly, the Iterative Dilated Convolution module (IDC) is introduced as the core of the encoder to extract correlations among distant pixels. Secondly, the Directional Awareness Enhancement module (DAE) is designed to enhance feature extraction in the vertical direction, providing the depth estimation model with additional depth cues. Furthermore, the problem of detail loss during the decoder upsampling process is addressed through the aggregation of disparity map features. Lastly, the Feature Attention Module (FAM) is employed to connect the encoder and decoder, effectively leveraging global contextual information to resolve adaptability issues in regions with weak textures. Experimental results on the KITTI dataset demonstrate that the proposed method has a model parameter count of only 2.9M, achieving an advanced performance with\begin{document}$ \delta $\end{document}
To address challenges such as high complexity in monocular depth estimation networks and low accuracy in regions with weak textures, a Direction-Aware Enhancement-based lightweight self-supervised monocular depth estimation Network (DAEN) is proposed in this paper. Firstly, the Iterative Dilated Convolution module (IDC) is introduced as the core of the encoder to extract correlations among distant pixels. Secondly, the Directional Awareness Enhancement module (DAE) is designed to enhance feature extraction in the vertical direction, providing the depth estimation model with additional depth cues. Furthermore, the problem of detail loss during the decoder upsampling process is addressed through the aggregation of disparity map features. Lastly, the Feature Attention Module (FAM) is employed to connect the encoder and decoder, effectively leveraging global contextual information to resolve adaptability issues in regions with weak textures. Experimental results on the KITTI dataset demonstrate that the proposed method has a model parameter count of only 2.9M, achieving an advanced performance with
Available online , doi: 10.11999/JEIT240297
Abstract:
With the continuous emergence of new applications, the issue of spectrum congestion is becoming increasingly severe. Dual-Functional Radar-Communication (DFRC) is consideredas a key enabling technology for many emerging applications and is one of the essential approaches to addressing the issue of spectrum congestion. However, how to solve the mutual interference between communication and radar and achieve high communication rate is a fundamental challenge that urgently needs to be solved in DFRC system. Based on multi carrier complementary coded division multiple access technology, a DFRC signal suitable for multi-user scenarios is designed in this paper. Theoretical analysis and simulation results show that compared with typical spread spectrum schemes, the proposed scheme can achieve non-interference transmission between communication and radar, with low bit error rate and high user communication rate.
With the continuous emergence of new applications, the issue of spectrum congestion is becoming increasingly severe. Dual-Functional Radar-Communication (DFRC) is consideredas a key enabling technology for many emerging applications and is one of the essential approaches to addressing the issue of spectrum congestion. However, how to solve the mutual interference between communication and radar and achieve high communication rate is a fundamental challenge that urgently needs to be solved in DFRC system. Based on multi carrier complementary coded division multiple access technology, a DFRC signal suitable for multi-user scenarios is designed in this paper. Theoretical analysis and simulation results show that compared with typical spread spectrum schemes, the proposed scheme can achieve non-interference transmission between communication and radar, with low bit error rate and high user communication rate.
Available online , doi: 10.11999/JEIT231272
Abstract:
In the information era, information security is the priority that cannot be ignored. Attacks and protection against password devices are research hotspots in this field. In recent years, various attacks on cryptographic devices have become well-known, all aimed at obtaining keys from the device. Among these attacks, power side channel attack is one of the most concerned attack techniques. Mask technology is an effective method to combat power side channel attacks, however, with the continuous progress of attack methods, the protection of first-order mask is no longer sufficient to cope with second-order and higher order power analysis attack, so the research on higher-order mask has considerable significance. To enhance the encryption circuit’s capability of anti-attack, high-order masking schemes: N-share masking is implemented on S-box in this paper, and a universal design method for galois field secure multiplication is proposed, which is based on the secure scheme published by Ishai et al. at Crypto 2003 (ISW framework). Through experiments, it has been shown that the encryption scheme adopted in this paper does not affect the functionality of the encryption algorithm, and can resist first-order and second-order correlation power analysis attack.
In the information era, information security is the priority that cannot be ignored. Attacks and protection against password devices are research hotspots in this field. In recent years, various attacks on cryptographic devices have become well-known, all aimed at obtaining keys from the device. Among these attacks, power side channel attack is one of the most concerned attack techniques. Mask technology is an effective method to combat power side channel attacks, however, with the continuous progress of attack methods, the protection of first-order mask is no longer sufficient to cope with second-order and higher order power analysis attack, so the research on higher-order mask has considerable significance. To enhance the encryption circuit’s capability of anti-attack, high-order masking schemes: N-share masking is implemented on S-box in this paper, and a universal design method for galois field secure multiplication is proposed, which is based on the secure scheme published by Ishai et al. at Crypto 2003 (ISW framework). Through experiments, it has been shown that the encryption scheme adopted in this paper does not affect the functionality of the encryption algorithm, and can resist first-order and second-order correlation power analysis attack.
Available online , doi: 10.11999/JEIT240062
Abstract:
The correction in Digital Elevation Models (DEMs) has always been a crucial aspect of remote sensing geoscience research. The burgeoning development of new machine learning methods in recent years has provided novel solutions for the correction of DEM elevation errors. Given the reliance of machine learning and other artificial intelligence methods on extensive training data, and considering the current lack of publicly available, unified, large-scale, and standardized multisource DEM elevation error prediction datasets for large areas, the multi-source DEM Elevation Error Prediction Dataset (DEEP-Dataset) is introduced in this paper. This dataset comprises four sub-datasets, based on the TerraSAR-X add-on for Digital Elevation Measurements (TanDEM-X) DEM and Advanced land observing satellite World 3D-30 m (AW3D30) DEM in the Guangdong Province study area of China, and the Shuttle Radar Topography Mission (SRTM) DEM and Advanced Spaceborne Thermal Emission and reflection Radiometer (ASTER) DEM in the Northern Territory study area of Australia. The Guangdong Province sample comprises approximately 40 000 instances, while the Northern Territory sample includes about 1 600 000 instances. Each sample in the dataset consists of ten features, encompassing geographic spatial information, land cover types, and topographic attributes. The effectiveness of the DEEP-Dataset in actual model training and DEM correction has been validated through a series of comparative experiments, including machine learning model testing, DEM correction, and feature importance assessment. These experiments demonstrate the dataset’s rationality, effectiveness, and comprehensiveness.
The correction in Digital Elevation Models (DEMs) has always been a crucial aspect of remote sensing geoscience research. The burgeoning development of new machine learning methods in recent years has provided novel solutions for the correction of DEM elevation errors. Given the reliance of machine learning and other artificial intelligence methods on extensive training data, and considering the current lack of publicly available, unified, large-scale, and standardized multisource DEM elevation error prediction datasets for large areas, the multi-source DEM Elevation Error Prediction Dataset (DEEP-Dataset) is introduced in this paper. This dataset comprises four sub-datasets, based on the TerraSAR-X add-on for Digital Elevation Measurements (TanDEM-X) DEM and Advanced land observing satellite World 3D-30 m (AW3D30) DEM in the Guangdong Province study area of China, and the Shuttle Radar Topography Mission (SRTM) DEM and Advanced Spaceborne Thermal Emission and reflection Radiometer (ASTER) DEM in the Northern Territory study area of Australia. The Guangdong Province sample comprises approximately 40 000 instances, while the Northern Territory sample includes about 1 600 000 instances. Each sample in the dataset consists of ten features, encompassing geographic spatial information, land cover types, and topographic attributes. The effectiveness of the DEEP-Dataset in actual model training and DEM correction has been validated through a series of comparative experiments, including machine learning model testing, DEM correction, and feature importance assessment. These experiments demonstrate the dataset’s rationality, effectiveness, and comprehensiveness.
Available online , doi: 10.11999/JEIT240123
Abstract:
The traditional radar adaptive target detectors in sea clutter usually assumes that the clutter texture is independent and identically distributed in the range dimension, ignoring the correlation information of the texture in the range dimension. In order to improve the adaptive detection performance of radar targets in sea clutter with texture range correlation, the texture component of compound Gaussian sea clutter is modeled as a lognormal random variable, and then a generalized likelihood ratio test with homogeneous lognormal texture detector is proposed based on the generalized likelihood ratio test. The proposed detector uses the prior distribution of texture and the correlated information of texture range. The simulation and the measured data beging used show that the detection probability of this detector for radar targets in compound Gaussian sea clutter with texture range correlation is higher than that of the existing detectors.
The traditional radar adaptive target detectors in sea clutter usually assumes that the clutter texture is independent and identically distributed in the range dimension, ignoring the correlation information of the texture in the range dimension. In order to improve the adaptive detection performance of radar targets in sea clutter with texture range correlation, the texture component of compound Gaussian sea clutter is modeled as a lognormal random variable, and then a generalized likelihood ratio test with homogeneous lognormal texture detector is proposed based on the generalized likelihood ratio test. The proposed detector uses the prior distribution of texture and the correlated information of texture range. The simulation and the measured data beging used show that the detection probability of this detector for radar targets in compound Gaussian sea clutter with texture range correlation is higher than that of the existing detectors.
Available online , doi: 10.11999/JEIT240318
Abstract:
For cross-modal person Re-IDentification (Re-ID) in visible-infrared images, methods using modality conversion and adversarial networks yield associative information between modalities. However, these approaches fall short in effective feature recognition. Thus, a two-stage approach using visual-text matching and graph embedding for enhanced re-identification effectiveness is proposed in this paper. A context-optimized scheme is utilized by the method to construct learnable text templates that generate person descriptions as associative information between modalities. Specifically, in the first stage, unified text descriptions of the same person across different modalities are utilized as prior information, assisting in the reduction of modality differences, based on the Contrastive Language–Image Pre-training (CLIP) model. Meanwhile, in the second stage, a cross-modal constraint framework based on graph embedding is applied, and a modality-adaptive loss function is designed, aiming to improve person recognition accuracy. The method’s efficacy has been confirmed through extensive experiments on the SYSU-MM01 and RegDB datasets, with a Rank-1 accuracy of 64.2% and mean Average Precision (mAP) of 60.2% on SYSU-MM01 being achieved, thereby demonstrating significant improvements in cross-modal person re-identification.
For cross-modal person Re-IDentification (Re-ID) in visible-infrared images, methods using modality conversion and adversarial networks yield associative information between modalities. However, these approaches fall short in effective feature recognition. Thus, a two-stage approach using visual-text matching and graph embedding for enhanced re-identification effectiveness is proposed in this paper. A context-optimized scheme is utilized by the method to construct learnable text templates that generate person descriptions as associative information between modalities. Specifically, in the first stage, unified text descriptions of the same person across different modalities are utilized as prior information, assisting in the reduction of modality differences, based on the Contrastive Language–Image Pre-training (CLIP) model. Meanwhile, in the second stage, a cross-modal constraint framework based on graph embedding is applied, and a modality-adaptive loss function is designed, aiming to improve person recognition accuracy. The method’s efficacy has been confirmed through extensive experiments on the SYSU-MM01 and RegDB datasets, with a Rank-1 accuracy of 64.2% and mean Average Precision (mAP) of 60.2% on SYSU-MM01 being achieved, thereby demonstrating significant improvements in cross-modal person re-identification.
Available online , doi: 10.11999/JEIT240025
Abstract:
The misuse of deepfake technology on social networks has raised serious concerns about the authenticity and reliability of visual content. The degradation phenomenon of deepfake videos on social networks has not been adequately considered in existing detection algorithms, resulting in deepfake detection performance being limited by challenging issues such as compression artifacts interference and lack of context-related information. Compression encoding and up-sampling operations in deepfake generation algorithms can leave artifacts on videos, which can result in fine-grained differences between real videos and deepfake videos. The common mechanisms between compression artifacts and deepfake artifacts are analyzed to reveal the structural similarities between them, which provides a reliable theoretical basis for enhancing the robustness of deepfake detection models against compression. Firstly, to address the interference of compression noise on deepfake features, the frequency-domain adaptive notch filter is designed based on the structural similarity of compression artifacts and deepfake artifacts to eliminate the interference of compression artifacts on specific frequency bands. Secondly, the denoising branch based on residual learning is designed to reduce the sensitivity of the deepfake detection model to unknown noise. Additionally, the attention-based feature fusion method is adopted to enhance the discriminative features of deepfakes. Metric learning strategies are adopted to optimize network models, achieving deepfake detection with resistance to compression. Theoretical analysis and experimental results indicate that the detection performance of compressed deepfake videos is significantly enhanced by using the algorithm proposed in this paper. It can be used as a plug-and-play model combined with existing detection methods to enhance their robustness against compression.
The misuse of deepfake technology on social networks has raised serious concerns about the authenticity and reliability of visual content. The degradation phenomenon of deepfake videos on social networks has not been adequately considered in existing detection algorithms, resulting in deepfake detection performance being limited by challenging issues such as compression artifacts interference and lack of context-related information. Compression encoding and up-sampling operations in deepfake generation algorithms can leave artifacts on videos, which can result in fine-grained differences between real videos and deepfake videos. The common mechanisms between compression artifacts and deepfake artifacts are analyzed to reveal the structural similarities between them, which provides a reliable theoretical basis for enhancing the robustness of deepfake detection models against compression. Firstly, to address the interference of compression noise on deepfake features, the frequency-domain adaptive notch filter is designed based on the structural similarity of compression artifacts and deepfake artifacts to eliminate the interference of compression artifacts on specific frequency bands. Secondly, the denoising branch based on residual learning is designed to reduce the sensitivity of the deepfake detection model to unknown noise. Additionally, the attention-based feature fusion method is adopted to enhance the discriminative features of deepfakes. Metric learning strategies are adopted to optimize network models, achieving deepfake detection with resistance to compression. Theoretical analysis and experimental results indicate that the detection performance of compressed deepfake videos is significantly enhanced by using the algorithm proposed in this paper. It can be used as a plug-and-play model combined with existing detection methods to enhance their robustness against compression.
Available online , doi: 10.11999/JEIT240102
Abstract:
Querying present and future traffic velocities of road segments is a routine task in urban intelligence transportation management, and a Vehicle-equipped-Edge Dominant (VED) method is proposed to answer the querying of present and future velocity of urban road segments. The collected data is exchanged with the other mobile sides by every vehicle-equipped mobile side when the mobile side’s speed falls below a given threshold, and the light-weighted present and history velocity indexes are constructed locally to support the querying of present velocity. To train as few models as possible to predict future velocities, a road network is proposed to be partitioned into a set of road-segment clusters based on the segments’ topological morphism and the spatio-temporal space is proposed to be partitioned into a set of model-equivalence classes according to the periodic time windows and road-segment clusters. The similar traffic patterns are exhibited by the road segments in the same model-equivalence class within the given time window. For every model-equivalence class, the federated learning is performed between the mobile sides and the data center to train the Long Short-Term Memories (LSTMs) which are stored at the mobile sides to answer the querying of future velocities of nearby areas. Data is indexed by every mobile side and queries are answered locally, thus the query response latency and possible communication congestion and can be avoided. Further, data is stored at the mobile sides, rather than at one data center, so as to prevent the privacy leakage due to security attacks.
Querying present and future traffic velocities of road segments is a routine task in urban intelligence transportation management, and a Vehicle-equipped-Edge Dominant (VED) method is proposed to answer the querying of present and future velocity of urban road segments. The collected data is exchanged with the other mobile sides by every vehicle-equipped mobile side when the mobile side’s speed falls below a given threshold, and the light-weighted present and history velocity indexes are constructed locally to support the querying of present velocity. To train as few models as possible to predict future velocities, a road network is proposed to be partitioned into a set of road-segment clusters based on the segments’ topological morphism and the spatio-temporal space is proposed to be partitioned into a set of model-equivalence classes according to the periodic time windows and road-segment clusters. The similar traffic patterns are exhibited by the road segments in the same model-equivalence class within the given time window. For every model-equivalence class, the federated learning is performed between the mobile sides and the data center to train the Long Short-Term Memories (LSTMs) which are stored at the mobile sides to answer the querying of future velocities of nearby areas. Data is indexed by every mobile side and queries are answered locally, thus the query response latency and possible communication congestion and can be avoided. Further, data is stored at the mobile sides, rather than at one data center, so as to prevent the privacy leakage due to security attacks.
Available online , doi: 10.11999/JEIT240130
Abstract:
An innovative aviation target tracking algorithm is presented in this paper, utilizing high-altitude unmanned airship dual photoelectric sensors in conjunction with Unmanned Aerial Vehicle (UAV)-borne two-coordinate radar. The algorithm addresses the challenge of integrating sensor data to accurately track targets when individual sensors lack complete target position information, thus overcoming limitations of traditional point-trace association methods. Initially, a two-level point-trace correlation algorithm based on angle and distance is introduced for multi-sensor measurement association following coordinate system transformation. Subsequently, a line-plane intersection fusion localization algorithm is proposed to determine the initial target track position through techniques such as least squares method, intersection projection, distance nearest point solution, and homologous data compression. Leveraging heterogeneous information from space-based multi-platform reconnaissance, an extended Unscented Kalman Filter (UKF) is designed to track aviation targets by enhancing the traditional UKF. Simulation results demonstrate that this algorithm achieves superior precision in tracking high-speed aerial targets.
An innovative aviation target tracking algorithm is presented in this paper, utilizing high-altitude unmanned airship dual photoelectric sensors in conjunction with Unmanned Aerial Vehicle (UAV)-borne two-coordinate radar. The algorithm addresses the challenge of integrating sensor data to accurately track targets when individual sensors lack complete target position information, thus overcoming limitations of traditional point-trace association methods. Initially, a two-level point-trace correlation algorithm based on angle and distance is introduced for multi-sensor measurement association following coordinate system transformation. Subsequently, a line-plane intersection fusion localization algorithm is proposed to determine the initial target track position through techniques such as least squares method, intersection projection, distance nearest point solution, and homologous data compression. Leveraging heterogeneous information from space-based multi-platform reconnaissance, an extended Unscented Kalman Filter (UKF) is designed to track aviation targets by enhancing the traditional UKF. Simulation results demonstrate that this algorithm achieves superior precision in tracking high-speed aerial targets.
Available online , doi: 10.11999/JEIT240210
Abstract:
With the advancement of robot automatic navigation technology, software-based path planning algorithms can no longer satisfy the needs in scenarios of many real-time applications. Fast and efficient hardware customization of the algorithm is required to achieve low-latency performance acceleration. In this work, High Level Synthesis (HLS) of classic A* algorithm is studied. Hardware-oriented data structure and function optimization, varying design constraints are explored to pick the right architecture, which is then followed by FPGA synthesis. Experimental results show that, compared to the conventional Register Transfer Level (RTL) method, the HLS-based FPGA implementation of the A* algorithm can achieve better productivity, improved hardware performance and resource utilization efficiency, which demonstrates the advantages of high level synthesis in hardware customization in algorithm-centric applications.
With the advancement of robot automatic navigation technology, software-based path planning algorithms can no longer satisfy the needs in scenarios of many real-time applications. Fast and efficient hardware customization of the algorithm is required to achieve low-latency performance acceleration. In this work, High Level Synthesis (HLS) of classic A* algorithm is studied. Hardware-oriented data structure and function optimization, varying design constraints are explored to pick the right architecture, which is then followed by FPGA synthesis. Experimental results show that, compared to the conventional Register Transfer Level (RTL) method, the HLS-based FPGA implementation of the A* algorithm can achieve better productivity, improved hardware performance and resource utilization efficiency, which demonstrates the advantages of high level synthesis in hardware customization in algorithm-centric applications.
Available online , doi: 10.11999/JEIT231324
Abstract:
In order to address the limitations of the joint beamforming method based on channel prior knowledge, which is constrained by multivariate Vehicle-to-Infrastructure (V2I) communication scenes and suffers from large overhead caused by channel estimation, a wireless propagation link prediction-based joint beamforming method assisted by environmental situation awareness is proposed in this paper. Firstly, a model of Reconfigurable Intelligent Surface (RIS) assisted mmWave communication system for V2I networks is established using a ray tracer. To build a dataset, diverse data of wireless propagation links is obtained by changing the environmental situation. Then, this dataset is used to train a machine learning-based wireless propagation link prediction model. Finally, the joint beamforming problem under the constraint of maximum transmission power is modeled. Additionally, based on the prediction outcome, the beamforming matrix of base station and the phase shift matrix of RIS are optimized using Alternating Iterative Optimization Algorithm (AIOA) to maximize the minimum Signal to Interference plus Noise Ratio (SINR) among synchronous communication vehicle users. Simulation results validate the effectiveness of the proposed method. Introducing non-channel prior knowledge driven reduces channel detection overhead and improves feasibility in applying the proposed method to V2I scenes.
In order to address the limitations of the joint beamforming method based on channel prior knowledge, which is constrained by multivariate Vehicle-to-Infrastructure (V2I) communication scenes and suffers from large overhead caused by channel estimation, a wireless propagation link prediction-based joint beamforming method assisted by environmental situation awareness is proposed in this paper. Firstly, a model of Reconfigurable Intelligent Surface (RIS) assisted mmWave communication system for V2I networks is established using a ray tracer. To build a dataset, diverse data of wireless propagation links is obtained by changing the environmental situation. Then, this dataset is used to train a machine learning-based wireless propagation link prediction model. Finally, the joint beamforming problem under the constraint of maximum transmission power is modeled. Additionally, based on the prediction outcome, the beamforming matrix of base station and the phase shift matrix of RIS are optimized using Alternating Iterative Optimization Algorithm (AIOA) to maximize the minimum Signal to Interference plus Noise Ratio (SINR) among synchronous communication vehicle users. Simulation results validate the effectiveness of the proposed method. Introducing non-channel prior knowledge driven reduces channel detection overhead and improves feasibility in applying the proposed method to V2I scenes.
Available online , doi: 10.11999/JEIT231470
Abstract:
To power Deep-Learning (DL) based Synthetic Aperture Radar Automatic Target Recognition (SAR ATR) systems with the capability of learning new-class targets incrementally and rapidly in openly dynamic non-cooperative situations, the problem of Few-Shot Class-Incremental Learning (FSCIL) of SAR ATR is researched and a Self-supervised Decoupled Dynamic Classifier (SDDC) is proposed. Considering solving both the intrinsic Catastrophic forgetting and Overfitting dilemma of the FSCIL and domain challenges of SAR ATR, a self-supervised learning task powered by Scattering Component Mixup and Rotation (SCMR) is designed to improve the model’s generalizability and stability for target representation, leveraged by the partiality and azimuth dependence of target information in SAR imagery. Meanwhile, a Class-Imprinting Cross-Entropy (CI-CE) and a Parameter Decoupled Learning (PDL) strategy are designed to fine-tune networks dynamically to identify old and new targets evenly. Experiments on various FSCIL scenarios constructed by the MSTAR and the SAR-AIRcraft-1.0 datasets covering diverse target categories, observing environments, and imaging payloads, verify the method’s adaptability to openly dynamic world.
To power Deep-Learning (DL) based Synthetic Aperture Radar Automatic Target Recognition (SAR ATR) systems with the capability of learning new-class targets incrementally and rapidly in openly dynamic non-cooperative situations, the problem of Few-Shot Class-Incremental Learning (FSCIL) of SAR ATR is researched and a Self-supervised Decoupled Dynamic Classifier (SDDC) is proposed. Considering solving both the intrinsic Catastrophic forgetting and Overfitting dilemma of the FSCIL and domain challenges of SAR ATR, a self-supervised learning task powered by Scattering Component Mixup and Rotation (SCMR) is designed to improve the model’s generalizability and stability for target representation, leveraged by the partiality and azimuth dependence of target information in SAR imagery. Meanwhile, a Class-Imprinting Cross-Entropy (CI-CE) and a Parameter Decoupled Learning (PDL) strategy are designed to fine-tune networks dynamically to identify old and new targets evenly. Experiments on various FSCIL scenarios constructed by the MSTAR and the SAR-AIRcraft-1.0 datasets covering diverse target categories, observing environments, and imaging payloads, verify the method’s adaptability to openly dynamic world.
Available online , doi: 10.11999/JEIT231436
Abstract:
Among various methods to detect the ship targets in Synthetic Aperture Radar (SAR) image, the Constant False Alarm Rate (CFAR) detection algorithm with an adaptive detection threshold is the most important and extensively used one. In order to improve the detection performance for ships in SAR image, various statistical distributions are applied, with an attempt to accurately model the SAR clutter backgrounds, such as Gamma, K, log-normal, G0, the alpha-stable distribution, etc. In modern radar systems, the use of the CFAR technique is necessary to keep the false alarms at a suitably low rate in an a priori unknown time-varying and spatially nonhomogeneous backgrounds, and to improve the detection probability as much as possible. The clutter background in SAR images is complicated and variable, when the actual clutter background deviates from the assumed statistical distribution, the performance of the parametric CFAR detectors deteriorates, whereas the nonparametric CFAR method exhibits its advantage. In this paper, the Wilcoxon nonparametric CFAR scheme for ship detection in SAR image is proposed and analyzed. By comparison with several typical parametric CFAR schemes on 3 real SAR images of Radarsat-2, ICEYE-X6 and Gaofen-3, the robustness of the Wilcoxon nonparametric detector to maintain a good false alarm performance in these different detection backgrounds is revealed, and its detection performance for the weak ship is improved evidently. Moreover, the detection speed of the Wilcoxon nonparametric detector is fast, and it has the simplicity of hardware implementation.
Among various methods to detect the ship targets in Synthetic Aperture Radar (SAR) image, the Constant False Alarm Rate (CFAR) detection algorithm with an adaptive detection threshold is the most important and extensively used one. In order to improve the detection performance for ships in SAR image, various statistical distributions are applied, with an attempt to accurately model the SAR clutter backgrounds, such as Gamma, K, log-normal, G0, the alpha-stable distribution, etc. In modern radar systems, the use of the CFAR technique is necessary to keep the false alarms at a suitably low rate in an a priori unknown time-varying and spatially nonhomogeneous backgrounds, and to improve the detection probability as much as possible. The clutter background in SAR images is complicated and variable, when the actual clutter background deviates from the assumed statistical distribution, the performance of the parametric CFAR detectors deteriorates, whereas the nonparametric CFAR method exhibits its advantage. In this paper, the Wilcoxon nonparametric CFAR scheme for ship detection in SAR image is proposed and analyzed. By comparison with several typical parametric CFAR schemes on 3 real SAR images of Radarsat-2, ICEYE-X6 and Gaofen-3, the robustness of the Wilcoxon nonparametric detector to maintain a good false alarm performance in these different detection backgrounds is revealed, and its detection performance for the weak ship is improved evidently. Moreover, the detection speed of the Wilcoxon nonparametric detector is fast, and it has the simplicity of hardware implementation.
Available online , doi: 10.11999/JEIT240007
Abstract:
A sum-rate maximization method is proposed for two Simultaneously Transmitting And Reflecting Reconfigurable Intelligent Surface (STAR-RIS) assisted downlink Non-Orthogonal Multiple Access (NOMA) systems. Firstly, the optimization problem for maximizing sum rate is constructed, with STAR-RIS phase shifts, power allocation and time allocation as optimization parameters. Then the Semi-Definite Programming method (SDP) is used to optimize the phase shifts of these two STAR-RISs. Finally, the power allocation and time allocation are optimized alternately by iterative method. In each iteration, Lagrange dual decomposition method is used to optimize power allocation and function extremum method is used to optimize time allocation. Simulation results show that the sum rate of the dual STAR-RIS-assisted NOMA system is higher than that of the single STAR-RIS-assisted NOMA system.
A sum-rate maximization method is proposed for two Simultaneously Transmitting And Reflecting Reconfigurable Intelligent Surface (STAR-RIS) assisted downlink Non-Orthogonal Multiple Access (NOMA) systems. Firstly, the optimization problem for maximizing sum rate is constructed, with STAR-RIS phase shifts, power allocation and time allocation as optimization parameters. Then the Semi-Definite Programming method (SDP) is used to optimize the phase shifts of these two STAR-RISs. Finally, the power allocation and time allocation are optimized alternately by iterative method. In each iteration, Lagrange dual decomposition method is used to optimize power allocation and function extremum method is used to optimize time allocation. Simulation results show that the sum rate of the dual STAR-RIS-assisted NOMA system is higher than that of the single STAR-RIS-assisted NOMA system.
Available online , doi: 10.11999/JEIT231033
Abstract:
The system energy consumption minimization problem is studied for a data compression based Non-Orthogonal Multiple Access-Mobile Edge Computing (NOMA-MEC) system. Considering the partial compression and offloading schemes and the limited computation capacity at the base station, a system energy consumption minimization optimization problem is formulated by jointly optimizing the users’ data compression and offloading ratios, transmit power, data compression time, etc. In order to solve this problem, closed-form expression of each user’s optimal transmit power is firstly derived. Then the Successive Convex Approximation (SCA) method is used to approximate the non-convex constraints of the formulated problem, and An SCA based efficient iterative algorithm is proposed to solve the formulated problem, obtaining the optimal resource allocation scheme of the system. Finally, the simulation results verify the advantages of the proposed scheme via computer simulations and show that compared with other benchmark schemes, the proposed scheme can effectively reduce the system energy consumption.
The system energy consumption minimization problem is studied for a data compression based Non-Orthogonal Multiple Access-Mobile Edge Computing (NOMA-MEC) system. Considering the partial compression and offloading schemes and the limited computation capacity at the base station, a system energy consumption minimization optimization problem is formulated by jointly optimizing the users’ data compression and offloading ratios, transmit power, data compression time, etc. In order to solve this problem, closed-form expression of each user’s optimal transmit power is firstly derived. Then the Successive Convex Approximation (SCA) method is used to approximate the non-convex constraints of the formulated problem, and An SCA based efficient iterative algorithm is proposed to solve the formulated problem, obtaining the optimal resource allocation scheme of the system. Finally, the simulation results verify the advantages of the proposed scheme via computer simulations and show that compared with other benchmark schemes, the proposed scheme can effectively reduce the system energy consumption.
Available online , doi: 10.11999/JEIT230953
Abstract:
With the growing use of smart devices, the ease of sharing digital media content has been enhanced. Concerns have been raised about unauthorized access, particularly via screen shooting. In this paper, a novel end-to-end watermarking framework is proposed, employing invertible neural networks and inverse gradient attention, to tackle the copyright infringement challenges related to screen content leakage. A single invertible neural network is employed by the proposed method for watermark embedding and extraction, ensuring information integrity during network propagation. Additionally, robustness and visual quality are enhanced by an inverse gradient attention module, which emphasizes pixel values and embeds the watermark in imperceptible areas for better invisibility and model resilience. Model parameters are optimized using the Learnable Perceptual Image Patch Similarity (LPIPS) loss function, minimizing perception differences in watermarked images. The superiority of this approach over existing learning-based screen-shooting resilient watermarking methods in terms of robustness and visual quality is demonstrated by experimental results.
With the growing use of smart devices, the ease of sharing digital media content has been enhanced. Concerns have been raised about unauthorized access, particularly via screen shooting. In this paper, a novel end-to-end watermarking framework is proposed, employing invertible neural networks and inverse gradient attention, to tackle the copyright infringement challenges related to screen content leakage. A single invertible neural network is employed by the proposed method for watermark embedding and extraction, ensuring information integrity during network propagation. Additionally, robustness and visual quality are enhanced by an inverse gradient attention module, which emphasizes pixel values and embeds the watermark in imperceptible areas for better invisibility and model resilience. Model parameters are optimized using the Learnable Perceptual Image Patch Similarity (LPIPS) loss function, minimizing perception differences in watermarked images. The superiority of this approach over existing learning-based screen-shooting resilient watermarking methods in terms of robustness and visual quality is demonstrated by experimental results.
Available online , doi: 10.11999/JEIT230763
Abstract:
The Multivariate Variational Mode Decomposition (MVMD), being an extension of the Variational Mode Decomposition (VMD), inherits the merits of VMD. However, it encounters an issue wherein its decomposition performance relies heavily on two predefined parameters, the number of modes (K) and the penalty factor (\begin{document}$ \alpha $\end{document} \begin{document}$ \alpha $\end{document} \begin{document}$ \alpha $\end{document}
The Multivariate Variational Mode Decomposition (MVMD), being an extension of the Variational Mode Decomposition (VMD), inherits the merits of VMD. However, it encounters an issue wherein its decomposition performance relies heavily on two predefined parameters, the number of modes (K) and the penalty factor (
Available online , doi: 10.11999/JEIT230719
Abstract:
Salient Object Detection (SOD) aims to recognize and segment visual salient objects in images, which is one of the important research contents in computer vision tasks and related fields. Existing Fully Convolutional Networks (FCNs)-based SOD methods have achieved good performance. However, the types and sizes of salient objects are variable and unfixed in real-world scenes, which makes it still a huge challenge to detect and segment salient objects accurately and completely. For that, in this paper, a novel integrating multiple context and hybrid interaction for SOD task is proposed to efficiently predict salient objects by collaborating Dense Context Information Exploration (DCIE) module and Multi-source Feature Hybrid Interaction (MFHI) module. The DCIE module uses dilated convolution, asymmetric convolution and dense guided connection to progressively capture the strongly correlated multi-scale and multi-receptive field context information, and enhances the expression ability of each initial input feature by aggregating context information. The MFHI module contains diverse feature aggregation operations, which can adaptively interact with complementary information from multi-level features to generate high-quality feature representations for accurately predicting saliency maps. The performance of the proposed method is tested on five public datasets. The performance of the proposed method is tested on five public datasets. Experimental results demonstrate that our method achieves superior prediction performance compared with 19 state-of-the-art SOD methods under different evaluation metrics.
Salient Object Detection (SOD) aims to recognize and segment visual salient objects in images, which is one of the important research contents in computer vision tasks and related fields. Existing Fully Convolutional Networks (FCNs)-based SOD methods have achieved good performance. However, the types and sizes of salient objects are variable and unfixed in real-world scenes, which makes it still a huge challenge to detect and segment salient objects accurately and completely. For that, in this paper, a novel integrating multiple context and hybrid interaction for SOD task is proposed to efficiently predict salient objects by collaborating Dense Context Information Exploration (DCIE) module and Multi-source Feature Hybrid Interaction (MFHI) module. The DCIE module uses dilated convolution, asymmetric convolution and dense guided connection to progressively capture the strongly correlated multi-scale and multi-receptive field context information, and enhances the expression ability of each initial input feature by aggregating context information. The MFHI module contains diverse feature aggregation operations, which can adaptively interact with complementary information from multi-level features to generate high-quality feature representations for accurately predicting saliency maps. The performance of the proposed method is tested on five public datasets. The performance of the proposed method is tested on five public datasets. Experimental results demonstrate that our method achieves superior prediction performance compared with 19 state-of-the-art SOD methods under different evaluation metrics.
Available online , doi: 10.11999/JEIT230986
Abstract:
In recent years, mounting Mobile Edge Computing (MEC) servers on Unmanned Aerial Vehicle (UAV) to provide services for mobile ground users has been widely researched in academia and industry. However, in malicious jamming environments, how to effectively schedule resources to reduce system delay and energy consumption becomes a key challenge. Therefore, this paper considers a UAV-assisted MEC system under a malicious jammer, where an optimization model is established to minimize the weighted energy consumption and delay by jointly optimizing UAV flight trajectories, resource scheduling, and task allocation. As the optimization problem is difficult to be solved and the malicious jamming behavior is dynamic, a Twin Delayed Deep Deterministic (TD3) policy gradient algorithm is proposed to search for the optimal policy. At the same time, the Prioritized Experience Replay (PER) technique is added to improve the convergence speed and stability of the algorithm, which is highly effective against malicious interference attacks. The simulation results show that the proposed algorithm can effectively reduce the delay and energy consumption, and achieve good convergence and stability compared with other algorithms.
In recent years, mounting Mobile Edge Computing (MEC) servers on Unmanned Aerial Vehicle (UAV) to provide services for mobile ground users has been widely researched in academia and industry. However, in malicious jamming environments, how to effectively schedule resources to reduce system delay and energy consumption becomes a key challenge. Therefore, this paper considers a UAV-assisted MEC system under a malicious jammer, where an optimization model is established to minimize the weighted energy consumption and delay by jointly optimizing UAV flight trajectories, resource scheduling, and task allocation. As the optimization problem is difficult to be solved and the malicious jamming behavior is dynamic, a Twin Delayed Deep Deterministic (TD3) policy gradient algorithm is proposed to search for the optimal policy. At the same time, the Prioritized Experience Replay (PER) technique is added to improve the convergence speed and stability of the algorithm, which is highly effective against malicious interference attacks. The simulation results show that the proposed algorithm can effectively reduce the delay and energy consumption, and achieve good convergence and stability compared with other algorithms.
Available online , doi: 10.11999/JEIT230887
Abstract:
Due to the complex marine environment, it is difficult for a maritime radar to achieve high-performance detection of slow and small targets on the sea surface. For such targets, the traditional energy-based statistical detection algorithms suffer from serious performance loss. Confronted with this problem, a detection algorithm of small targets based on Deep Graph Infomax framework is proposed to realize unsupervised target anomaly detection in the background of sea clutter. In the traditional neural networks, there is an assumption that the samples are independent and identically distributed, which, however, the high-resolution radar echo does not meet. Therefore, this paper re-models the data from the perspective of graph and constructs the graph topological structure according to the correlation characteristics of the echo. Moreover, this paper puts forward the relative maximum node degree, and combines it with the relative average amplitude and the relative Doppler vector entropy to be the initial representation vectors of the graph nodes. With the graph modeling done, the graph attention network is used as the encoder in the Deep Graph Infomax framework to learn representation vectors. Finally, the anomaly detection algorithm is used to detect the targets, and the false alarm can be controlled. The detection result on the measured datasets shows that the performance of the proposed detector is improved by 9.2% compared to the three-feature detector when using the fast convex hull learning algorithm. Compared to the time-frequency three-feature detector, the performance is improved by 7.9%. When the network outputs a higher-dimensional representation vectors, the performance of the detector using the isolated forest algorithm is improved by 27.4%.
Due to the complex marine environment, it is difficult for a maritime radar to achieve high-performance detection of slow and small targets on the sea surface. For such targets, the traditional energy-based statistical detection algorithms suffer from serious performance loss. Confronted with this problem, a detection algorithm of small targets based on Deep Graph Infomax framework is proposed to realize unsupervised target anomaly detection in the background of sea clutter. In the traditional neural networks, there is an assumption that the samples are independent and identically distributed, which, however, the high-resolution radar echo does not meet. Therefore, this paper re-models the data from the perspective of graph and constructs the graph topological structure according to the correlation characteristics of the echo. Moreover, this paper puts forward the relative maximum node degree, and combines it with the relative average amplitude and the relative Doppler vector entropy to be the initial representation vectors of the graph nodes. With the graph modeling done, the graph attention network is used as the encoder in the Deep Graph Infomax framework to learn representation vectors. Finally, the anomaly detection algorithm is used to detect the targets, and the false alarm can be controlled. The detection result on the measured datasets shows that the performance of the proposed detector is improved by 9.2% compared to the three-feature detector when using the fast convex hull learning algorithm. Compared to the time-frequency three-feature detector, the performance is improved by 7.9%. When the network outputs a higher-dimensional representation vectors, the performance of the detector using the isolated forest algorithm is improved by 27.4%.
Available online , doi: 10.11999/JEIT230966
Abstract:
A hybrid detector AEM-YOLO based on the adaptive fusion of different scale features is proposed, aiming at the problems of difficult detection of small objects in remote sensing images caused by the high background noise, dense arrangement of small objects, and wide-scale distribution. Firstly, a two-axes k-means clustering algorithm combining width and height information with scale and ratio information is proposed to generate anchors with high matching degrees with remote sensing datasets. Secondly, an adaptive enhance module is designed to address information conflicts caused by direct fusion between different scale features. A lower feature layer is introduced to broadcast small object details along the bottom-up path. By using multi-task learning and scale guidance factor, the recall for objects with a high aspect ratio can be effectively improved. Finally, the experiments on the DIOR dataset show that compared with the original model, the AP of AEM-YOLO is improved by 7.8%, and increased by 5.4%, 7.2%, and 8.6% in small, medium, and large object detection, respectively.
A hybrid detector AEM-YOLO based on the adaptive fusion of different scale features is proposed, aiming at the problems of difficult detection of small objects in remote sensing images caused by the high background noise, dense arrangement of small objects, and wide-scale distribution. Firstly, a two-axes k-means clustering algorithm combining width and height information with scale and ratio information is proposed to generate anchors with high matching degrees with remote sensing datasets. Secondly, an adaptive enhance module is designed to address information conflicts caused by direct fusion between different scale features. A lower feature layer is introduced to broadcast small object details along the bottom-up path. By using multi-task learning and scale guidance factor, the recall for objects with a high aspect ratio can be effectively improved. Finally, the experiments on the DIOR dataset show that compared with the original model, the AP of AEM-YOLO is improved by 7.8%, and increased by 5.4%, 7.2%, and 8.6% in small, medium, and large object detection, respectively.
Available online , doi: 10.11999/JEIT231421
Abstract:
Non-Line-Of-Sight (NLOS) propagation will cause the pseudo-range measurement error of the Global Navigation Satellite System (GNSS) receivers, and eventually lead to a large positioning error, which is especially prominent in complex environments such as urban canyons. To solve this problem, a robust positioning method for Kernel Density Estimation (KDE) is proposed. The core idea is to introduce robust estimation theory into localization solution to alleviate the influence of NLOS. Considering that the pseudo-range observation error caused by NLOS deviates from the Gaussian distribution, the proposed method firstly uses the method based on KDE to estimate the probability density function of the observation error, and then uses the probability density function to construct a robust cost function for positioning solution, so as to alleviate the positioning error caused by NLOS. The experimental results show that the proposed method can effectively reduce GNSS positioning error in NLOS propagation environment.
Non-Line-Of-Sight (NLOS) propagation will cause the pseudo-range measurement error of the Global Navigation Satellite System (GNSS) receivers, and eventually lead to a large positioning error, which is especially prominent in complex environments such as urban canyons. To solve this problem, a robust positioning method for Kernel Density Estimation (KDE) is proposed. The core idea is to introduce robust estimation theory into localization solution to alleviate the influence of NLOS. Considering that the pseudo-range observation error caused by NLOS deviates from the Gaussian distribution, the proposed method firstly uses the method based on KDE to estimate the probability density function of the observation error, and then uses the probability density function to construct a robust cost function for positioning solution, so as to alleviate the positioning error caused by NLOS. The experimental results show that the proposed method can effectively reduce GNSS positioning error in NLOS propagation environment.
Available online , doi: 10.11999/JEIT240149
Abstract:
Passive intelligent reflecting surfaces hold great potential in enhancing wireless communication systems and improving physical layer security, but they suffer from significant drawbacks such as “double fading” and partial coverage. Therefore, an active Simultaneously Transmitting And Reflecting Reconfigurable Intelligent Surface (STAR-RIS) is conducted in this paper. Considering the coupling between reflection and transmission phase shifts, a joint optimization problem for maximizing the security energy efficiency of base station beams and active STAR-RIS beams is formulated. To solve the resulting non-convex optimization problem, continuous convex approximation, penalty function method, semi-definite relaxation, and alternating optimization techniques are employed to transform the original problem into a convex one. Additionally, a penalty dual decomposition algorithm is proposed. Simulation results validate the effectiveness of the algorithm proposed in this paper.
Passive intelligent reflecting surfaces hold great potential in enhancing wireless communication systems and improving physical layer security, but they suffer from significant drawbacks such as “double fading” and partial coverage. Therefore, an active Simultaneously Transmitting And Reflecting Reconfigurable Intelligent Surface (STAR-RIS) is conducted in this paper. Considering the coupling between reflection and transmission phase shifts, a joint optimization problem for maximizing the security energy efficiency of base station beams and active STAR-RIS beams is formulated. To solve the resulting non-convex optimization problem, continuous convex approximation, penalty function method, semi-definite relaxation, and alternating optimization techniques are employed to transform the original problem into a convex one. Additionally, a penalty dual decomposition algorithm is proposed. Simulation results validate the effectiveness of the algorithm proposed in this paper.
Available online , doi: 10.11999/JEIT240172
Abstract:
The significant advancement of deep learning has facilitated the emergence of high-precision interpretation models for remote-sensing images. However, a notable drawback is that the majority of interpretation models are trained independently on static datasets, rendering them incapable of adapting to open environments and dynamic demands. This limitation poses a substantial obstacle to the widespread and long-term application of remote-sensing interpretation models. Incremental learning, empowering models to continuously learn new knowledge while retaining previous knowledge, has been recently utilized to drive the evolution of interpretation models and improve their performance. A comprehensive investigation of incremental learning methods for multi-modal remote sensing data and diverse interpretation tasks is provided in this paper. Existing research efforts are organized and reviewed in terms of mitigating catastrophic forgetting and facilitating interpretation model evolution. Drawing from this research progress, this study deliberates on the future research directions for incremental learning in remote sensing, with the aim of advancing research in model evolution for remote sensing image interpretation.
The significant advancement of deep learning has facilitated the emergence of high-precision interpretation models for remote-sensing images. However, a notable drawback is that the majority of interpretation models are trained independently on static datasets, rendering them incapable of adapting to open environments and dynamic demands. This limitation poses a substantial obstacle to the widespread and long-term application of remote-sensing interpretation models. Incremental learning, empowering models to continuously learn new knowledge while retaining previous knowledge, has been recently utilized to drive the evolution of interpretation models and improve their performance. A comprehensive investigation of incremental learning methods for multi-modal remote sensing data and diverse interpretation tasks is provided in this paper. Existing research efforts are organized and reviewed in terms of mitigating catastrophic forgetting and facilitating interpretation model evolution. Drawing from this research progress, this study deliberates on the future research directions for incremental learning in remote sensing, with the aim of advancing research in model evolution for remote sensing image interpretation.
Available online , doi: 10.11999/JEIT240142
Abstract:
Unsupervised Continual Learning(UCL) refers to the ability to learn over time while remembering previous patterns without supervision. Although significant progress has been made in this direction, existing works often assume strong prior knowledge about forthcoming data (e.g., knowing class boundaries), which may not be obtainable in complex and unpredictable open environments. Inspired by real-world scenarios, a more practical problem setting called unsupervised online continual learning without prior knowledge is proposed in this paper. The proposed setting is challenging because the data are non-i.i.d. and lack external supervision or prior knowledge. To address these challenges, a method called EvolveNet is intriduced, which is an adaptive unsupervised continual learning approach capable of purely extracting and memorizing representations from data streams. EvolveNet is designed around three main components: adversarial pseudo-supervised learning loss, self-supervised forgetting loss, and online memory update for uniform subset selection. The design of these three components aims to synergize and maximize learning performance. We conduct comprehensive experiments on five public datasets with EvolveNet. The results show that EvolveNet outperforms existing algorithms in all settings, achieving significantly improved accuracy on CIFAR-10, CIFAR-100, and TinyImageNet datasets, as well as performing best on the multimodal datasets Core-50 and iLab-20M for incremental learning. We also conduct cross-dataset generalization experiments, demonstrating EvolveNet’s robustness in generalization. Finally, we open-source the EvolveNet model and core code on GitHub, facilitating progress in unsupervised continual learning and providing a useful tool and platform for the research community.
Unsupervised Continual Learning(UCL) refers to the ability to learn over time while remembering previous patterns without supervision. Although significant progress has been made in this direction, existing works often assume strong prior knowledge about forthcoming data (e.g., knowing class boundaries), which may not be obtainable in complex and unpredictable open environments. Inspired by real-world scenarios, a more practical problem setting called unsupervised online continual learning without prior knowledge is proposed in this paper. The proposed setting is challenging because the data are non-i.i.d. and lack external supervision or prior knowledge. To address these challenges, a method called EvolveNet is intriduced, which is an adaptive unsupervised continual learning approach capable of purely extracting and memorizing representations from data streams. EvolveNet is designed around three main components: adversarial pseudo-supervised learning loss, self-supervised forgetting loss, and online memory update for uniform subset selection. The design of these three components aims to synergize and maximize learning performance. We conduct comprehensive experiments on five public datasets with EvolveNet. The results show that EvolveNet outperforms existing algorithms in all settings, achieving significantly improved accuracy on CIFAR-10, CIFAR-100, and TinyImageNet datasets, as well as performing best on the multimodal datasets Core-50 and iLab-20M for incremental learning. We also conduct cross-dataset generalization experiments, demonstrating EvolveNet’s robustness in generalization. Finally, we open-source the EvolveNet model and core code on GitHub, facilitating progress in unsupervised continual learning and providing a useful tool and platform for the research community.
Available online , doi: 10.11999/JEIT230945
Abstract:
Considering the trajectory prediction problem of drift buoys, an end-to-end prediction model based on the depth learning framework is proposed in this paper.The hydrodynamic models in different sea areas are quite different, and the calculation of fluid load of floating buoys on the sea surface is also complicated. Therefore, a more universal data-driven trajectory prediction model based on the multidimensional time series formed by the historical trajectories of drifting buoys is proposed. In this model, Particle Swarm Optimization (PSO) is combined with Gated Recurrent Unit (GRU), and the PSO is used to initialize the hyperparameters of the GRU neural network. The optimal drifting buoy trajectory prediction model is obtained after multiple migration iteration training. Finally, several real drifting buoy track data in the North Atlantic are used to verify the results. The results show that the PSOGRU algorithm can achieve accurate drifting buoy track prediction results.
Considering the trajectory prediction problem of drift buoys, an end-to-end prediction model based on the depth learning framework is proposed in this paper.The hydrodynamic models in different sea areas are quite different, and the calculation of fluid load of floating buoys on the sea surface is also complicated. Therefore, a more universal data-driven trajectory prediction model based on the multidimensional time series formed by the historical trajectories of drifting buoys is proposed. In this model, Particle Swarm Optimization (PSO) is combined with Gated Recurrent Unit (GRU), and the PSO is used to initialize the hyperparameters of the GRU neural network. The optimal drifting buoy trajectory prediction model is obtained after multiple migration iteration training. Finally, several real drifting buoy track data in the North Atlantic are used to verify the results. The results show that the PSOGRU algorithm can achieve accurate drifting buoy track prediction results.
Available online , doi: 10.11999/JEIT231103
Abstract:
In order to balance the transmission reliability and efficiency of large-scale multi-mode mesh networks in the new power system, a two-stage algorithm is proposed based on reinforcement learning for joint routing selection and resource scheduling in large-scale multi-mode mesh networks, building upon the description and analysis of optimization problems. In the first stage, based on the network topology information and service requirements, a multi shortest path routing algorithm is utilized to generate all the shortest paths. In the second stage, a resource scheduling algorithm based on Multi-Armed Bandit (MAB) is proposed. The algorithm constructs the arms of the MAB based on the obtained set of shortest paths, then calculates the reward according to the service demands, and finally gives the optimal route selection and resource scheduling mode for service transmission. Simulation results show that the proposed algorithm can meet different service transmission requirements, and achieve an efficient balance between the average end-to-end path delay and the average transmission success rate.
In order to balance the transmission reliability and efficiency of large-scale multi-mode mesh networks in the new power system, a two-stage algorithm is proposed based on reinforcement learning for joint routing selection and resource scheduling in large-scale multi-mode mesh networks, building upon the description and analysis of optimization problems. In the first stage, based on the network topology information and service requirements, a multi shortest path routing algorithm is utilized to generate all the shortest paths. In the second stage, a resource scheduling algorithm based on Multi-Armed Bandit (MAB) is proposed. The algorithm constructs the arms of the MAB based on the obtained set of shortest paths, then calculates the reward according to the service demands, and finally gives the optimal route selection and resource scheduling mode for service transmission. Simulation results show that the proposed algorithm can meet different service transmission requirements, and achieve an efficient balance between the average end-to-end path delay and the average transmission success rate.
Available online , doi: 10.11999/JEIT230999
Abstract:
In the radar target adaptive detection problem, the presence of clutter edges in the auxiliary data will cause a serious decrease in the estimation performance of the Clutter Covariance Matrix (CCM), which greatly affects the target detection performance. In order to solve this problem, a clutter edge detection method is proposed, which can adaptively discriminate the number and position of clutter edges in auxiliary data. Firstly, assuming the presence of clutter edges in the auxiliary data, the model order selection algorithm and the maximum likelihood estimation method are used to complete the clutter parameter estimation, and the clutter edge position is obtained by the cyclic search method. Then, the clutter parameter estimation results are applied to the detection algorithm, and the existence of clutter edges is determined by the generalized likelihood ratio test method. In addition, in order to further improve the robustness of the algorithm under the condition of small samples, the special structure of CCM is introduced as a priori knowledge, and the algorithm is generalized to the situation where CCM is persymmetry, spectrum symmetry and central-symmetry. Both simulation and measured data show that the proposed algorithm can efficiently identify the number and location of clutter edges in radar auxiliary data, and the introduction of prior knowledge can further improve the performance of the algorithm when the amount of auxiliary data is small.
In the radar target adaptive detection problem, the presence of clutter edges in the auxiliary data will cause a serious decrease in the estimation performance of the Clutter Covariance Matrix (CCM), which greatly affects the target detection performance. In order to solve this problem, a clutter edge detection method is proposed, which can adaptively discriminate the number and position of clutter edges in auxiliary data. Firstly, assuming the presence of clutter edges in the auxiliary data, the model order selection algorithm and the maximum likelihood estimation method are used to complete the clutter parameter estimation, and the clutter edge position is obtained by the cyclic search method. Then, the clutter parameter estimation results are applied to the detection algorithm, and the existence of clutter edges is determined by the generalized likelihood ratio test method. In addition, in order to further improve the robustness of the algorithm under the condition of small samples, the special structure of CCM is introduced as a priori knowledge, and the algorithm is generalized to the situation where CCM is persymmetry, spectrum symmetry and central-symmetry. Both simulation and measured data show that the proposed algorithm can efficiently identify the number and location of clutter edges in radar auxiliary data, and the introduction of prior knowledge can further improve the performance of the algorithm when the amount of auxiliary data is small.
Available online , doi: 10.11999/JEIT231004
Abstract:
To achieve formation control of multiple snake robots, an error-constrained anti-interference path-following method is proposed in this paper. A highly coupled dynamic frequency compensator is used to adjust the motion speed of each robot to ensure consistency in the position and velocity of the formation members. In dynamic control, the singularity phenomenon of virtual variables is eliminated by the equivalent principle of barrier functions, improving the stability of path following. In addition, predictive values for model uncertainty and external interference are designed to pre-compensate for joint offsets and torque inputs of the robots, further improving the convergence rate and steady-state performance of the following errors. Finally, the Lyapunov theory is used to prove the Uniform Ultimate Boundedness (UUB) of this system. Simulation data demonstrate that the proposed method and control strategy have higher following accuracy compared to other classic methods.
To achieve formation control of multiple snake robots, an error-constrained anti-interference path-following method is proposed in this paper. A highly coupled dynamic frequency compensator is used to adjust the motion speed of each robot to ensure consistency in the position and velocity of the formation members. In dynamic control, the singularity phenomenon of virtual variables is eliminated by the equivalent principle of barrier functions, improving the stability of path following. In addition, predictive values for model uncertainty and external interference are designed to pre-compensate for joint offsets and torque inputs of the robots, further improving the convergence rate and steady-state performance of the following errors. Finally, the Lyapunov theory is used to prove the Uniform Ultimate Boundedness (UUB) of this system. Simulation data demonstrate that the proposed method and control strategy have higher following accuracy compared to other classic methods.
Available online , doi: 10.11999/JEIT231016
Abstract:
The Space-Air-Ground Integrated Network (SAGIN) can effectively meet the communication needs of various service types by improving the resource utilization of the ground network, but ignoring the adaptive ability and robustness of the system and the Quality of Service (QoS) in different users. In response to this problem, a Deep Reinforcement Learning (DRL) Resource allocation algorithm for urban and suburban communications under the SAGIN architecture is proposed in this paper. Based on Reference Signal Reception Power (RSRP) defined in the 3rd Generation Partnership Project (3GPP) standard, considering ground co-frequency interference, and using the time-frequency resources of base stations in different domains as constraints, an optimization problem to maxmize the downlink throughput of system users is constructed. When using the Deep Q-network (DQN) algorithm to solve the optimization problem, a reward function which can comprehensively consider the user’s QoS requirements, system adaptability and system robustness is defined. Considering the service requirements of unmanned vehicles, immersive services and ordinary mobile communication services, the simulation results show that the value of the reward function which represents the performance of the system is increased by 39.1% compared with the greedy algorithm under 2 000 iterations. For the unmanned vehicle services, the average packet loss rate by the DQN algorithm is 38.07% lower than that by the greedy algorithm, and the delay by the DQN algorithm is also 6.05% lower than that by the greedy algorithm.
The Space-Air-Ground Integrated Network (SAGIN) can effectively meet the communication needs of various service types by improving the resource utilization of the ground network, but ignoring the adaptive ability and robustness of the system and the Quality of Service (QoS) in different users. In response to this problem, a Deep Reinforcement Learning (DRL) Resource allocation algorithm for urban and suburban communications under the SAGIN architecture is proposed in this paper. Based on Reference Signal Reception Power (RSRP) defined in the 3rd Generation Partnership Project (3GPP) standard, considering ground co-frequency interference, and using the time-frequency resources of base stations in different domains as constraints, an optimization problem to maxmize the downlink throughput of system users is constructed. When using the Deep Q-network (DQN) algorithm to solve the optimization problem, a reward function which can comprehensively consider the user’s QoS requirements, system adaptability and system robustness is defined. Considering the service requirements of unmanned vehicles, immersive services and ordinary mobile communication services, the simulation results show that the value of the reward function which represents the performance of the system is increased by 39.1% compared with the greedy algorithm under 2 000 iterations. For the unmanned vehicle services, the average packet loss rate by the DQN algorithm is 38.07% lower than that by the greedy algorithm, and the delay by the DQN algorithm is also 6.05% lower than that by the greedy algorithm.
Available online , doi: 10.11999/JEIT230898
Abstract:
Considering the problem of unknown and highly heterogeneous user preference in the current edge caching scenario, a joint optimization strategy of user request perceived edge caching and user recommendation is proposed. Firstly, the basic model of Click Through Rate (CTR) prediction is established, and the contrastive learning method is introduced to generate high-quality feature representation, which could better help Factorization Machine(FM) model to predict user preference. Then, based on the predicted user preference, a dynamic recommendation mechanism is designed to reshape the content request probability of different users, thereby affecting cache decision; Finally, a joint optimization problem of edge caching and user recommendation is established with the goal of minimizing the average user content acquisition delay. It is decoupled into edge caching subproblem and user recommendation subproblem, and solved based on the region greedy algorithm and one-to-one exchange matching algorithm, respectively. The convergence optimization results are obtained through iterative update. The results show that compared with the benchmark model, the contrastive learning method has improved Area Under Curve (AUC) and ACCuracy (ACC) by 1.65% and 1.30%, respectively, and the joint optimization algorithm can effectively reduce the average content acquisition latency of users and improve the system cache performance.
Considering the problem of unknown and highly heterogeneous user preference in the current edge caching scenario, a joint optimization strategy of user request perceived edge caching and user recommendation is proposed. Firstly, the basic model of Click Through Rate (CTR) prediction is established, and the contrastive learning method is introduced to generate high-quality feature representation, which could better help Factorization Machine(FM) model to predict user preference. Then, based on the predicted user preference, a dynamic recommendation mechanism is designed to reshape the content request probability of different users, thereby affecting cache decision; Finally, a joint optimization problem of edge caching and user recommendation is established with the goal of minimizing the average user content acquisition delay. It is decoupled into edge caching subproblem and user recommendation subproblem, and solved based on the region greedy algorithm and one-to-one exchange matching algorithm, respectively. The convergence optimization results are obtained through iterative update. The results show that compared with the benchmark model, the contrastive learning method has improved Area Under Curve (AUC) and ACCuracy (ACC) by 1.65% and 1.30%, respectively, and the joint optimization algorithm can effectively reduce the average content acquisition latency of users and improve the system cache performance.
Available online , doi: 10.11999/JEIT231111
Abstract:
Short Quasi-Cyclic (QC) Low-Density Parity-Check (LDPC) codes without small cycles suitable for an arbitrary row weight (i.e., Row-Weight Universal (RWU)), are of great significance for both theoretical research and engineering application. Existing methods having RWU property and guaranteeing the nonexistence of 4-cycles and 6-cycles, can only offer short QC-LDPC codes for the column weights of 3 and 4. Based on the Greatest Common Divisor (GCD) framework, three new methods are proposed in this paper for the column weights of 5 and 6, which can possess RWU property and at the same time remove all 4-cycles and 6-cycles. Compared with existing methods with RWU property, the code lengths of the novel methods are sharply reduced from the fourth power of row weight to the third power of row weight. Therefore, the new methods can provide short RWU QC-LDPC codes without 4-cycles and 6-cycles for occasions where base codes with large column weights are required, such as composite constructions and advanced optimization pertaining to QC-LDPC codes. Moreover, compared with the search-based symmetric QC-LDPC codes, the new codes need no search, have lower description complexity, and exhibit better decoding performance.
Short Quasi-Cyclic (QC) Low-Density Parity-Check (LDPC) codes without small cycles suitable for an arbitrary row weight (i.e., Row-Weight Universal (RWU)), are of great significance for both theoretical research and engineering application. Existing methods having RWU property and guaranteeing the nonexistence of 4-cycles and 6-cycles, can only offer short QC-LDPC codes for the column weights of 3 and 4. Based on the Greatest Common Divisor (GCD) framework, three new methods are proposed in this paper for the column weights of 5 and 6, which can possess RWU property and at the same time remove all 4-cycles and 6-cycles. Compared with existing methods with RWU property, the code lengths of the novel methods are sharply reduced from the fourth power of row weight to the third power of row weight. Therefore, the new methods can provide short RWU QC-LDPC codes without 4-cycles and 6-cycles for occasions where base codes with large column weights are required, such as composite constructions and advanced optimization pertaining to QC-LDPC codes. Moreover, compared with the search-based symmetric QC-LDPC codes, the new codes need no search, have lower description complexity, and exhibit better decoding performance.
Available online , doi: 10.11999/JEIT230801
Abstract:
To address the challenge of aligning multimodal data and improving the slow translation speed in sign language translation, a Transformer Sign Language Translation Non-Autoregression (Trans-SLT-NA) is proposed in this paper, which utilizes a self-attention mechanism. Additionally, it incorporates a contrastive learning loss function to align the multimodal data. By capturing the contextual and interaction information between the input sequence (sign language videos) and the target sequence (text), the proposed model is able to perform sign language translation to natural language in s single step. The effectiveness of the proposed model is evaluated on publicly available datasets, including PHOENIX-2014-T (German), CSL (Chinese) and How2Sign (English). Results demonstrate that the proposed method achieves a significant improvement in translation speed, with a speed boost ranging from 11.6 to 17.6 times compared to autoregressive models, while maintaining comparable performance in terms of BiLingual Evaluation Understudy (BLEU-4) and Recall-Oriented Understudy for Gisting Evaluation (ROUGE) metrics.
To address the challenge of aligning multimodal data and improving the slow translation speed in sign language translation, a Transformer Sign Language Translation Non-Autoregression (Trans-SLT-NA) is proposed in this paper, which utilizes a self-attention mechanism. Additionally, it incorporates a contrastive learning loss function to align the multimodal data. By capturing the contextual and interaction information between the input sequence (sign language videos) and the target sequence (text), the proposed model is able to perform sign language translation to natural language in s single step. The effectiveness of the proposed model is evaluated on publicly available datasets, including PHOENIX-2014-T (German), CSL (Chinese) and How2Sign (English). Results demonstrate that the proposed method achieves a significant improvement in translation speed, with a speed boost ranging from 11.6 to 17.6 times compared to autoregressive models, while maintaining comparable performance in terms of BiLingual Evaluation Understudy (BLEU-4) and Recall-Oriented Understudy for Gisting Evaluation (ROUGE) metrics.
Available online , doi: 10.11999/JEIT231113
Abstract:
The style transfer method, due to its excellent domain adaptation capability, is widely used to alleviate domain gap of computer vision domain. Currently, stereo matching based on style transfer faces the following challenges: (1) The transformed left and right images need to remain matched; (2) The content and spatial information of the transformed images should remain consistent with the original images. To address these challenges, an Edge Domain Adaptation Stereo matching (EDA-Stereo) method is proposed. First, an Edge-guided Generative Adversarial Network (Edge-GAN) is constructed. by incorporating edge cues and synthetic features through the Spatial Feature Transform (SFT) layer. the Edge-GAN guides the generator to produce pseudo-images that retain the structural features of syntheitic domain images. Second, a warping loss is introduced to guarantee the left image to be reconstructed based on the transformed right image to approximate the original left image, preventing mismatches between the transformed left and right images. Finally, a normal loss based stetreo matching network is proposed to capture more geometric details by characterizing local depth variations, thereby improving matching accuracy. By training on synthetic datasets and comparing with various methods on real datasets, results show the effectiveness in mitigating domain gaps. On the KITTI 2012 and KITTI 2015 datasets, the D1 error is 3.9% and 4.8%, respectively, which is a relative reduction of 37% and 26% compared to the state-of-the-art Domain-invariant Stereo Matching Networks (DSM-Net) method.
The style transfer method, due to its excellent domain adaptation capability, is widely used to alleviate domain gap of computer vision domain. Currently, stereo matching based on style transfer faces the following challenges: (1) The transformed left and right images need to remain matched; (2) The content and spatial information of the transformed images should remain consistent with the original images. To address these challenges, an Edge Domain Adaptation Stereo matching (EDA-Stereo) method is proposed. First, an Edge-guided Generative Adversarial Network (Edge-GAN) is constructed. by incorporating edge cues and synthetic features through the Spatial Feature Transform (SFT) layer. the Edge-GAN guides the generator to produce pseudo-images that retain the structural features of syntheitic domain images. Second, a warping loss is introduced to guarantee the left image to be reconstructed based on the transformed right image to approximate the original left image, preventing mismatches between the transformed left and right images. Finally, a normal loss based stetreo matching network is proposed to capture more geometric details by characterizing local depth variations, thereby improving matching accuracy. By training on synthetic datasets and comparing with various methods on real datasets, results show the effectiveness in mitigating domain gaps. On the KITTI 2012 and KITTI 2015 datasets, the D1 error is 3.9% and 4.8%, respectively, which is a relative reduction of 37% and 26% compared to the state-of-the-art Domain-invariant Stereo Matching Networks (DSM-Net) method.
Available online , doi: 10.11999/JEIT231118
Abstract:
In 5G Vehicular to Everything (5G-V2X), service messages are provided to a group of vehicles belonging to a specific region by means of point-to-multipoint transmission. To address security threats and privacy leakage, an authentication and key negotiation scheme is proposed for multicast service message transmission between content providers and vehicles. A certificate-less aggregated signature technique is used to batch-verify all vehicles in the group, and improves the efficiency of authentication requests. Secure key negotiation is realized based on the polynomial key management technique, which makes it impossible for illegal users or the core network to obtain the shared session key. Finally, a dynamic key update mechanism for vehicles in the group is implemented, so that when a vehicle joins or leaves the group, the content provider only needs to send a key update message to update the session key. The proposed scheme can guarantee security requirements such as anonymity, unlinkability, forward and backward security, and resistance to conspiracy attacks, as shown by formal verification tools and further security analysis. The computational efficiency is improved by about 34.2% compared to existing schemes.
In 5G Vehicular to Everything (5G-V2X), service messages are provided to a group of vehicles belonging to a specific region by means of point-to-multipoint transmission. To address security threats and privacy leakage, an authentication and key negotiation scheme is proposed for multicast service message transmission between content providers and vehicles. A certificate-less aggregated signature technique is used to batch-verify all vehicles in the group, and improves the efficiency of authentication requests. Secure key negotiation is realized based on the polynomial key management technique, which makes it impossible for illegal users or the core network to obtain the shared session key. Finally, a dynamic key update mechanism for vehicles in the group is implemented, so that when a vehicle joins or leaves the group, the content provider only needs to send a key update message to update the session key. The proposed scheme can guarantee security requirements such as anonymity, unlinkability, forward and backward security, and resistance to conspiracy attacks, as shown by formal verification tools and further security analysis. The computational efficiency is improved by about 34.2% compared to existing schemes.
Available online , doi: 10.11999/JEIT230984
Abstract:
In order to mitigate the problem of inaccurate synchronization sensory information in Digital Twins (DTs) caused by unreliable and delayed transmission in Radio Access Networks (RAN), a sensory information synchronization strategy for DTs based on intelligent hierarchical slicing technology is proposed. The strategy aims to optimize the allocation of wireless resources for slicing and the synchronization of DTs' sensing information in dual time scales, with the goals of maximizing the satisfaction of sensing information and minimizing the costs associated with slicing reconfiguration and DTs' synchronization. Firstly, at large time scales, network slicing is employed to provide isolation for DTs with varying Quality of Service (QoS) and resolve deployment challenges; At small time scales, a more flexible wireless resource allocation is utilized to enhance the adaptability of DTs' sensory information synchronization to dynamic environments. Secondly, in order to optimize the synchronization of DTs' sensory information at different time scales, a two-layer Deep Reinforcement Learning (DRL) framework is introduced to facilitate efficient network resource interaction, and in the framework the lower-layer control algorithm incorporates the Prioritized Experience Replay (PER) mechanism to accelerate convergence speed. Finally, the effectiveness of the proposed strategy is validated through simulation results.
In order to mitigate the problem of inaccurate synchronization sensory information in Digital Twins (DTs) caused by unreliable and delayed transmission in Radio Access Networks (RAN), a sensory information synchronization strategy for DTs based on intelligent hierarchical slicing technology is proposed. The strategy aims to optimize the allocation of wireless resources for slicing and the synchronization of DTs' sensing information in dual time scales, with the goals of maximizing the satisfaction of sensing information and minimizing the costs associated with slicing reconfiguration and DTs' synchronization. Firstly, at large time scales, network slicing is employed to provide isolation for DTs with varying Quality of Service (QoS) and resolve deployment challenges; At small time scales, a more flexible wireless resource allocation is utilized to enhance the adaptability of DTs' sensory information synchronization to dynamic environments. Secondly, in order to optimize the synchronization of DTs' sensory information at different time scales, a two-layer Deep Reinforcement Learning (DRL) framework is introduced to facilitate efficient network resource interaction, and in the framework the lower-layer control algorithm incorporates the Prioritized Experience Replay (PER) mechanism to accelerate convergence speed. Finally, the effectiveness of the proposed strategy is validated through simulation results.
Available online , doi: 10.11999/JEIT230949
Abstract:
In order to solve the problems of time synchronization accuracy degradation of two-way time transfer due to relative motion between nodes in Flying Ad hoc NETwork (FANET), a full-duplex Two-Way Time Transfer (TWTT) method is proposed. Firstly, a system of equations to be solved is constructed according to the full-duplex two-way time transfer procedure, and the synchronization error expression for single full-duplex two-way time transfer is derived. Then, the convergence of iteratively performing full-duplex two-way time transfer with or without frequency offset is analyzed. Finally, the performance of full-duplex two-way time transfer method is compared with traditional two-way time transfer methods by simulation analysis and experimental validation. The simulation and experimental results show that full-duplex two-way time transfer method can achieve the same time synchronization accuracy as the physical layer timestamps under high-speed maneuvering between nodes, and the synchronization accuracy is better than the existing motion compensation methods.
In order to solve the problems of time synchronization accuracy degradation of two-way time transfer due to relative motion between nodes in Flying Ad hoc NETwork (FANET), a full-duplex Two-Way Time Transfer (TWTT) method is proposed. Firstly, a system of equations to be solved is constructed according to the full-duplex two-way time transfer procedure, and the synchronization error expression for single full-duplex two-way time transfer is derived. Then, the convergence of iteratively performing full-duplex two-way time transfer with or without frequency offset is analyzed. Finally, the performance of full-duplex two-way time transfer method is compared with traditional two-way time transfer methods by simulation analysis and experimental validation. The simulation and experimental results show that full-duplex two-way time transfer method can achieve the same time synchronization accuracy as the physical layer timestamps under high-speed maneuvering between nodes, and the synchronization accuracy is better than the existing motion compensation methods.
Available online , doi: 10.11999/JEIT230803
Abstract:
Due to the interference and background drift between new and old task parameters, semantic segmentation model based on deep neural networks promotes catastrophic forgetting of old knowledge. Furthermore, information frequently cannot be stored owing to privacy concerns, security concerns, and other issues, which leads to model failure. Therefore, a continual semantic segmentation method based on gating mechanism and replay strategy is proposed. First, without storing old data, generative adversarial network and webpage crawling are used as data sources, the label evaluation module is used to solve the unsupervised problem and the background self-drawing module is used to solve background drift problem. Then, catastrophic forgetting is mitigateed by replay strategy; Finally, gated variables are used as a regularization means to increase the sparsity of the module and study the special case of gated variables combined with continual learning replay strategy. Our evaluation results on the Pascal VOC2012 dataset show that in the settings of complex scenario 10-2, Generative Adversarial Networks (GAN) and Web, the performance of the old task after all incremental steps are improved by 3.8% and 3.7% compared with the baseline, and in scenario 10-1, they are improved by 2.7% and 1.3% compared with the baseline, respectively.
Due to the interference and background drift between new and old task parameters, semantic segmentation model based on deep neural networks promotes catastrophic forgetting of old knowledge. Furthermore, information frequently cannot be stored owing to privacy concerns, security concerns, and other issues, which leads to model failure. Therefore, a continual semantic segmentation method based on gating mechanism and replay strategy is proposed. First, without storing old data, generative adversarial network and webpage crawling are used as data sources, the label evaluation module is used to solve the unsupervised problem and the background self-drawing module is used to solve background drift problem. Then, catastrophic forgetting is mitigateed by replay strategy; Finally, gated variables are used as a regularization means to increase the sparsity of the module and study the special case of gated variables combined with continual learning replay strategy. Our evaluation results on the Pascal VOC2012 dataset show that in the settings of complex scenario 10-2, Generative Adversarial Networks (GAN) and Web, the performance of the old task after all incremental steps are improved by 3.8% and 3.7% compared with the baseline, and in scenario 10-1, they are improved by 2.7% and 1.3% compared with the baseline, respectively.
Available online , doi: 10.11999/JEIT230935
Abstract:
Considering the problem of blind identification of Unique Words (UW) for Time Division Multiple Access (TDMA) signals in non-cooperative communication, a blind identification algorithm for distributed UW is proposed in this paper. Different from the unique codes recognition algorithm at the bit layer, a unique words recognition algorithm at the waveform layer oriented to the correlation is proposed between different windows of the modulated data for centralised unique words and distributed unique words, respectively. The algorithm takes advantage of the consistency and correlation of the unique words and proceeds in two steps: firstly, the unique words of different burst signals are vertically aligned by eliminating the effects of frequency and phase bias between the different burst signals through differential accumulation, and then the positions and lengths of the unique words are identified by the multilayer differential conjugate fourth order correlation algorithm. The performance of the algorithm is simulated and analysed with different number of bursts, signal-to-noise ratios, and with or without frequency and phase biases, and the effectiveness of the waveform layer identification of unique words is verified, and the algorithm achieves more than 95% of the identification rate at a signal-to-noise ratio of 5dB for both centralized and distributed unique words, which is of certain value for engineering applications.
Considering the problem of blind identification of Unique Words (UW) for Time Division Multiple Access (TDMA) signals in non-cooperative communication, a blind identification algorithm for distributed UW is proposed in this paper. Different from the unique codes recognition algorithm at the bit layer, a unique words recognition algorithm at the waveform layer oriented to the correlation is proposed between different windows of the modulated data for centralised unique words and distributed unique words, respectively. The algorithm takes advantage of the consistency and correlation of the unique words and proceeds in two steps: firstly, the unique words of different burst signals are vertically aligned by eliminating the effects of frequency and phase bias between the different burst signals through differential accumulation, and then the positions and lengths of the unique words are identified by the multilayer differential conjugate fourth order correlation algorithm. The performance of the algorithm is simulated and analysed with different number of bursts, signal-to-noise ratios, and with or without frequency and phase biases, and the effectiveness of the waveform layer identification of unique words is verified, and the algorithm achieves more than 95% of the identification rate at a signal-to-noise ratio of 5dB for both centralized and distributed unique words, which is of certain value for engineering applications.
Available online , doi: 10.11999/JEIT230902
Abstract:
In order to solve the high-dimensional Service Function Chain (SFC) deployment problem of high reliability and low cost in the Network Function Virtualization (NFV) environment, a Improving Service and Reducing Consumption based on Proximal Policy Optimization (PPO-ISRC) is proposed. Firstly, considering the characteristics of the underlying physical server and SFC, the state transition process of the underlying server network is descried, and the deployment of SFC is taken as a Markov Decision Process. Then the reward function is set with the optimization goal of maximizing the service rate and minimizing resource consumption. Finally the PPO method is used to solve the SFC deployment strategy. The results show that compared with the heuristic algorithm First-Fit Dijkstra (FFD) and the Deep Deterministic Policy Gradient (DDPG) algorithm, the proposed algorithm has the characteristics of fast convergence speed and higher stability. Under the requirements of service quality, the deployment cost is reduced and the reliability of network service is improved.
In order to solve the high-dimensional Service Function Chain (SFC) deployment problem of high reliability and low cost in the Network Function Virtualization (NFV) environment, a Improving Service and Reducing Consumption based on Proximal Policy Optimization (PPO-ISRC) is proposed. Firstly, considering the characteristics of the underlying physical server and SFC, the state transition process of the underlying server network is descried, and the deployment of SFC is taken as a Markov Decision Process. Then the reward function is set with the optimization goal of maximizing the service rate and minimizing resource consumption. Finally the PPO method is used to solve the SFC deployment strategy. The results show that compared with the heuristic algorithm First-Fit Dijkstra (FFD) and the Deep Deterministic Policy Gradient (DDPG) algorithm, the proposed algorithm has the characteristics of fast convergence speed and higher stability. Under the requirements of service quality, the deployment cost is reduced and the reliability of network service is improved.
Available online , doi: 10.11999/JEIT231187
Abstract:
To overcome the “double path loss” in Intelligent Reflecting Surface (IRS) assisted communication system and further enhance reliability and spectral efficiency, a Reflection Modulation (RM) system scheme based on grouping of active IRS reflecting elements is proposed. This scheme utilizes the number of active reflecting element groups to transmit additional information. Then the average pairwise error probability of both the symbols transmitted by base station and the number of active reflecting element groups under the maximum likelihood detection algorithm is derived based on the moment generating function, and an upper bound on the theoretical Bit Error Probability (BEP) as well as the achievable data rate of the system are obtained. Simulation results verify the accuracy of the theoretical derivation and demonstrate the superior error performance and spectral efficiency of the proposed scheme.
To overcome the “double path loss” in Intelligent Reflecting Surface (IRS) assisted communication system and further enhance reliability and spectral efficiency, a Reflection Modulation (RM) system scheme based on grouping of active IRS reflecting elements is proposed. This scheme utilizes the number of active reflecting element groups to transmit additional information. Then the average pairwise error probability of both the symbols transmitted by base station and the number of active reflecting element groups under the maximum likelihood detection algorithm is derived based on the moment generating function, and an upper bound on the theoretical Bit Error Probability (BEP) as well as the achievable data rate of the system are obtained. Simulation results verify the accuracy of the theoretical derivation and demonstrate the superior error performance and spectral efficiency of the proposed scheme.
Available online , doi: 10.11999/JEIT230974
Abstract:
Unmanned Aerial Vehicles (UAV) loaded with various payloads can achieve multiple tasks such as sensing, communication, and computing, and are often deployed in fields such as Data Acquisition (DA) and auxiliary computing. However, so far, the vast majority of research has only focused on single function drone resource allocation and trajectory optimization, and the problem of multi task oriented drone resource allocation and trajectory optimization has not been solved yet. Therefore, an allocation strategy for optimizing drone communication network resources is proposed that comprehensively considers drone data acquisition, data broadcasting, and computing task offloading. The aim is to maximize user offloading by jointly optimizing transmission duty cycle, user transmission power, and drone trajectory, while meeting the real-time broadcast of target location data collection. In order to solve the problem of multivariable coupled optimization, an efficient iterative optimization algorithm based on Block Coordinate Descent (BCD) and Successive Convex Approximate (SCA) is proposed. The coupled optimization problem is decomposed into three sub problems for iterative optimization. Finally, a large number of simulation results show that the algorithm outperforms other testing schemes in terms of fairness and total offloading computation.
Unmanned Aerial Vehicles (UAV) loaded with various payloads can achieve multiple tasks such as sensing, communication, and computing, and are often deployed in fields such as Data Acquisition (DA) and auxiliary computing. However, so far, the vast majority of research has only focused on single function drone resource allocation and trajectory optimization, and the problem of multi task oriented drone resource allocation and trajectory optimization has not been solved yet. Therefore, an allocation strategy for optimizing drone communication network resources is proposed that comprehensively considers drone data acquisition, data broadcasting, and computing task offloading. The aim is to maximize user offloading by jointly optimizing transmission duty cycle, user transmission power, and drone trajectory, while meeting the real-time broadcast of target location data collection. In order to solve the problem of multivariable coupled optimization, an efficient iterative optimization algorithm based on Block Coordinate Descent (BCD) and Successive Convex Approximate (SCA) is proposed. The coupled optimization problem is decomposed into three sub problems for iterative optimization. Finally, a large number of simulation results show that the algorithm outperforms other testing schemes in terms of fairness and total offloading computation.
Available online , doi: 10.11999/JEIT231074
Abstract:
In order to solve the problem that the travel information of vehicle users in Intelligent Transportation System (ITS) is easy to be illegally stolen and the traffic data stored in the cloud server of transportation system is abused by malicious users, a new Attribute Based Searchable Encryption (ABSE) scheme is proposed in this paper, which has the functions of privacy protection, key aggregation and lightweight calculation. The scheme realizes full privacy protection in key generation stage, access control stage and partial decryption stage. The search keyword is embedded into the access structure, which can realize partial policy hiding and keyword security. Through key aggregation technology, all file identities that meet the search conditions and access policies are aggregated into one aggregate key, which reduces the burden of key storage for users, and further ensures the security of file keys and data. The security analysis shows that the scheme has the advantages of hidden access structure security, keyword ciphertext indistinguishable security and trapdoor indistinguishable security. The theoretical analysis and numerical simulation showed the proposed scheme was efficient and practical in terms of communication and computing overhead.
In order to solve the problem that the travel information of vehicle users in Intelligent Transportation System (ITS) is easy to be illegally stolen and the traffic data stored in the cloud server of transportation system is abused by malicious users, a new Attribute Based Searchable Encryption (ABSE) scheme is proposed in this paper, which has the functions of privacy protection, key aggregation and lightweight calculation. The scheme realizes full privacy protection in key generation stage, access control stage and partial decryption stage. The search keyword is embedded into the access structure, which can realize partial policy hiding and keyword security. Through key aggregation technology, all file identities that meet the search conditions and access policies are aggregated into one aggregate key, which reduces the burden of key storage for users, and further ensures the security of file keys and data. The security analysis shows that the scheme has the advantages of hidden access structure security, keyword ciphertext indistinguishable security and trapdoor indistinguishable security. The theoretical analysis and numerical simulation showed the proposed scheme was efficient and practical in terms of communication and computing overhead.
Available online , doi: 10.11999/JEIT231082
Abstract:
Considering the problem of the non-uniformity of the ramp signal in the large-array CMOS Image Sensors (CIS), a ramp uniformity correction method for CMOS image sensors is proposed in this paper. The correction method is based on error storage and level shift ideas. Storage capacitor that are used to store ramp non-uniformity error are introduced in column readout circuit. According to the stored ramp non-uniformity error, the ramp signal of each column is shifted. So as to ensure the uniformity of the ramp signal. Based on the 55 nm 1P4M CMOS process, this paper has completed the detailed circuit design and comprehensive simulation verification of the proposed method. Under the design conditions that the voltage range of the ramp signal is 1.4 V, the slope of the ramp signal is 71.908 V/ms, the number of pixel area arrays is 8192(H)×8192(V), and a single pixel size is 10 μm, the proposed correction method reduces the ramp non-uniformity error from 7.89mV to 36.8 μV. The Differential NonLinearity (DNL) of the ramp signal is +0.001 3/–0.004 LSB and the Integral NonLinearity (INL) is +0.045/–0.02 LSB. The Column Fixed Pattern Noise(CFPN) is reduced from 1.9% to 0.01%. The proposed ramp uniformity correction method reduces the ramp non-uniformity error by 99.54% on the basis of ensuring the high linearity of the ramp signal, without significantly increasing the chip area and without introducing additional power consumption. It provides a certain theoretical support for the design of high-precision CMOS image sensors.
Considering the problem of the non-uniformity of the ramp signal in the large-array CMOS Image Sensors (CIS), a ramp uniformity correction method for CMOS image sensors is proposed in this paper. The correction method is based on error storage and level shift ideas. Storage capacitor that are used to store ramp non-uniformity error are introduced in column readout circuit. According to the stored ramp non-uniformity error, the ramp signal of each column is shifted. So as to ensure the uniformity of the ramp signal. Based on the 55 nm 1P4M CMOS process, this paper has completed the detailed circuit design and comprehensive simulation verification of the proposed method. Under the design conditions that the voltage range of the ramp signal is 1.4 V, the slope of the ramp signal is 71.908 V/ms, the number of pixel area arrays is 8192(H)×8192(V), and a single pixel size is 10 μm, the proposed correction method reduces the ramp non-uniformity error from 7.89mV to 36.8 μV. The Differential NonLinearity (DNL) of the ramp signal is +0.001 3/–0.004 LSB and the Integral NonLinearity (INL) is +0.045/–0.02 LSB. The Column Fixed Pattern Noise(CFPN) is reduced from 1.9% to 0.01%. The proposed ramp uniformity correction method reduces the ramp non-uniformity error by 99.54% on the basis of ensuring the high linearity of the ramp signal, without significantly increasing the chip area and without introducing additional power consumption. It provides a certain theoretical support for the design of high-precision CMOS image sensors.
Available online , doi: 10.11999/JEIT231093
Abstract:
As an ideal designated verifier signature, Chameleon Signature (CS) can solve the problem of signature secondary transmission more subtly by embedding an efficient Chameleon Hash Function (CHF) into the signing algorithm. In addition to non-transferability, CS also should satisfy unforgeability, deniability, non-repudiation for the signer, and so on. To solve the problems that cryptosystems based on the traditional number theory problems, such as the large integer factorization or discrete logarithm cannot resist quantum computing attacks, and the schemes that provably secure in the random oracle model may not be secure in a practical implementation, a lattice-based CS scheme in the standard model is proposed; Furthermore, to solve the problem of requiring a significant local storage to obtain deniability for the signer, a lattice-based CS scheme without local storage in the standard model is proposed, the new scheme completely eliminates the signer's dependence on the local signature library, and enables the signer to assist an arbitrator to reject a forged signature of any adversary without storing the original message and signature. Particularly, based on the hardness of the small integer solution problem and learning with errors problem, both schemes are proved secure in the standard model.
As an ideal designated verifier signature, Chameleon Signature (CS) can solve the problem of signature secondary transmission more subtly by embedding an efficient Chameleon Hash Function (CHF) into the signing algorithm. In addition to non-transferability, CS also should satisfy unforgeability, deniability, non-repudiation for the signer, and so on. To solve the problems that cryptosystems based on the traditional number theory problems, such as the large integer factorization or discrete logarithm cannot resist quantum computing attacks, and the schemes that provably secure in the random oracle model may not be secure in a practical implementation, a lattice-based CS scheme in the standard model is proposed; Furthermore, to solve the problem of requiring a significant local storage to obtain deniability for the signer, a lattice-based CS scheme without local storage in the standard model is proposed, the new scheme completely eliminates the signer's dependence on the local signature library, and enables the signer to assist an arbitrator to reject a forged signature of any adversary without storing the original message and signature. Particularly, based on the hardness of the small integer solution problem and learning with errors problem, both schemes are proved secure in the standard model.
Available online , doi: 10.11999/JEIT231123
Abstract:
Two-Dimensional (2-D) convolution with different kernel sizes enriches the overall performance in computer vision tasks. Currently, there is a lack of an efficient design method of reconfigurable 2-D convolver, which limits the deployment of Convolution Neural Network (CNN) models at the edge. In this paper, a new approach based on multiplication management and triangular numbers decomposition is proposed. The proposed 2-D convolver includes a certain number of Processing Elements (PE) and corresponding control units, where the former is responsible for computing tasks and the latter manages the combination of multiplication operations to achieve different convolution sizes. Specifically, an odd number list is determined based on the application scenario, which represents the supported sizes of the 2-D convolutional kernel. The corresponding triangular number list is obtained using the triangular numbers decomposition method. Then, the total number of PEs is determined based on the triangular number list and computational requirements. Finally, the corresponding control units and the interconnection of PEs are determined by the addition combinations of triangular numbers. The proposed reconfigurable 2-D convolver is designed by Verilog Hardware Description Language (HDL) and implemented by Vivado 2022.2 software on the XCZU7EG board. Compared with similar methods, the proposed 2-D convolver significantly improves the efficiency of multiplication resources, increasing from 20%～50% to 89%, and achieves a throughput of 1 500 MB/s with 514 logic units, thereby demonstrating its wide applicability.
Two-Dimensional (2-D) convolution with different kernel sizes enriches the overall performance in computer vision tasks. Currently, there is a lack of an efficient design method of reconfigurable 2-D convolver, which limits the deployment of Convolution Neural Network (CNN) models at the edge. In this paper, a new approach based on multiplication management and triangular numbers decomposition is proposed. The proposed 2-D convolver includes a certain number of Processing Elements (PE) and corresponding control units, where the former is responsible for computing tasks and the latter manages the combination of multiplication operations to achieve different convolution sizes. Specifically, an odd number list is determined based on the application scenario, which represents the supported sizes of the 2-D convolutional kernel. The corresponding triangular number list is obtained using the triangular numbers decomposition method. Then, the total number of PEs is determined based on the triangular number list and computational requirements. Finally, the corresponding control units and the interconnection of PEs are determined by the addition combinations of triangular numbers. The proposed reconfigurable 2-D convolver is designed by Verilog Hardware Description Language (HDL) and implemented by Vivado 2022.2 software on the XCZU7EG board. Compared with similar methods, the proposed 2-D convolver significantly improves the efficiency of multiplication resources, increasing from 20%～50% to 89%, and achieves a throughput of 1 500 MB/s with 514 logic units, thereby demonstrating its wide applicability.
Available online , doi: 10.11999/JEIT230942
Abstract:
Considering the secure communication of the distributed Reconfigurable Intelligent Surfaces (RISs) under imperfect Channel State Information (CSI), a joint optimization problem of the secrecy rate maximization based on the active beamforming, Artificial Noise(AN), and RISs’ phase shifts is formulated. Then an efficient algorithm based on alternating optimization and 1-Dimensional linear search is proposed to solve the non-convex optimization problem. Simulation results demonstrate that, compared with the random phase optimization scheme and the secure transmission without AN scheme, the proposed scheme can achieve a higher secrecy rate. The superiority of the proposed scheme over the other transmission schemes becomes more prominent with the increase of the number of distribution units. The proposed scheme has better robustness than the other transmission schemes to the uncertainty of communication channel in our considered network.
Considering the secure communication of the distributed Reconfigurable Intelligent Surfaces (RISs) under imperfect Channel State Information (CSI), a joint optimization problem of the secrecy rate maximization based on the active beamforming, Artificial Noise(AN), and RISs’ phase shifts is formulated. Then an efficient algorithm based on alternating optimization and 1-Dimensional linear search is proposed to solve the non-convex optimization problem. Simulation results demonstrate that, compared with the random phase optimization scheme and the secure transmission without AN scheme, the proposed scheme can achieve a higher secrecy rate. The superiority of the proposed scheme over the other transmission schemes becomes more prominent with the increase of the number of distribution units. The proposed scheme has better robustness than the other transmission schemes to the uncertainty of communication channel in our considered network.
Available online , doi: 10.11999/JEIT231012
Abstract:
Most of the existing intelligent algorithms for identifying real and false targets are based on supervised learning and perform poorly under a low signal-to-noise ratio. Considering the above problems, an unsupervised clustering identification method of real and false targets based on frequency response features in multi-Coherent Processing Intervals(CPIs) is proposed by using the variability and uniqueness of the scattering characteristics of real and false targets in multi-CPIs, respectively. Firstly, the real and false targets are windowed and truncated along the fast time dimension in the fast-slow time domain, and the fast-slow time domain frequency response features are extracted to construct a preliminary sample set. Then, the real and false targets are identified by a two-step recognition algorithm composed of an Agglomerative clustering and a feature fusion network. Finally, a multi-CPI joint decision method is proposed to improve the recognition performance and reliability. It is proved by simulation and measured data that the proposed method can effectively identify real targets and multiple active false targets.
Most of the existing intelligent algorithms for identifying real and false targets are based on supervised learning and perform poorly under a low signal-to-noise ratio. Considering the above problems, an unsupervised clustering identification method of real and false targets based on frequency response features in multi-Coherent Processing Intervals(CPIs) is proposed by using the variability and uniqueness of the scattering characteristics of real and false targets in multi-CPIs, respectively. Firstly, the real and false targets are windowed and truncated along the fast time dimension in the fast-slow time domain, and the fast-slow time domain frequency response features are extracted to construct a preliminary sample set. Then, the real and false targets are identified by a two-step recognition algorithm composed of an Agglomerative clustering and a feature fusion network. Finally, a multi-CPI joint decision method is proposed to improve the recognition performance and reliability. It is proved by simulation and measured data that the proposed method can effectively identify real targets and multiple active false targets.
Available online , doi: 10.11999/JEIT230976
Abstract:
Federated Learning (FL) emerges as a distributed training method in the Internet of Vehicle (IoV), allowing Connected and Automated Vehicles (CAVs) to train a global model by exchanging models instead of raw data, protecting data privacy. Due to the limitation of model accuracy and communication overhead in FL, in this paper, a Directed Acyclic Graph (DAG) blockchain-based IoV is proposed that comprises a DAG layer and a CAV layer for model sharing and training, respectively. Furthermore, a DAG blockchain-based Asynchronous Federated Mutual distillation Learning (DAFML) algorithm is introduced to improve the model performance, which utilizes a teacher model and a student model to mutual distillation in the local training. Specifically, the teacher model with a professional network could achieve higher model accuracy, while the student model with a lightweight network could reduce the communication overhead in contrast. Moreover, to further improve the model accuracy, the personalized weight based on global epoch and model accuracy is designed to adjust the mutual distillation in the model updating. Simulation results demonstrate that the proposed DAFML algorithm outperforms other benchmarks in terms of the model accuracy and distillation ratio.
Federated Learning (FL) emerges as a distributed training method in the Internet of Vehicle (IoV), allowing Connected and Automated Vehicles (CAVs) to train a global model by exchanging models instead of raw data, protecting data privacy. Due to the limitation of model accuracy and communication overhead in FL, in this paper, a Directed Acyclic Graph (DAG) blockchain-based IoV is proposed that comprises a DAG layer and a CAV layer for model sharing and training, respectively. Furthermore, a DAG blockchain-based Asynchronous Federated Mutual distillation Learning (DAFML) algorithm is introduced to improve the model performance, which utilizes a teacher model and a student model to mutual distillation in the local training. Specifically, the teacher model with a professional network could achieve higher model accuracy, while the student model with a lightweight network could reduce the communication overhead in contrast. Moreover, to further improve the model accuracy, the personalized weight based on global epoch and model accuracy is designed to adjust the mutual distillation in the model updating. Simulation results demonstrate that the proposed DAFML algorithm outperforms other benchmarks in terms of the model accuracy and distillation ratio.
Available online , doi: 10.11999/JEIT231092
Abstract:
A decentralized, revocable, and privacy-preserving Self-Sovereign Identity (SSI) solution based on blockchain is proposed to address digital identity management challenges for users in the context of the Vehicular Ad hoc NETworks (VANETs). The Road Side Units (RSU) authorized by a Trusted Authority (TA) to form a committee are responsible for user registration, credential issuer and management.The threshold BLS signature and the Practical Byzantine Fault Tolerance (PBFT) consensus mechanism are uesd to create digital identity credentials to resolve the issues resulted in the centralized certification authorities. The combining secret sharing and zero-knowledge proof addresses privacy concerns during identity credential issuance and verification. The cryptographic accumulator is employed to tackle the revocation of user identity credentials in distributed storage scenarios. Finally, our comprehensive security analysis demonstrates the novel scheme can meet the proposed security objectives. The novel solution is implemented on an Ethereum private blockchain by using smart contracts, and experimental results show the reliability, feasibility and effectiveness of our scheme.
A decentralized, revocable, and privacy-preserving Self-Sovereign Identity (SSI) solution based on blockchain is proposed to address digital identity management challenges for users in the context of the Vehicular Ad hoc NETworks (VANETs). The Road Side Units (RSU) authorized by a Trusted Authority (TA) to form a committee are responsible for user registration, credential issuer and management.The threshold BLS signature and the Practical Byzantine Fault Tolerance (PBFT) consensus mechanism are uesd to create digital identity credentials to resolve the issues resulted in the centralized certification authorities. The combining secret sharing and zero-knowledge proof addresses privacy concerns during identity credential issuance and verification. The cryptographic accumulator is employed to tackle the revocation of user identity credentials in distributed storage scenarios. Finally, our comprehensive security analysis demonstrates the novel scheme can meet the proposed security objectives. The novel solution is implemented on an Ethereum private blockchain by using smart contracts, and experimental results show the reliability, feasibility and effectiveness of our scheme.
Available online , doi: 10.11999/JEIT231274
Abstract:
Deep learning methods have gained popularity in multimodal sentiment analysis due to their impressive representation and fusion capabilities in recent years. Existing studies often analyze the emotions of individuals using multimodal information such as text, facial expressions, and speech intonation, primarily employing complex fusion methods. However, existing models inadequately consider the dynamic changes in emotions over long time sequences, resulting in suboptimal performance in sentiment analysis. In response to this issue, a Multimodal Sentiment Analysis Model Enhanced with Non-verbal Information and Contrastive Learning is proposed in this paper. Firstly, the paper employs long-term textual information to enable the model to learn dynamic changes in audio and video across extended time sequences. Subsequently, a gating mechanism is employed to eliminate redundant information and semantic ambiguity between modalities. Finally, contrastive learning is applied to strengthen the interaction between modalities, enhancing the model’s generalization. Experimental results demonstrate that on the CMU-MOSI dataset, the model improves the Pearson Correlation coefficient (Corr) and F1 score by 3.7% and 2.1%, respectively. On the CMU-MOSEI dataset, the model increases “Corr” and “F1 score” by 1.4% and 1.1%, respectively. Therefore, the proposed model effectively utilizes intermodal interaction information while eliminating information redundancy.
Deep learning methods have gained popularity in multimodal sentiment analysis due to their impressive representation and fusion capabilities in recent years. Existing studies often analyze the emotions of individuals using multimodal information such as text, facial expressions, and speech intonation, primarily employing complex fusion methods. However, existing models inadequately consider the dynamic changes in emotions over long time sequences, resulting in suboptimal performance in sentiment analysis. In response to this issue, a Multimodal Sentiment Analysis Model Enhanced with Non-verbal Information and Contrastive Learning is proposed in this paper. Firstly, the paper employs long-term textual information to enable the model to learn dynamic changes in audio and video across extended time sequences. Subsequently, a gating mechanism is employed to eliminate redundant information and semantic ambiguity between modalities. Finally, contrastive learning is applied to strengthen the interaction between modalities, enhancing the model’s generalization. Experimental results demonstrate that on the CMU-MOSI dataset, the model improves the Pearson Correlation coefficient (Corr) and F1 score by 3.7% and 2.1%, respectively. On the CMU-MOSEI dataset, the model increases “Corr” and “F1 score” by 1.4% and 1.1%, respectively. Therefore, the proposed model effectively utilizes intermodal interaction information while eliminating information redundancy.
Available online , doi: 10.11999/JEIT231170
Abstract:
To comprehensively explore the information content of camouflaged target features, leverage the potential of target detection algorithms, and address issues such as low camouflage target detection accuracy and high false positive rates, a camouflage target detection algorithm named CAFM-YOLOv5 (Cross Attention Fusion Module Based on YOLOv5) is proposed. Firstly, a camouflaged target multispectral dataset is constructed for the performance validation of the multimodal image fusion method; secondly, a dual-stream convolution channel is constructed for visible and infrared image feature extraction; and finally, a cross-attention fusion module is proposed based on the channel-attention mechanism and spatial-attention mechanism in order to realise the effective fusion of two different features.Experimental results demonstrate that the model achieves a detection accuracy of 96.4% and a recognition probability of 88.1%, surpassing the YOLOv5 baseline network. Moreover, when compared with unimodal detection algorithms like YOLOv8 and multimodal detection algorithms such as SLBAF-Net, our algorithm exhibits superior performance in detection accuracy metrics. These findings highlight the practical value of our method for military target detection on the battlefield, enhancing situational awareness capabilities significantly.
To comprehensively explore the information content of camouflaged target features, leverage the potential of target detection algorithms, and address issues such as low camouflage target detection accuracy and high false positive rates, a camouflage target detection algorithm named CAFM-YOLOv5 (Cross Attention Fusion Module Based on YOLOv5) is proposed. Firstly, a camouflaged target multispectral dataset is constructed for the performance validation of the multimodal image fusion method; secondly, a dual-stream convolution channel is constructed for visible and infrared image feature extraction; and finally, a cross-attention fusion module is proposed based on the channel-attention mechanism and spatial-attention mechanism in order to realise the effective fusion of two different features.Experimental results demonstrate that the model achieves a detection accuracy of 96.4% and a recognition probability of 88.1%, surpassing the YOLOv5 baseline network. Moreover, when compared with unimodal detection algorithms like YOLOv8 and multimodal detection algorithms such as SLBAF-Net, our algorithm exhibits superior performance in detection accuracy metrics. These findings highlight the practical value of our method for military target detection on the battlefield, enhancing situational awareness capabilities significantly.
Available online , doi: 10.11999/JEIT231234
Abstract:
In vertical federated learning, the datasets of the clients have overlapping sample IDs and features of different dimensions, thus the data alignment is necessary for model training. As the intersection of the sample IDs is public in current data alignment technologies, how to align the data without any leakage of the intersection becomes a key issue. The proposed private-preserving data alignment framework is based on interchangeable encryption and homomorphic encryption technologies, mainly including data encryption, ciphertext blinding, private intersecting, and feature splicing. The sample IDs are encrypted twice based on an interchangeable encryption algorithm, where the same ciphertexts correspond to the same plaintexts, and the sample features are encrypted and then randomly blinded based on a homomorphic encryption algorithm. The intersection of the encrypted sample IDs is obtained, and the corresponding features are then spliced and secretly shared with the participants. Compared to the existing technologies, the privacy of the ID intersection is protected, and the samples corresponding to the IDs outside intersection can be removed safely in our framework. The security proof shows that each participant cannot obtain any knowledge of each other except for the data size, which guarantees the effectiveness of the private-preserving strategies. The simulation experiments demonstrate that the runtime is shortened about 1.3 seconds and the model accuracy keeps higher than 85% with every 10% reduction of the redundant data. The simulation experimental results show that using the ALIGN framework for vertical federated learning data alignment is beneficial for improving the efficiency and accuracy of subsequent model training.
In vertical federated learning, the datasets of the clients have overlapping sample IDs and features of different dimensions, thus the data alignment is necessary for model training. As the intersection of the sample IDs is public in current data alignment technologies, how to align the data without any leakage of the intersection becomes a key issue. The proposed private-preserving data alignment framework is based on interchangeable encryption and homomorphic encryption technologies, mainly including data encryption, ciphertext blinding, private intersecting, and feature splicing. The sample IDs are encrypted twice based on an interchangeable encryption algorithm, where the same ciphertexts correspond to the same plaintexts, and the sample features are encrypted and then randomly blinded based on a homomorphic encryption algorithm. The intersection of the encrypted sample IDs is obtained, and the corresponding features are then spliced and secretly shared with the participants. Compared to the existing technologies, the privacy of the ID intersection is protected, and the samples corresponding to the IDs outside intersection can be removed safely in our framework. The security proof shows that each participant cannot obtain any knowledge of each other except for the data size, which guarantees the effectiveness of the private-preserving strategies. The simulation experiments demonstrate that the runtime is shortened about 1.3 seconds and the model accuracy keeps higher than 85% with every 10% reduction of the redundant data. The simulation experimental results show that using the ALIGN framework for vertical federated learning data alignment is beneficial for improving the efficiency and accuracy of subsequent model training.
Available online , doi: 10.11999/JEIT240151
Abstract:
Ubiquitous coverage will become the main form of 6G networks, and complete the deployment in the mountains, hills, deserts and other blind area, to achieve full-area wireless coverage. However, the large-scale deployment of terrestrial base stations in remote areas is extremely difficult. For this reason, combining Unmanned Aerial Vehicle (UAV) communications with Non-Orthogonal Multiple Access (NOMA) technology, an energy-efficient three-dimensional coverage scheme to maximize the energy efficiency of network throughput is proposed in this paper. First, the system model is established and a user pairing algorithm is proposed based on the K-Means algorithm and the Gale-Shapley algorithm. Then, after user pairing is completed, the initial problem is split into two optimization subproblems, which are transformed to convex respectively. Finally, the block coordinate ascent method is used to alternately optimize the UAV trajectory and transmit power to maximize the energy efficiency. Simulation results show that compared with benchmarks, the proposed scheme can significantly improve the throughput energy efficiency of air-ground networks under large-scale wireless coverage.
Ubiquitous coverage will become the main form of 6G networks, and complete the deployment in the mountains, hills, deserts and other blind area, to achieve full-area wireless coverage. However, the large-scale deployment of terrestrial base stations in remote areas is extremely difficult. For this reason, combining Unmanned Aerial Vehicle (UAV) communications with Non-Orthogonal Multiple Access (NOMA) technology, an energy-efficient three-dimensional coverage scheme to maximize the energy efficiency of network throughput is proposed in this paper. First, the system model is established and a user pairing algorithm is proposed based on the K-Means algorithm and the Gale-Shapley algorithm. Then, after user pairing is completed, the initial problem is split into two optimization subproblems, which are transformed to convex respectively. Finally, the block coordinate ascent method is used to alternately optimize the UAV trajectory and transmit power to maximize the energy efficiency. Simulation results show that compared with benchmarks, the proposed scheme can significantly improve the throughput energy efficiency of air-ground networks under large-scale wireless coverage.
Available online , doi: 10.11999/JEIT240083
Abstract:
In order to solve the problems of information security, and spectrum limitation in Integrated Sensing And Communications (ISAC) systems, a secure resource allocation scheme in Intelligent Reflecting Surface (IRS)-assisted ISAC systems is investigated in this paper. To start with, in this IRS-ISAC system, where the user is being maliciously attacked by eavesdroppers, the security of the system is ensured by incorporating a jammer and deploying an IRS that utilizes its intelligent regulation of the wireless environment. Then, a secrecy rate maximization problem that subjects to the maximum transmit power constraints of the base station and the jammer, the IRS reflecting phase shift constraints, and the radar’s signal-to-noise ratio constraints is formulated by jointly designing the transmit beamforming of base station, jammer precoding vectors, and IRS phase shifts. Next, utilizing techniques such as alternating optimization and Semi-Definite Relaxation (SDR) algorithm, the original non-convex optimization problem is reformulated into a convex optimization problem, capable of determining a definitive solution. Finally, simulation results verify the security and effectiveness of the proposed algorithm and the superiority of the IRS-ISAC system.
In order to solve the problems of information security, and spectrum limitation in Integrated Sensing And Communications (ISAC) systems, a secure resource allocation scheme in Intelligent Reflecting Surface (IRS)-assisted ISAC systems is investigated in this paper. To start with, in this IRS-ISAC system, where the user is being maliciously attacked by eavesdroppers, the security of the system is ensured by incorporating a jammer and deploying an IRS that utilizes its intelligent regulation of the wireless environment. Then, a secrecy rate maximization problem that subjects to the maximum transmit power constraints of the base station and the jammer, the IRS reflecting phase shift constraints, and the radar’s signal-to-noise ratio constraints is formulated by jointly designing the transmit beamforming of base station, jammer precoding vectors, and IRS phase shifts. Next, utilizing techniques such as alternating optimization and Semi-Definite Relaxation (SDR) algorithm, the original non-convex optimization problem is reformulated into a convex optimization problem, capable of determining a definitive solution. Finally, simulation results verify the security and effectiveness of the proposed algorithm and the superiority of the IRS-ISAC system.
Available online , doi: 10.11999/JEIT240012
Abstract:
The Integrated Sensing And Communication (ISAC) requires that communication and sensing share the same radio frequency band and hardware resource. The characteristics of multi bands, large bandwidth, communication and sensing’s different requirements for hardware put forward higher requirements for ISAC hardware design. The hardware designs, verification technologies and systemic hardware verification platforms of beyond 5G, 6G and WiFi ISACs are summarized. The relevant hardware designs and verification researches at home and abroad in recent years are summarized also. The hardware design challenges such as the hardware requirement contradictions between communication and sensing systems, the In Band Full Duplex (IBFD) Self-Interference Cancellation (SIC), the Power Amplifier (PA) efficiency, and the more accurate modeling required by circuit performance are paid attention to. First of all, the design of ISAC transceiver architectures in existing researches are summarized and compared. Then, the existing ISAC IBFD self-interference suppression schemes, the low Peak to Average Power Ratio (PAPR) waveform or high-performance PA designs, the high precision device modeling methods and the systemic hardware verification platforms are introduced and analyzed. At last, the full text is summarized, the future open issues for ISAC hardware design are analyzed.
The Integrated Sensing And Communication (ISAC) requires that communication and sensing share the same radio frequency band and hardware resource. The characteristics of multi bands, large bandwidth, communication and sensing’s different requirements for hardware put forward higher requirements for ISAC hardware design. The hardware designs, verification technologies and systemic hardware verification platforms of beyond 5G, 6G and WiFi ISACs are summarized. The relevant hardware designs and verification researches at home and abroad in recent years are summarized also. The hardware design challenges such as the hardware requirement contradictions between communication and sensing systems, the In Band Full Duplex (IBFD) Self-Interference Cancellation (SIC), the Power Amplifier (PA) efficiency, and the more accurate modeling required by circuit performance are paid attention to. First of all, the design of ISAC transceiver architectures in existing researches are summarized and compared. Then, the existing ISAC IBFD self-interference suppression schemes, the low Peak to Average Power Ratio (PAPR) waveform or high-performance PA designs, the high precision device modeling methods and the systemic hardware verification platforms are introduced and analyzed. At last, the full text is summarized, the future open issues for ISAC hardware design are analyzed.
Available online , doi: 10.11999/JEIT240091
Abstract:
The anonymity network represented by The onion router(Tor) is one of the most widely used encrypted communication networks, criminals utilize encrypted networks to conceal their illegal activities, posing significant challenges to network regulation and cybersecurity. The emergence of website fingerprinting attack has made the analysis of encrypted traffic possible, enabling supervisors to identify Tor traffic and infer the web pages being visited by users by utilizing features such as packet direction and so on. In this paper, a wide survey and analysis of website fingerprinting attack and defense methods on Tor are conducted. Firstly, relevant techniques of website fingerprinting attacks on Tor are summarized and compared. The emphasis is placed on website fingerprinting attacks based on traditional machine learning and deep learning technologies. Secondly, a comprehensive survey and analysis of various existing defense methods are conducted. The limitations in the field of website fingerprinting attack methods on Tor are analyzed and summarized, and the future development directions and prospects are looked forward to.
The anonymity network represented by The onion router(Tor) is one of the most widely used encrypted communication networks, criminals utilize encrypted networks to conceal their illegal activities, posing significant challenges to network regulation and cybersecurity. The emergence of website fingerprinting attack has made the analysis of encrypted traffic possible, enabling supervisors to identify Tor traffic and infer the web pages being visited by users by utilizing features such as packet direction and so on. In this paper, a wide survey and analysis of website fingerprinting attack and defense methods on Tor are conducted. Firstly, relevant techniques of website fingerprinting attacks on Tor are summarized and compared. The emphasis is placed on website fingerprinting attacks based on traditional machine learning and deep learning technologies. Secondly, a comprehensive survey and analysis of various existing defense methods are conducted. The limitations in the field of website fingerprinting attack methods on Tor are analyzed and summarized, and the future development directions and prospects are looked forward to.
Available online , doi: 10.11999/JEIT231240
Abstract:
In the face of large-scale, diverse, and time-evolving data, as well as machine learning tasks in industrial production processes, a Federated Incremental Learning(FIL) and optimization method based on information entropy is proposed in this paper. Within the federated framework, local computing nodes utilize local data for model training, and compute the average entropy to be transmitted to the server to assist in identifying class-incremental tasks. The global server then selects local nodes for current round training based on the locally provided average entropy and makes decisions on task incrementality, followed by global model deployment and aggregation updates. The proposed method combines average entropy and thresholds for nodes selection in various situations, achieving stable model learning under low average entropy and incremental model expansion under high average entropy. Additionally, convex optimization is employed to adaptively adjust aggregation frequency and resource allocation in resource-constrained scenarios, ultimately achieving effective model convergence. Simulation results demonstrate that the proposed method accelerates model convergence and enhances training accuracy in different scenarios.
In the face of large-scale, diverse, and time-evolving data, as well as machine learning tasks in industrial production processes, a Federated Incremental Learning(FIL) and optimization method based on information entropy is proposed in this paper. Within the federated framework, local computing nodes utilize local data for model training, and compute the average entropy to be transmitted to the server to assist in identifying class-incremental tasks. The global server then selects local nodes for current round training based on the locally provided average entropy and makes decisions on task incrementality, followed by global model deployment and aggregation updates. The proposed method combines average entropy and thresholds for nodes selection in various situations, achieving stable model learning under low average entropy and incremental model expansion under high average entropy. Additionally, convex optimization is employed to adaptively adjust aggregation frequency and resource allocation in resource-constrained scenarios, ultimately achieving effective model convergence. Simulation results demonstrate that the proposed method accelerates model convergence and enhances training accuracy in different scenarios.
Available online , doi: 10.11999/JEIT231376
Abstract:
According to the problem that the maximum likelihood DOA estimation algorithm requires multi-dimensional search, is computationally intensive, and there is a problem in grid estimation, an Off-grid alternating projection maximum likelihood algorithm based on Taylor expansion is proposed. Firstly, the alternating projection method is used to transform the multi-dimensional search into multiple one-dimensional searches to obtain the rough estimation results corresponding to the preset large grid. Then, the second-order Taylor expansion of the one-dimensional cost function at the rough estimation results is carried out by using the matrix derivation theory. Finally, by calculating the partial derivative of the second-order Taylor expansion and making the derivative equal to zero, the closed-form solution of the off-grid parameters is obtained. Compared with the alternating projection maximum likelihood algorithm, this method breaks through the limitation of the search grid size. While ensuring the accuracy of the algorithm, it effectively reduces the number of points in the grid calculation of the algorithm while ensuring the accuracy of the algorithm, and improves the operation efficiency. Simulation results show the effectiveness of the algorithm.
According to the problem that the maximum likelihood DOA estimation algorithm requires multi-dimensional search, is computationally intensive, and there is a problem in grid estimation, an Off-grid alternating projection maximum likelihood algorithm based on Taylor expansion is proposed. Firstly, the alternating projection method is used to transform the multi-dimensional search into multiple one-dimensional searches to obtain the rough estimation results corresponding to the preset large grid. Then, the second-order Taylor expansion of the one-dimensional cost function at the rough estimation results is carried out by using the matrix derivation theory. Finally, by calculating the partial derivative of the second-order Taylor expansion and making the derivative equal to zero, the closed-form solution of the off-grid parameters is obtained. Compared with the alternating projection maximum likelihood algorithm, this method breaks through the limitation of the search grid size. While ensuring the accuracy of the algorithm, it effectively reduces the number of points in the grid calculation of the algorithm while ensuring the accuracy of the algorithm, and improves the operation efficiency. Simulation results show the effectiveness of the algorithm.
Available online , doi: 10.11999/JEIT231267
Abstract:
The use of semantic segmentation technology to extract high-resolution remote sensing image object segmentation has important application prospects. With the rapid development of multi-sensor technology, the good complementary advantages between multimodal remote sensing images have received widespread attention, and joint analysis of them has become a research hotspot. This article analyzes both optical remote sensing images and elevation data, and proposes a multi-task collaborative model based on multimodal remote sensing data (United Refined PSPNet, UR-PSPNet) to address the issue of insufficient fusion classification accuracy of the two types of data due to insufficient fully registered elevation data in real scenarios. This model extracts deep features of optical images, predicts semantic labels and elevation values, and embeds elevation data as supervised information, to improve the accuracy of target segmentation. This article designs a comparative experiment based on ISPRS, which proves that this algorithm can better fuse multimodal data features and improve the accuracy of object segmentation in optical remote sensing images.
The use of semantic segmentation technology to extract high-resolution remote sensing image object segmentation has important application prospects. With the rapid development of multi-sensor technology, the good complementary advantages between multimodal remote sensing images have received widespread attention, and joint analysis of them has become a research hotspot. This article analyzes both optical remote sensing images and elevation data, and proposes a multi-task collaborative model based on multimodal remote sensing data (United Refined PSPNet, UR-PSPNet) to address the issue of insufficient fusion classification accuracy of the two types of data due to insufficient fully registered elevation data in real scenarios. This model extracts deep features of optical images, predicts semantic labels and elevation values, and embeds elevation data as supervised information, to improve the accuracy of target segmentation. This article designs a comparative experiment based on ISPRS, which proves that this algorithm can better fuse multimodal data features and improve the accuracy of object segmentation in optical remote sensing images.
Available online , doi: 10.11999/JEIT231064
Abstract:
To address the catastrophic forgetting problem in Class Incremental Learning (CIL), a class incremental learning algorithm with dual separation of data flow and feature space for various classes is proposed in this paper. The Dual Separation (S2) algorithm is composed of two stages in an incremental task. In the first stage, the network training is achieved through the comprehensive constraint of classification loss, distillation loss, and contrastive loss. The data flows from different classes are separated depending on module functions, in order to enhance the network's ability to recognize new classes. By utilizing contrastive loss, the distance between different classes in the feature space is increased to prevent the feature space of old class from being eroded by the new class due to the incompleteness of the old class samples. In the second stage, the imbalanced dataset is subjected to dynamic balancing sampling to provide a balanced dataset for the new network’s dynamic fine-tuning. A high-resolution range profile incremental learning dataset of aircraft targets was created using observed and simulated data. The experimental results demonstrate that the algorithm proposed in this paper outperforms other algorithms in terms of overall performance and higher stability, while maintaining high plasticity.
To address the catastrophic forgetting problem in Class Incremental Learning (CIL), a class incremental learning algorithm with dual separation of data flow and feature space for various classes is proposed in this paper. The Dual Separation (S2) algorithm is composed of two stages in an incremental task. In the first stage, the network training is achieved through the comprehensive constraint of classification loss, distillation loss, and contrastive loss. The data flows from different classes are separated depending on module functions, in order to enhance the network's ability to recognize new classes. By utilizing contrastive loss, the distance between different classes in the feature space is increased to prevent the feature space of old class from being eroded by the new class due to the incompleteness of the old class samples. In the second stage, the imbalanced dataset is subjected to dynamic balancing sampling to provide a balanced dataset for the new network’s dynamic fine-tuning. A high-resolution range profile incremental learning dataset of aircraft targets was created using observed and simulated data. The experimental results demonstrate that the algorithm proposed in this paper outperforms other algorithms in terms of overall performance and higher stability, while maintaining high plasticity.
Available online , doi: 10.11999/JEIT240057
Abstract:
Considering the problem of 3D weak multi target detection, a 3-level Parallel-line-coordinate Transformation (PT) Track Before Detect (TBD) algorithm based on the dual accumulation self-feedback optimized is proposed in this paper. By introducing PT into TBD technology, the measurement points are transformed and accumulated sequentially on the normalized radial distance-time, azimuth angle-time and elevation angle-time planes, then the power accumulation are used to feedback the optimized binary accumulation, effectively mitigating the mutual interference between strong targets overwhelming weak targets and formation targets. Simulation results show that when the overall signal-to-clutter ratio reaches 10 dB, the overall detection probability of the proposed algorithm is close to 80%, demonstrating the effectiveness of the algorithm.
Considering the problem of 3D weak multi target detection, a 3-level Parallel-line-coordinate Transformation (PT) Track Before Detect (TBD) algorithm based on the dual accumulation self-feedback optimized is proposed in this paper. By introducing PT into TBD technology, the measurement points are transformed and accumulated sequentially on the normalized radial distance-time, azimuth angle-time and elevation angle-time planes, then the power accumulation are used to feedback the optimized binary accumulation, effectively mitigating the mutual interference between strong targets overwhelming weak targets and formation targets. Simulation results show that when the overall signal-to-clutter ratio reaches 10 dB, the overall detection probability of the proposed algorithm is close to 80%, demonstrating the effectiveness of the algorithm.
Available online , doi: 10.11999/JEIT231290
Abstract:
In the 3D maneuvering target tracking, unknown prior and coordinate coupling errors can cause model-mode mismatch and state estimation bias. In this paper, the state transition matrices are modified based on the target velocity-orthogonal condition, the spherical feasible domain is approximated by using the primal-dual regularization, and the adaptive turn rate model is combined in the frame of Unscented Kalman Filtering (UKF) to estimate the model-conditioned state, attaining the consistent output processing. 3D Variable Structure Multi-Model UKF (VSMMUKF) algorithm is derived. Simulation results show that, compared to the Multimode Importance UKF (MIUKF) algorithm, VSMMUKF can more accurately fit the maneuvering motion of 3D spatial point target with the comparable computational complexity; Compared to the Interactive Multi-model Maximum Minimum Particle Filtering (IMM-MPF) algorithm, the filtering accuracy of VSMMUKF for tracking a fixed-wing Unmanned Aerial Vehicle (UAV) has improved by 2.8%～59.9%, and the overall computation burden has reduced an order of magnitude.
In the 3D maneuvering target tracking, unknown prior and coordinate coupling errors can cause model-mode mismatch and state estimation bias. In this paper, the state transition matrices are modified based on the target velocity-orthogonal condition, the spherical feasible domain is approximated by using the primal-dual regularization, and the adaptive turn rate model is combined in the frame of Unscented Kalman Filtering (UKF) to estimate the model-conditioned state, attaining the consistent output processing. 3D Variable Structure Multi-Model UKF (VSMMUKF) algorithm is derived. Simulation results show that, compared to the Multimode Importance UKF (MIUKF) algorithm, VSMMUKF can more accurately fit the maneuvering motion of 3D spatial point target with the comparable computational complexity; Compared to the Interactive Multi-model Maximum Minimum Particle Filtering (IMM-MPF) algorithm, the filtering accuracy of VSMMUKF for tracking a fixed-wing Unmanned Aerial Vehicle (UAV) has improved by 2.8%～59.9%, and the overall computation burden has reduced an order of magnitude.
Available online , doi: 10.11999/JEIT240010
Abstract:
Memristors have a natural plasticity that enables silicon-based neurons and nano-synapses with similar or the same mechanisms as biological neurons and synapses. Using a memristor as a synapse to couple two heterogeneous memristive cellular neural networks, a memristor-coupled heterogeneous cellular neural network is constructed in this paper. The coupled network contains a space equilibrium set related to the initial value conditions of memristor synapse and subnets, which can exhibit complex dynamic evolution. The multi-stable behaviors of the coupling network, such as stable point, period, chaos, hyperchaos and unbounded oscillation, which depend on the initial value conditions, are revealed by numerical simulation method. In addition, under the control of memristor synapse, two heterogeneous subnets can achieve phase synchronization. Finally, the experimental verification of the circuit is completed based on STM32 MCU hardware platform.
Memristors have a natural plasticity that enables silicon-based neurons and nano-synapses with similar or the same mechanisms as biological neurons and synapses. Using a memristor as a synapse to couple two heterogeneous memristive cellular neural networks, a memristor-coupled heterogeneous cellular neural network is constructed in this paper. The coupled network contains a space equilibrium set related to the initial value conditions of memristor synapse and subnets, which can exhibit complex dynamic evolution. The multi-stable behaviors of the coupling network, such as stable point, period, chaos, hyperchaos and unbounded oscillation, which depend on the initial value conditions, are revealed by numerical simulation method. In addition, under the control of memristor synapse, two heterogeneous subnets can achieve phase synchronization. Finally, the experimental verification of the circuit is completed based on STM32 MCU hardware platform.
Available online , doi: 10.11999/JEIT231273
Abstract:
The Visibility Region (VR) information can be used to reduce the complexity in transmission design of Extremely Large-scale massive Multiple-Input Multiple-Output (XL-MIMO) systems. Existing theoretical analysis and transmission design are mostly based on simplified VR models. In order to evaluate and analyze the performance of XL-MIMO in realistic propagation scenarios, this paper discloses a VR spatial distribution dataset for XL-MIMO systems, which is constructed by steps including environmental parameter setting, ray tracing simulation, field strength data preprocessing and VR determination. For typical urban scenarios, the dataset establishes the connections between user locations, field strength data, and VR data, with a total number of hundreds of millions of data entries. Furthermore, the VR distribution is visualized and analyzed, and a VR-based XL-MIMO user access protocol is taken as an example usecase, with its performance evaluated with the proposed VR dataset.
The Visibility Region (VR) information can be used to reduce the complexity in transmission design of Extremely Large-scale massive Multiple-Input Multiple-Output (XL-MIMO) systems. Existing theoretical analysis and transmission design are mostly based on simplified VR models. In order to evaluate and analyze the performance of XL-MIMO in realistic propagation scenarios, this paper discloses a VR spatial distribution dataset for XL-MIMO systems, which is constructed by steps including environmental parameter setting, ray tracing simulation, field strength data preprocessing and VR determination. For typical urban scenarios, the dataset establishes the connections between user locations, field strength data, and VR data, with a total number of hundreds of millions of data entries. Furthermore, the VR distribution is visualized and analyzed, and a VR-based XL-MIMO user access protocol is taken as an example usecase, with its performance evaluated with the proposed VR dataset.
Available online , doi: 10.11999/JEIT231440
Abstract:
Considering mismatched problem between theoretical steering vector and actual steering vector causes false-alarm-rate increase in the process of maritime radar detection, the adaptive mismatched detectors are studied under Compound Gaussian Model (CGM). In order to reject mismatched signal, the fictitious signal orthogonal to theoretical steering vector is introduced in the hull hypothesis, and a target detection with mismatched signal is given. The texture component of CGM is represented by generalized inverse distribution, and the Adaptive Beamformer Orthogonal Rejection Test (ABORT) are developed based on two-step Generalized Likelihood Ratio Test (GLRT) and Maximum A Posteriori GLRT (MAP GLRT) criterions respectively. Both the proposed detectors are testified to be Constant False Alarm (CFAR) characteristics for speckle covariance matrix and target doppler steering vector. Experimental results based on simulated and real measured sea clutter data indicate that the proposed mismatched detectors show preferable target detection performance under the matched steering vector condition and anti-mismatch capability under the mismatched steering vector condition.
Considering mismatched problem between theoretical steering vector and actual steering vector causes false-alarm-rate increase in the process of maritime radar detection, the adaptive mismatched detectors are studied under Compound Gaussian Model (CGM). In order to reject mismatched signal, the fictitious signal orthogonal to theoretical steering vector is introduced in the hull hypothesis, and a target detection with mismatched signal is given. The texture component of CGM is represented by generalized inverse distribution, and the Adaptive Beamformer Orthogonal Rejection Test (ABORT) are developed based on two-step Generalized Likelihood Ratio Test (GLRT) and Maximum A Posteriori GLRT (MAP GLRT) criterions respectively. Both the proposed detectors are testified to be Constant False Alarm (CFAR) characteristics for speckle covariance matrix and target doppler steering vector. Experimental results based on simulated and real measured sea clutter data indicate that the proposed mismatched detectors show preferable target detection performance under the matched steering vector condition and anti-mismatch capability under the mismatched steering vector condition.
Available online , doi: 10.11999/JEIT231236
Abstract:
In order to achieve identification of radar emitter unaffected by signal parameters and modulation methods, a method based on Dual Radio Frequency Fingerprint Convolutional Neural Network (Dual RFF-CNN2) and feature fusion is proposed in this paper. Firstly, Raw-I/Q signals are extracted from the received radio frequency signals. Secondly, Axially Integral Bispectrum (AIB) and Square Integral Bispectrum (SIB) dimensionality reduction are performed separately on Raw-In-phase/Quadrature (Raw-I/Q) signals to construct the bispectrum integration matrix. Finally, both the Raw-I/Q signals and the bispectrum integration matrix are fed into the Dual RFF-CNN2 network for feature fusion to achieve identification of radar emitter. Experimental results demonstrate that this method achieves high identification accuracy, and the extracted "fingerprint features" exhibit stability and robustness.
In order to achieve identification of radar emitter unaffected by signal parameters and modulation methods, a method based on Dual Radio Frequency Fingerprint Convolutional Neural Network (Dual RFF-CNN2) and feature fusion is proposed in this paper. Firstly, Raw-I/Q signals are extracted from the received radio frequency signals. Secondly, Axially Integral Bispectrum (AIB) and Square Integral Bispectrum (SIB) dimensionality reduction are performed separately on Raw-In-phase/Quadrature (Raw-I/Q) signals to construct the bispectrum integration matrix. Finally, both the Raw-I/Q signals and the bispectrum integration matrix are fed into the Dual RFF-CNN2 network for feature fusion to achieve identification of radar emitter. Experimental results demonstrate that this method achieves high identification accuracy, and the extracted "fingerprint features" exhibit stability and robustness.
Available online , doi: 10.11999/JEIT231332
Abstract:
Perfect complementary sequence is a kind of signal with ideal correlation function, which is widely used in multiple access communication system, radar waveform design and so on. However, the set size of perfect complementary sequences is at most equal to the number of its subsequences. In order to expand the number of complementary sequences, the construction methods of aperiodic low correlation zone complementary sequence set are studied in this paper. First, two kinds of mapping functions on finite fields are proposed, and then two kinds of low correlation zone complementary sequence set with asymptotically optimal parameters are obtained. The number of these kinds of low correlation zone complementary sequence set are more than the perfect complementary sequence set, and which could support more users in the communication system.
Perfect complementary sequence is a kind of signal with ideal correlation function, which is widely used in multiple access communication system, radar waveform design and so on. However, the set size of perfect complementary sequences is at most equal to the number of its subsequences. In order to expand the number of complementary sequences, the construction methods of aperiodic low correlation zone complementary sequence set are studied in this paper. First, two kinds of mapping functions on finite fields are proposed, and then two kinds of low correlation zone complementary sequence set with asymptotically optimal parameters are obtained. The number of these kinds of low correlation zone complementary sequence set are more than the perfect complementary sequence set, and which could support more users in the communication system.
Available online , doi: 10.11999/JEIT231291
Abstract:
The study of information dissemination models is an important component of the Internet of Things field, which helps to improve the performance and efficiency of IoT systems, promote the further development of IoT technology. In response to the complex and unstable factors that affect information dissemination in IoT communication, a double-layer coupled information dissemination model SIVR-UAD (Susceptible, Infection, Variant, Recovered-Unknown, Aware, Disinterest) is proposed, which analyzes the impact of devices and users in different states on information dissemination in the Internet of Things, Six coupling states were established, and the Markov method was used to analyze the state change process of the coupling nodes, finding the information dissemination equilibrium point. Finally, the uniqueness and stability of the equilibrium point of the model were proved through theoretical analysis. The simulation results show that under three different initial coupling node numbers, the number of six coupling nodes in the SIVR-UAD model always tends to the same stable level, proving the equilibrium point and stability of the model.
The study of information dissemination models is an important component of the Internet of Things field, which helps to improve the performance and efficiency of IoT systems, promote the further development of IoT technology. In response to the complex and unstable factors that affect information dissemination in IoT communication, a double-layer coupled information dissemination model SIVR-UAD (Susceptible, Infection, Variant, Recovered-Unknown, Aware, Disinterest) is proposed, which analyzes the impact of devices and users in different states on information dissemination in the Internet of Things, Six coupling states were established, and the Markov method was used to analyze the state change process of the coupling nodes, finding the information dissemination equilibrium point. Finally, the uniqueness and stability of the equilibrium point of the model were proved through theoretical analysis. The simulation results show that under three different initial coupling node numbers, the number of six coupling nodes in the SIVR-UAD model always tends to the same stable level, proving the equilibrium point and stability of the model.
Available online , doi: 10.11999/JEIT231235
Abstract:
A hybrid active-passive Reconfigurable Intelligent reflecting Surface (RIS) and Artificial Noise (AN) based transmission scheme is proposed for the secret communication of the RIS assisted wireless communication system. Aiming at maximizing the secrecy rate, a joint optimization problem over the transmit beamforming and AN vector of the base station and the reflecting coefficient matrix of the RIS is formulated. Then, the Alternating Optimization (AO) method, the weighted Minimum Mean Square Error (MMSE) algorithm and the semi-definite relaxation algorithm are proposed to solve this non-convex optimization problem with highly-coupled variables. The simulation results show that the proposed hybrid RIS and AN based scheme can efficiently improve the secrecy rate of the considered system and overcome the secrecy rate decrease due to the "double fading" effect of the passive RIS. Compared with the full active RIS, the proposed scheme achieves higher energy efficiency.
A hybrid active-passive Reconfigurable Intelligent reflecting Surface (RIS) and Artificial Noise (AN) based transmission scheme is proposed for the secret communication of the RIS assisted wireless communication system. Aiming at maximizing the secrecy rate, a joint optimization problem over the transmit beamforming and AN vector of the base station and the reflecting coefficient matrix of the RIS is formulated. Then, the Alternating Optimization (AO) method, the weighted Minimum Mean Square Error (MMSE) algorithm and the semi-definite relaxation algorithm are proposed to solve this non-convex optimization problem with highly-coupled variables. The simulation results show that the proposed hybrid RIS and AN based scheme can efficiently improve the secrecy rate of the considered system and overcome the secrecy rate decrease due to the "double fading" effect of the passive RIS. Compared with the full active RIS, the proposed scheme achieves higher energy efficiency.
Available online , doi: 10.11999/JEIT230882
Abstract:
Tomographic Synthetic Aperture Radar (TomoSAR) can effectively recover the information of ground objects in steep terrain, and is one of the research hotspots in urban mapping. However, the current public data sets lack the true values of the object models, and cannot quantitatively verify the TomoSAR algorithm. To solve this problem and further promote the development of TomoSAR technology, this paper first proposes an Ray Tracing Based Space Borne Radar Advanced Simulator (RT-SBRAS), which can quickly and stably simulate the spaceborne SAR images of complex buildings compared with previous methods. Based on this, the 1.0 version of the 3D SAR Building Simulation (3DSARBuSim) data set is constructed, which contains the full-link simulation data of eight typical building scenes in dual-band and multi-pass. Finally, Orthogonal Matching Pursuit (OMP) and dual-frequency OMP algorithms are verified on the proposed data set, and the data set can provide clear and accurate quantitative comparison for the algorithms.
Tomographic Synthetic Aperture Radar (TomoSAR) can effectively recover the information of ground objects in steep terrain, and is one of the research hotspots in urban mapping. However, the current public data sets lack the true values of the object models, and cannot quantitatively verify the TomoSAR algorithm. To solve this problem and further promote the development of TomoSAR technology, this paper first proposes an Ray Tracing Based Space Borne Radar Advanced Simulator (RT-SBRAS), which can quickly and stably simulate the spaceborne SAR images of complex buildings compared with previous methods. Based on this, the 1.0 version of the 3D SAR Building Simulation (3DSARBuSim) data set is constructed, which contains the full-link simulation data of eight typical building scenes in dual-band and multi-pass. Finally, Orthogonal Matching Pursuit (OMP) and dual-frequency OMP algorithms are verified on the proposed data set, and the data set can provide clear and accurate quantitative comparison for the algorithms.
Available online , doi: 10.11999/JEIT231161
Abstract:
Currently, the Contrastive Language-Image Pre-training (CLIP) has shown great potential in zero-shot 3D shape classification. However, there is a large modality gap between 3D shapes and texts, which limits further improvement of classification accuracy. To address the problem, a zero-shot 3D shape classification method based on semantic-enhanced CLIP is proposed in this paper. Firstly, 3D shapes are represented as views. Then, in order to improve recognition ability of unknown categories in zero-shot learning, the semantic descriptive text of each view and its corresponding category are obtained through a visual language generative model, and it is used as the semantic bridge between views and category prompt texts. The semantic descriptive texts are obtained through image captioning and visual question answering. Finally, the finely-adjusted semantic encoder is used to concretize the semantic descriptive texts to the semantic descriptions of each category, which have rich semantic information and strong interpretability, and effectively reduce the semantic gap between views and category prompt texts. Experiments show that our method outperforms existing zero-shot classification methods on the ModelNet10 and ModelNet40 datasets.
Currently, the Contrastive Language-Image Pre-training (CLIP) has shown great potential in zero-shot 3D shape classification. However, there is a large modality gap between 3D shapes and texts, which limits further improvement of classification accuracy. To address the problem, a zero-shot 3D shape classification method based on semantic-enhanced CLIP is proposed in this paper. Firstly, 3D shapes are represented as views. Then, in order to improve recognition ability of unknown categories in zero-shot learning, the semantic descriptive text of each view and its corresponding category are obtained through a visual language generative model, and it is used as the semantic bridge between views and category prompt texts. The semantic descriptive texts are obtained through image captioning and visual question answering. Finally, the finely-adjusted semantic encoder is used to concretize the semantic descriptive texts to the semantic descriptions of each category, which have rich semantic information and strong interpretability, and effectively reduce the semantic gap between views and category prompt texts. Experiments show that our method outperforms existing zero-shot classification methods on the ModelNet10 and ModelNet40 datasets.
Available online , doi: 10.11999/JEIT231457
Abstract:
Logic synthesis plays a crucial role in the modern electronic design automation process. With the continuous enhancement of computational capabilities and the emergence of new computing paradigms, various efficient Boolean SATisfiability (SAT) solvers and circuit simulators have been developed and applied in the context of logic synthesis. First, the overview of the Boolean Satisfiability problem and circuit logic simulator is briefly described. Subsequently, the historical development of the matrix semi-tensor product is reviewed, and based on the fundamental principles of the semi-tensor product, its research progress in inference engines and logic synthesis is expounded. Finally, a prospective analysis is conducted on emerging technologies that may significantly impact logic synthesis in the future.
Logic synthesis plays a crucial role in the modern electronic design automation process. With the continuous enhancement of computational capabilities and the emergence of new computing paradigms, various efficient Boolean SATisfiability (SAT) solvers and circuit simulators have been developed and applied in the context of logic synthesis. First, the overview of the Boolean Satisfiability problem and circuit logic simulator is briefly described. Subsequently, the historical development of the matrix semi-tensor product is reviewed, and based on the fundamental principles of the semi-tensor product, its research progress in inference engines and logic synthesis is expounded. Finally, a prospective analysis is conducted on emerging technologies that may significantly impact logic synthesis in the future.
Available online , doi: 10.11999/JEIT231367
Abstract:
A sparse Bayesian estimation for spatial Radio Frequency Interference (RFI) of synthetic aperture microwave radiometers is proposed in this paper. Firstly, an interferometry measurement model of the visibility function for synthetic aperture microwave radiometers is established. The observed data are expressed as the product of the observation matrix of the aperture synthesis antenna baseline correlation steering vector and the brightness temperature of the field of view. Due to the orthogonality of the observation matrix and the sparsity of the RFI spatial angle distribution, the transformation coefficients of brightness temperature in the support domain are sparse. Under the Sparse Bayesian Learning (SBL) framework, brightness temperature is sparsely reconstructed. Notably, this method can obtain high reconstruction performance without the prior information of sparsity and regularization parameters. The effectiveness of this method is verified through computer simulations.
A sparse Bayesian estimation for spatial Radio Frequency Interference (RFI) of synthetic aperture microwave radiometers is proposed in this paper. Firstly, an interferometry measurement model of the visibility function for synthetic aperture microwave radiometers is established. The observed data are expressed as the product of the observation matrix of the aperture synthesis antenna baseline correlation steering vector and the brightness temperature of the field of view. Due to the orthogonality of the observation matrix and the sparsity of the RFI spatial angle distribution, the transformation coefficients of brightness temperature in the support domain are sparse. Under the Sparse Bayesian Learning (SBL) framework, brightness temperature is sparsely reconstructed. Notably, this method can obtain high reconstruction performance without the prior information of sparsity and regularization parameters. The effectiveness of this method is verified through computer simulations.
Available online , doi: 10.11999/JEIT231423
Abstract:
In response to the challenge of ensuring positioning accuracy in environments where the Global Navigation Satellite System (GNSS) is denied, a positioning scheme based on opportunistic New Radio (NR) signals is devised and an Interference Cancellation Subspace Pursuit (ICSP) algorithm is proposed in this paper. This algorithm aims to resolve the issue of inadequate precision in the extraction of positioning observations due to co-channel interference within Ultra-Dense Networks (UDNs) and Heterogeneous Networks (HetNets). The effectiveness of the ICSP algorithm in optimizing the performance of 5G opportunistic signal receivers and enhancing positioning accuracy in complex network environments has been validated through simulation experiments and semi-physical simulations utilizing the Universal Software Radio Peripheral (USRP).
In response to the challenge of ensuring positioning accuracy in environments where the Global Navigation Satellite System (GNSS) is denied, a positioning scheme based on opportunistic New Radio (NR) signals is devised and an Interference Cancellation Subspace Pursuit (ICSP) algorithm is proposed in this paper. This algorithm aims to resolve the issue of inadequate precision in the extraction of positioning observations due to co-channel interference within Ultra-Dense Networks (UDNs) and Heterogeneous Networks (HetNets). The effectiveness of the ICSP algorithm in optimizing the performance of 5G opportunistic signal receivers and enhancing positioning accuracy in complex network environments has been validated through simulation experiments and semi-physical simulations utilizing the Universal Software Radio Peripheral (USRP).
Available online , doi: 10.11999/JEIT231374
Abstract:
Addressing the issues of inadequate performance in constructing Radio Environment Maps (REMs) in complex scenarios due to non-penetrable obstacles for electromagnetic waves, and the arbitrary selection of interpolation neighborhoods imposed by Inverse Distance Weighted (IDW), a Voronoi-based Inverse Obstacle Distance Weighted algorithm (VIODW) is proposed in this paper. This algorithm adaptively defines interpolation neighborhoods for each interpolation point by creating Voronoi diagrams incorporating obstacles for numerical computation. Then, using the ANY-Angle (ANYA) Algorithm to calculate the obstacle distance between the interpolation point and each monitoring station within the interpolation neighborhood. Finally, by calculating the weighted mean with the inverse power of the obstacle distance as the weight, the value at the point is obtained, achieving high-precision construction of REMs in complex scenarios. Both theoretical analysis and simulation results demonstrate that this method offers excellent construction accuracy and can accurately model the power distribution of electromagnetic waves in complex scenarios. Hence, it provides an effective approach for high-precision REM construction in complex scenarios.
Addressing the issues of inadequate performance in constructing Radio Environment Maps (REMs) in complex scenarios due to non-penetrable obstacles for electromagnetic waves, and the arbitrary selection of interpolation neighborhoods imposed by Inverse Distance Weighted (IDW), a Voronoi-based Inverse Obstacle Distance Weighted algorithm (VIODW) is proposed in this paper. This algorithm adaptively defines interpolation neighborhoods for each interpolation point by creating Voronoi diagrams incorporating obstacles for numerical computation. Then, using the ANY-Angle (ANYA) Algorithm to calculate the obstacle distance between the interpolation point and each monitoring station within the interpolation neighborhood. Finally, by calculating the weighted mean with the inverse power of the obstacle distance as the weight, the value at the point is obtained, achieving high-precision construction of REMs in complex scenarios. Both theoretical analysis and simulation results demonstrate that this method offers excellent construction accuracy and can accurately model the power distribution of electromagnetic waves in complex scenarios. Hence, it provides an effective approach for high-precision REM construction in complex scenarios.
Available online , doi: 10.11999/JEIT231283
Abstract:
To better leverage complementary image information from infrared and visible light images and generate fused images that align with human perception characteristics, a two-stage training strategy is proposed to obtain a novel infrared-visible image fusion Network based on pre-trained fixed Parameters and Deep feature modulation (PDNet). Specifically, in the self-supervised pre-training stage, a substantial dataset of clear natural images is employed as both inputs and outputs for the UNet backbone network, and pre-training is accomplished with autoencoder technology. As such, the resulting encoder module can proficiently extract multi-scale depth features from the input image, while the decoder module can faithfully reconstruct it into an output image with minimal deviation from the input. In the unsupervised fusion training stage, the pre-trained encoder and decoder module parameters remain fixed, and a fusion module featuring a Transformer structure is introduced between them. Within the Transformer structure, the multi-head self-attention mechanism allocates deep feature weights, extracted by the encoder from both infrared and visible light images, in a rational manner. This process fuses and modulates the deep image features at various scales into the manifold space of deep features of clear natural image, thereby ensuring the visual perception quality of the fused image after reconstruction by the decoder. Extensive experimental results demonstrate that, in comparison to current mainstream fusion models (algorithms), the proposed PDNet model exhibits substantial advantages across various objective evaluation metrics. Furthermore, in subjective visual evaluations, it aligns more closely with human visual perception characteristics.
To better leverage complementary image information from infrared and visible light images and generate fused images that align with human perception characteristics, a two-stage training strategy is proposed to obtain a novel infrared-visible image fusion Network based on pre-trained fixed Parameters and Deep feature modulation (PDNet). Specifically, in the self-supervised pre-training stage, a substantial dataset of clear natural images is employed as both inputs and outputs for the UNet backbone network, and pre-training is accomplished with autoencoder technology. As such, the resulting encoder module can proficiently extract multi-scale depth features from the input image, while the decoder module can faithfully reconstruct it into an output image with minimal deviation from the input. In the unsupervised fusion training stage, the pre-trained encoder and decoder module parameters remain fixed, and a fusion module featuring a Transformer structure is introduced between them. Within the Transformer structure, the multi-head self-attention mechanism allocates deep feature weights, extracted by the encoder from both infrared and visible light images, in a rational manner. This process fuses and modulates the deep image features at various scales into the manifold space of deep features of clear natural image, thereby ensuring the visual perception quality of the fused image after reconstruction by the decoder. Extensive experimental results demonstrate that, in comparison to current mainstream fusion models (algorithms), the proposed PDNet model exhibits substantial advantages across various objective evaluation metrics. Furthermore, in subjective visual evaluations, it aligns more closely with human visual perception characteristics.
Available online , doi: 10.11999/JEIT240051
Abstract:
General low-complexity joint beamforming designs are proposed for Reconfigurable Intelligent Surface (RIS) assisted multi-user systems. First, the non-convex optimization problem of joint beamforming design is analyzed to maximize sum data rate for RIS-aided multi-user systems. Second, the RIS reflection matrix is designed by using the approximation orthogonality of the beam steering vectors, and the transmit beamforming at the base station is derived from the zero forcing method, and the power allocation is optimized for multiple users. Finally, it is found that the proposed scheme has wide applicability and an order of magnitude reduction on computational complexity than that of existing work. Numerical results show that the proposed beamforming design can achieve high sum data rate, which can be further improved by employing the optimal power allocation. Besides, both the simulation results and theoretical analysis indicate that the sum data rate changes with the RIS location, which provides reference standards for the selection of RIS location.
General low-complexity joint beamforming designs are proposed for Reconfigurable Intelligent Surface (RIS) assisted multi-user systems. First, the non-convex optimization problem of joint beamforming design is analyzed to maximize sum data rate for RIS-aided multi-user systems. Second, the RIS reflection matrix is designed by using the approximation orthogonality of the beam steering vectors, and the transmit beamforming at the base station is derived from the zero forcing method, and the power allocation is optimized for multiple users. Finally, it is found that the proposed scheme has wide applicability and an order of magnitude reduction on computational complexity than that of existing work. Numerical results show that the proposed beamforming design can achieve high sum data rate, which can be further improved by employing the optimal power allocation. Besides, both the simulation results and theoretical analysis indicate that the sum data rate changes with the RIS location, which provides reference standards for the selection of RIS location.
Available online , doi: 10.11999/JEIT231460
Abstract:
The high-quality generation and precise phase synchronization of high-speed modulated baseband signals are key technologies of space optical communication ranging system. Traditional approaches relying on FPGA or Digital Signal Processor (DSP) and high-speed Digital to Analog Convertor (DAC) technology often suffer from limited phase synchronization accuracy and high hardware complexity. A method for high-speed optical communication baseband signal generation and phase synchronization is proposed and a phase-locked dynamic control loop is designed in this paper. By dynamically adjusting the phase of the high-speed signal transmission clock in real time, the deterministic relationship between the I/Q high-speed baseband signal phase and the external reference clock phase can be achieved. The experimental results demonstrate impressive performance metrics: When the code rate is of the Quadrature Phase Shift Keying (QPSK) optical modulated signal is 2.5 Gbit/s, the phase synchronization accuracy is less than 2 ps and the Error Vector Magnitude (EVM) is less than 8%; the bit error rate is 10–7 at a 5 Gbit/s optical communication rate, the receiver sensitivity is better than –47 dBm, and the ranging accuracy is better than 2 mm. Compared with traditional methods, both sensitivity and ranging accuracy are significantly improved.
The high-quality generation and precise phase synchronization of high-speed modulated baseband signals are key technologies of space optical communication ranging system. Traditional approaches relying on FPGA or Digital Signal Processor (DSP) and high-speed Digital to Analog Convertor (DAC) technology often suffer from limited phase synchronization accuracy and high hardware complexity. A method for high-speed optical communication baseband signal generation and phase synchronization is proposed and a phase-locked dynamic control loop is designed in this paper. By dynamically adjusting the phase of the high-speed signal transmission clock in real time, the deterministic relationship between the I/Q high-speed baseband signal phase and the external reference clock phase can be achieved. The experimental results demonstrate impressive performance metrics: When the code rate is of the Quadrature Phase Shift Keying (QPSK) optical modulated signal is 2.5 Gbit/s, the phase synchronization accuracy is less than 2 ps and the Error Vector Magnitude (EVM) is less than 8%; the bit error rate is 10–7 at a 5 Gbit/s optical communication rate, the receiver sensitivity is better than –47 dBm, and the ranging accuracy is better than 2 mm. Compared with traditional methods, both sensitivity and ranging accuracy are significantly improved.
Available online , doi: 10.11999/JEIT231232
Abstract:
Narrowband radar is widely used in the field of air defense guidance due to its advantages of low cost and long operating range. With the development of high-speed mobile platforms, traditional target recognition methods based on feature modeling of long-term observation echo sequences are no longer applicable. In response to the problem of poor feature recognition ability of narrowband radar for Observe Echoes for a Short period of Time (OEST) sequences and susceptibility to bait target interference, resulting in low reliability of recognition results, a narrowband radar OEST sequence air target recognition method using multi feature adaptive fusion is proposed in this paper. Firstly, the encoder and classification layers are constructed with channel-spatial attention modules and trained to adaptively enhance features with high separability. Then, the maximum edge orthogonal loss function is proposed to increase the feature spacing between different classes, reduce the feature spacing between the same classes, and make the feature vectors orthogonal between different classes; Finally, the parameters of the encoder layer and classification layer are fixed, and the decoder layer is trained using reconstruction loss value to ensure that the model has accurate identification ability for decoy targets. Under the condition of an observation sequence length of 100, the classification accuracy and discrimination rate of the experimental part reached 94.37% and 96.78%, respectively. It can be concluded that the proposed method can effectively improve the classification performance of narrowband radar and the discrimination ability against bait targets, thereby improving the reliability of recognition results.
Narrowband radar is widely used in the field of air defense guidance due to its advantages of low cost and long operating range. With the development of high-speed mobile platforms, traditional target recognition methods based on feature modeling of long-term observation echo sequences are no longer applicable. In response to the problem of poor feature recognition ability of narrowband radar for Observe Echoes for a Short period of Time (OEST) sequences and susceptibility to bait target interference, resulting in low reliability of recognition results, a narrowband radar OEST sequence air target recognition method using multi feature adaptive fusion is proposed in this paper. Firstly, the encoder and classification layers are constructed with channel-spatial attention modules and trained to adaptively enhance features with high separability. Then, the maximum edge orthogonal loss function is proposed to increase the feature spacing between different classes, reduce the feature spacing between the same classes, and make the feature vectors orthogonal between different classes; Finally, the parameters of the encoder layer and classification layer are fixed, and the decoder layer is trained using reconstruction loss value to ensure that the model has accurate identification ability for decoy targets. Under the condition of an observation sequence length of 100, the classification accuracy and discrimination rate of the experimental part reached 94.37% and 96.78%, respectively. It can be concluded that the proposed method can effectively improve the classification performance of narrowband radar and the discrimination ability against bait targets, thereby improving the reliability of recognition results.
Available online , doi: 10.11999/JEIT231338
Abstract:
Realizing high accuracy and low computational burden is a serious challenge faced by Convolutional Neural Network (CNN) for real-time semantic segmentation. In this paper, an efficient real-time semantic segmentation Adaptive Attention mechanism Fusion Network(AAFNet) is designed for complex urban street scenes with numerous types of targets and large changes in lighting. Image spatial details and semantic information are respectively extracted by the network, and then, through Feature Fusion Network(FFN), accurate semantic images are obtained. Dilated Deep-Wise separable convolution (DDW) is adopted by AAFNet to increase the receptive field of semantic feature extraction, an Adaptive Attention mechanism Fusion Module (AAFM) is proposed, which combines Adaptive average pooling(Avp) and Adaptive max pooling(Amp) to refine the edge segmentation effect of the target and reduce the leakage rate of small targets. Finally, semantic segmentation experiments are performed on the Cityscapes and CamVid datasets for complex urban street scenes. The designed AAFNet achieves 73.0% and 69.8% mean Intersection over Union (mIoU) at inference speeds of 32 fps (Cityscapes) and 52 fps (CamVid). Compared with Dilated Spatial Attention Network (DSANet), Multi-Scale Context Fusion Network (MSCFNet), and Lightweight Bilateral Asymmetric Residual Network (LBARNet), AAFNet has the highest segmentation accuracy.
Realizing high accuracy and low computational burden is a serious challenge faced by Convolutional Neural Network (CNN) for real-time semantic segmentation. In this paper, an efficient real-time semantic segmentation Adaptive Attention mechanism Fusion Network(AAFNet) is designed for complex urban street scenes with numerous types of targets and large changes in lighting. Image spatial details and semantic information are respectively extracted by the network, and then, through Feature Fusion Network(FFN), accurate semantic images are obtained. Dilated Deep-Wise separable convolution (DDW) is adopted by AAFNet to increase the receptive field of semantic feature extraction, an Adaptive Attention mechanism Fusion Module (AAFM) is proposed, which combines Adaptive average pooling(Avp) and Adaptive max pooling(Amp) to refine the edge segmentation effect of the target and reduce the leakage rate of small targets. Finally, semantic segmentation experiments are performed on the Cityscapes and CamVid datasets for complex urban street scenes. The designed AAFNet achieves 73.0% and 69.8% mean Intersection over Union (mIoU) at inference speeds of 32 fps (Cityscapes) and 52 fps (CamVid). Compared with Dilated Spatial Attention Network (DSANet), Multi-Scale Context Fusion Network (MSCFNet), and Lightweight Bilateral Asymmetric Residual Network (LBARNet), AAFNet has the highest segmentation accuracy.
Available online , doi: 10.11999/JEIT231165
Abstract:
To relieve the impact of data heterogeneity problems caused by full overlapping attribute skew between clients in Federated Learning (FL), a local adaptive FL algorithm that incorporates channel personalized normalization is proposed in this paper. Specifically, an FL model oriented to data attribute skew is constructed, and a series of random enhancement operations are performed on the images data set in the client before training begins. Next, the client calculates the mean and standard deviation of the data set separately by color channel to achieve channel personalized normalization. Furthermore, a local adaptive update FL algorithm is designed, that is, the global model and the local model are adaptively aggregated for local initialization. The uniqueness of this aggregation method is that it not only retains the personalized characteristics of the client model, but also can capture necessary information in the global model to improve the generalization performance of the model. Finally, the experimental results demonstrate that the proposed algorithm obtains competitive convergence speed compared with existing representative works and the accuracy is 3%～19% higher.
To relieve the impact of data heterogeneity problems caused by full overlapping attribute skew between clients in Federated Learning (FL), a local adaptive FL algorithm that incorporates channel personalized normalization is proposed in this paper. Specifically, an FL model oriented to data attribute skew is constructed, and a series of random enhancement operations are performed on the images data set in the client before training begins. Next, the client calculates the mean and standard deviation of the data set separately by color channel to achieve channel personalized normalization. Furthermore, a local adaptive update FL algorithm is designed, that is, the global model and the local model are adaptively aggregated for local initialization. The uniqueness of this aggregation method is that it not only retains the personalized characteristics of the client model, but also can capture necessary information in the global model to improve the generalization performance of the model. Finally, the experimental results demonstrate that the proposed algorithm obtains competitive convergence speed compared with existing representative works and the accuracy is 3%～19% higher.
Available online , doi: 10.11999/JEIT231262
Abstract:
An end-to-end quadruple Super-Resolution Inpainting Generative Adversarial Network (SRIGAN) is proposed in this paper, for low-resolution random occlusion face images. The generative network consists of an encoder, a feature compensation subnetwork, and a decoder constructed with a pyramid attention module. The discriminant network is an improved Patch discriminant network. The network can effectively learn the absent features of the occluded region through a feature compensation subnetwork and a two-stage training strategy. Then, the information is constructed with the decoder with a pyramid attention module and multi-scale reconstruction loss. Hence, the generative network can transform a low-resolution occlusion image into a quadruple high-resolution complete image. Furthermore, the improvements of the loss function and Patch discriminant network are employed to ensure the stability of network training and enhance the performance of the generated network. The effectiveness of the proposed algorithm is verified by comparison and module verification experiments.
An end-to-end quadruple Super-Resolution Inpainting Generative Adversarial Network (SRIGAN) is proposed in this paper, for low-resolution random occlusion face images. The generative network consists of an encoder, a feature compensation subnetwork, and a decoder constructed with a pyramid attention module. The discriminant network is an improved Patch discriminant network. The network can effectively learn the absent features of the occluded region through a feature compensation subnetwork and a two-stage training strategy. Then, the information is constructed with the decoder with a pyramid attention module and multi-scale reconstruction loss. Hence, the generative network can transform a low-resolution occlusion image into a quadruple high-resolution complete image. Furthermore, the improvements of the loss function and Patch discriminant network are employed to ensure the stability of network training and enhance the performance of the generated network. The effectiveness of the proposed algorithm is verified by comparison and module verification experiments.
Available online , doi: 10.11999/JEIT231145
Abstract:
Elements such as pulse interference and outlier measurement information usually lead to abnormal heavy-tailed noise, which sharply reduces the performance of the Extended Target Tracking (ETT) estimator based on the Gaussian hypothesis. To address this problem, a Student’s t Inverse Wishart Smoothing (StIWS) algorithm based on the Random Matrix Model (RMM) is proposed. Firstly, the kinematic state of the target, process noise and measurement noise are modeled as a Student’s t distribution to characterize the effect of anomalous noise on the probability distribution of extended target, and the extended state of target is modeled as a random matrix which obeys inverse Wishart distribution. Then, in a Student’s t bayesian smoothing frame, the StIWS algorithm is derived in detail, which can effectively estimate target state in the process of the dynamic evolution of multiple characteristics of extended target. Finally, the effectiveness of the proposed algorithm is verified by the simulation experiment and the engineering experiment of extended target tracking.
Elements such as pulse interference and outlier measurement information usually lead to abnormal heavy-tailed noise, which sharply reduces the performance of the Extended Target Tracking (ETT) estimator based on the Gaussian hypothesis. To address this problem, a Student’s t Inverse Wishart Smoothing (StIWS) algorithm based on the Random Matrix Model (RMM) is proposed. Firstly, the kinematic state of the target, process noise and measurement noise are modeled as a Student’s t distribution to characterize the effect of anomalous noise on the probability distribution of extended target, and the extended state of target is modeled as a random matrix which obeys inverse Wishart distribution. Then, in a Student’s t bayesian smoothing frame, the StIWS algorithm is derived in detail, which can effectively estimate target state in the process of the dynamic evolution of multiple characteristics of extended target. Finally, the effectiveness of the proposed algorithm is verified by the simulation experiment and the engineering experiment of extended target tracking.
Available online , doi: 10.11999/JEIT231061
Abstract:
To address the high computational performance demands of delay-sensitive tasks in complex terrains, the collaborative computation offloading scheme for reusable tasks in mobile edge computing with the assistance of Unmanned Aerial Vehicle (UAV) is proposed. Firstly, the minimization of the average total energy consumption is formulated by jointly optimizing user offloading, user transmission power, server assignment on UAV, computation frequencies of users and UAV servers, as well as UAV flight trajectory, while meeting the latency constraints. Secondly, a deep reinforcement learning approach is employed to solve the optimization problem, and a Soft Actor-Critic (SAC) based optimization algorithm is introduced. The SAC algorithm utilizes a maximum entropy policy to encourage exploration that enhances the algorithm’s exploration capabilities and accelerates the training convergence speed. Simulation results demonstrate that the proposed SAC algorithm effectively reduces the average total energy consumption of the system while exhibiting good convergence.
To address the high computational performance demands of delay-sensitive tasks in complex terrains, the collaborative computation offloading scheme for reusable tasks in mobile edge computing with the assistance of Unmanned Aerial Vehicle (UAV) is proposed. Firstly, the minimization of the average total energy consumption is formulated by jointly optimizing user offloading, user transmission power, server assignment on UAV, computation frequencies of users and UAV servers, as well as UAV flight trajectory, while meeting the latency constraints. Secondly, a deep reinforcement learning approach is employed to solve the optimization problem, and a Soft Actor-Critic (SAC) based optimization algorithm is introduced. The SAC algorithm utilizes a maximum entropy policy to encourage exploration that enhances the algorithm’s exploration capabilities and accelerates the training convergence speed. Simulation results demonstrate that the proposed SAC algorithm effectively reduces the average total energy consumption of the system while exhibiting good convergence.
Available online , doi: 10.11999/JEIT230830
Abstract:
Unmanned Aerial Vehicle (UAV) equipped with Intelligent Reflecting Surface (IRS) can effectively solve the problem of inefficient information and energy transmission between the hybrid access point and nodes in complex wireless scenarios due to obstacle occlusion. A novel framework for aerial IRS-assisted wireless powered communication networks is proposed that exploits the flexibility of aerial IRS to improve the performance of the network. The architecture achieves the transmission of energy and data for each time slot employing the harvest-then-transmit scheme. A multi-variable coupled optimization problem that combines the flight trajectory, node selection association variable, time slot allocation ratio, and the phase is established while satisfying the node energy harvesting threshold. Thus, the block coordinate descent algorithm is utilized to decompose the optimization problem into four sub-problems to be solved separately. Firstly, the closed-form solution for the optimal phase of the intelligent reflecting surface is derived based on the beam alignment theory. Secondly, the non-convex problem is transformed into a convex problem by introducing auxiliary variables and employing a successive convex approximation algorithm. Finally, the solution is iteratively solved utilizing the block coordinate descent algorithm. Simulation results show that the proposed scheme has excellent convergence performance and significantly improve the average throughput.
Unmanned Aerial Vehicle (UAV) equipped with Intelligent Reflecting Surface (IRS) can effectively solve the problem of inefficient information and energy transmission between the hybrid access point and nodes in complex wireless scenarios due to obstacle occlusion. A novel framework for aerial IRS-assisted wireless powered communication networks is proposed that exploits the flexibility of aerial IRS to improve the performance of the network. The architecture achieves the transmission of energy and data for each time slot employing the harvest-then-transmit scheme. A multi-variable coupled optimization problem that combines the flight trajectory, node selection association variable, time slot allocation ratio, and the phase is established while satisfying the node energy harvesting threshold. Thus, the block coordinate descent algorithm is utilized to decompose the optimization problem into four sub-problems to be solved separately. Firstly, the closed-form solution for the optimal phase of the intelligent reflecting surface is derived based on the beam alignment theory. Secondly, the non-convex problem is transformed into a convex problem by introducing auxiliary variables and employing a successive convex approximation algorithm. Finally, the solution is iteratively solved utilizing the block coordinate descent algorithm. Simulation results show that the proposed scheme has excellent convergence performance and significantly improve the average throughput.
Available online , doi: 10.11999/JEIT231139
Abstract:
For Semi-Coprime Arrays (SCA), the performance of classical Direction of Arrival (DoA) estimation algorithm degrades under the presence of coherent adjacent sources. To address this problem, a high-precison DoA estimation method for SCA is proposed. Firstly, the array is divided into three subarrays (Subarray 1 to 3 respectively). And conventional beamforming algorithm is applied to obtain the signals of the three subarrays, respectively. Then, the output signal of subarray 3 is weighted and added to the output signals of subarray 1 and 2 to construct one sum beam. Meanwhile, the difference between output signals of subarray 1 and subarray 2 is used to construct one difference beam. Finally, the final output signal is obtained by the sum beam signal and the difference beam signal. The azimuth spectrum is the power of the final output signal. This method is based on the characteristic of SCA arrays to construct sum beam and difference beam, fully utilizing the overlapping sensors of the three subarrays to improve estimation accuracy. Simulations and lake experiments are implemented to validate the effectiveness for the proposed method used for DoA estimation in SCA. The proposed method performs better than the existing approaches, such as Minimum Variance Distortionless Response (MVDR) and Min Processing (MP) when facing adjacent coherent sources.
For Semi-Coprime Arrays (SCA), the performance of classical Direction of Arrival (DoA) estimation algorithm degrades under the presence of coherent adjacent sources. To address this problem, a high-precison DoA estimation method for SCA is proposed. Firstly, the array is divided into three subarrays (Subarray 1 to 3 respectively). And conventional beamforming algorithm is applied to obtain the signals of the three subarrays, respectively. Then, the output signal of subarray 3 is weighted and added to the output signals of subarray 1 and 2 to construct one sum beam. Meanwhile, the difference between output signals of subarray 1 and subarray 2 is used to construct one difference beam. Finally, the final output signal is obtained by the sum beam signal and the difference beam signal. The azimuth spectrum is the power of the final output signal. This method is based on the characteristic of SCA arrays to construct sum beam and difference beam, fully utilizing the overlapping sensors of the three subarrays to improve estimation accuracy. Simulations and lake experiments are implemented to validate the effectiveness for the proposed method used for DoA estimation in SCA. The proposed method performs better than the existing approaches, such as Minimum Variance Distortionless Response (MVDR) and Min Processing (MP) when facing adjacent coherent sources.
Available online , doi: 10.11999/JEIT230918
Abstract:
In response to the characteristics of heterogeneous node resources and dynamic changes in the network topology in the Internet of Vehicles (IoV), a Two-layer Asynchronous Federated Learning with Two-factor updating (TTAFL) framework is established in this paper. Considering the impact of model version differences and the number of times that vehicles participate in Federated Learning (FL) on server model updates, a model update scheme based on staleness factor and contribution factor is proposed. Furthermore, to avoid the problem of roadside unit switching caused by vehicle mobility during the training process, a node selection scheme considering the residence time is given. Finally, in order to reduce the accuracy loss and system energy consumption, a reinforcement learning method is used to optimize the number of local iterations of FL and the number of local model updates of roadside units. Simulation results show that the proposed algorithm effectively improves the training efficiency and training accuracy of federated learning and reduces the system energy consumption.
In response to the characteristics of heterogeneous node resources and dynamic changes in the network topology in the Internet of Vehicles (IoV), a Two-layer Asynchronous Federated Learning with Two-factor updating (TTAFL) framework is established in this paper. Considering the impact of model version differences and the number of times that vehicles participate in Federated Learning (FL) on server model updates, a model update scheme based on staleness factor and contribution factor is proposed. Furthermore, to avoid the problem of roadside unit switching caused by vehicle mobility during the training process, a node selection scheme considering the residence time is given. Finally, in order to reduce the accuracy loss and system energy consumption, a reinforcement learning method is used to optimize the number of local iterations of FL and the number of local model updates of roadside units. Simulation results show that the proposed algorithm effectively improves the training efficiency and training accuracy of federated learning and reduces the system energy consumption.
Available online , doi: 10.11999/JEIT231067
Abstract:
The integrated scheduling algorithm of flexible equipment network is difficult to reasonably select the relevant processes of processing equipment, which affects the completion time of products. An Integrated Scheduling algorithm for Flexible Equipment Network considering the Same layer after Process (SP-FENIS) is proposed. Firstly, the priority strategy of the reverse order layer is adopted, which assigns each process to the set of processes to be scheduled in the reverse layer. Then, the average reverse-order compact path strategy is proposed to determine the scheduling sequence of the processes to be scheduled in each reverse order layer. Finally, the earliest completion time strategy and equipment idle insertion strategy are proposed. When the earliest completion time of the process on the flexible equipment is the same, the processing time on the flexible equipment and the processing equipment of the same layer after the process are considered, and the processing equipment and processing time of the target process are determined. The example shows that, compared with the existing algorithm, the proposed algorithm can shorten the product completion time.
The integrated scheduling algorithm of flexible equipment network is difficult to reasonably select the relevant processes of processing equipment, which affects the completion time of products. An Integrated Scheduling algorithm for Flexible Equipment Network considering the Same layer after Process (SP-FENIS) is proposed. Firstly, the priority strategy of the reverse order layer is adopted, which assigns each process to the set of processes to be scheduled in the reverse layer. Then, the average reverse-order compact path strategy is proposed to determine the scheduling sequence of the processes to be scheduled in each reverse order layer. Finally, the earliest completion time strategy and equipment idle insertion strategy are proposed. When the earliest completion time of the process on the flexible equipment is the same, the processing time on the flexible equipment and the processing equipment of the same layer after the process are considered, and the processing equipment and processing time of the target process are determined. The example shows that, compared with the existing algorithm, the proposed algorithm can shorten the product completion time.
Available online , doi: 10.11999/JEIT231222
Abstract:
Time Series Classification(TSC) is one of the most important and challenging tasks in the field of data mining. Deep learning techniques have achieved revolutionary progress in natural language processing and computer vision, and have also demonstrated great potential in areas such as time series analysis. A detailed review of the latest research advances in deep learning-based TSC is provided in this paper. Firstly, key terms and related concepts are defined. Secondly, the latest time series classification models are classified from four perspectives of network architectures: multilayer perceptron, convolutional neural networks, recurrent neural networks, and attention mechanisms, along with their respective advantages and limitations. Additionally, the latest developments and challenges in time series classification in the fields of human activity recognition and electroencephalogram-based emotion recognition are outlined. Finally, the unresolved issues and future research directions when applying deep learning to time series data are discussed. This paper provides researchers with a reference for understanding the latest research dynamics, new technologies, and development trends in the deep learning-based time series classification field.
Time Series Classification(TSC) is one of the most important and challenging tasks in the field of data mining. Deep learning techniques have achieved revolutionary progress in natural language processing and computer vision, and have also demonstrated great potential in areas such as time series analysis. A detailed review of the latest research advances in deep learning-based TSC is provided in this paper. Firstly, key terms and related concepts are defined. Secondly, the latest time series classification models are classified from four perspectives of network architectures: multilayer perceptron, convolutional neural networks, recurrent neural networks, and attention mechanisms, along with their respective advantages and limitations. Additionally, the latest developments and challenges in time series classification in the fields of human activity recognition and electroencephalogram-based emotion recognition are outlined. Finally, the unresolved issues and future research directions when applying deep learning to time series data are discussed. This paper provides researchers with a reference for understanding the latest research dynamics, new technologies, and development trends in the deep learning-based time series classification field.
Available online , doi: 10.11999/JEIT230923
Abstract:
Federated learning is an emerging distributed machine learning framework that effectively solves the problems of data silos and privacy leakage in traditional machine learning by performing joint modeling training without leaving the user’s private data out of the domain. However, federated learning suffers from the problem of training-lagged clients dragging down the global training speed. Related research has proposed asynchronous federated learning, which allows the users to upload to the server and participate in the aggregation task as soon as they finish updating their models locally, without waiting for the other users. However, asynchronous federated learning also suffers from the inability to recognize malicious models uploaded by malicious users and the problem of leaking user’s privacy. To address these issues, a privacy-preserving Secure Aggregation scheme for asynchronous Federated Learning(SAFL) is designed. The users add perturbations to locally trained models and upload the perturbed models to the server. The server detects and rejects the malicious users through a poisoning detection algorithm to achieve Secure Aggregation(SA). Finally, theoretical analysis and experiments show that in the scenario of asynchronous federated learning, the proposed scheme can effectively detect malicious users while protecting the privacy of users’ local models and reducing the risk of privacy leakage. The proposed scheme has also a significant improvement in the accuracy of the model compared with other schemes.
Federated learning is an emerging distributed machine learning framework that effectively solves the problems of data silos and privacy leakage in traditional machine learning by performing joint modeling training without leaving the user’s private data out of the domain. However, federated learning suffers from the problem of training-lagged clients dragging down the global training speed. Related research has proposed asynchronous federated learning, which allows the users to upload to the server and participate in the aggregation task as soon as they finish updating their models locally, without waiting for the other users. However, asynchronous federated learning also suffers from the inability to recognize malicious models uploaded by malicious users and the problem of leaking user’s privacy. To address these issues, a privacy-preserving Secure Aggregation scheme for asynchronous Federated Learning(SAFL) is designed. The users add perturbations to locally trained models and upload the perturbed models to the server. The server detects and rejects the malicious users through a poisoning detection algorithm to achieve Secure Aggregation(SA). Finally, theoretical analysis and experiments show that in the scenario of asynchronous federated learning, the proposed scheme can effectively detect malicious users while protecting the privacy of users’ local models and reducing the risk of privacy leakage. The proposed scheme has also a significant improvement in the accuracy of the model compared with other schemes.
Available online , doi: 10.11999/JEIT230992
Abstract:
To overcome the drawback of poor adaptability of existing unbiased self-synchronous scrambling code recognition algorithms at low Signal-to-Noise Ratios (SNR), a soft-judgement recognition method based on the log conditional likelihood ratio is proposed. Firstly, the linear constraint equations for the pairwise even-vector product of the self-synchronous scrambler of linear grouping codes and the self-synchronous scrambler of convolutional codes are constructed, and then the log likelihood ratio function is introduced, the log conditional likelihood ratio function based on the soft judgment is constructed, and the distribution characteristics of its mean and variance are analyzed. Finally the identification of self-synchronous scrambler of linear grouping codes and self-synchronous scrambler of convolutional codes is accomplished through binary assumption and the derived coresponding judgement threshold value. The simulations show that the proposed algorithm is able to complete the recognition of generating polynomials at low signal-to-noise ratios, and has a good low signal-to-noise adaptation capability. Compared with the recognition method based on solving the cost function, the recognition rate of the algorithm is greatly improved at signal-to-noise ratios below 3 dB, and when the recognition rate is 90%, the proposed algorithm achieves a performance gain of about 3 dB.
To overcome the drawback of poor adaptability of existing unbiased self-synchronous scrambling code recognition algorithms at low Signal-to-Noise Ratios (SNR), a soft-judgement recognition method based on the log conditional likelihood ratio is proposed. Firstly, the linear constraint equations for the pairwise even-vector product of the self-synchronous scrambler of linear grouping codes and the self-synchronous scrambler of convolutional codes are constructed, and then the log likelihood ratio function is introduced, the log conditional likelihood ratio function based on the soft judgment is constructed, and the distribution characteristics of its mean and variance are analyzed. Finally the identification of self-synchronous scrambler of linear grouping codes and self-synchronous scrambler of convolutional codes is accomplished through binary assumption and the derived coresponding judgement threshold value. The simulations show that the proposed algorithm is able to complete the recognition of generating polynomials at low signal-to-noise ratios, and has a good low signal-to-noise adaptation capability. Compared with the recognition method based on solving the cost function, the recognition rate of the algorithm is greatly improved at signal-to-noise ratios below 3 dB, and when the recognition rate is 90%, the proposed algorithm achieves a performance gain of about 3 dB.
Available online , doi: 10.11999/JEIT231047
Abstract:
To address the challenges of poor recognition effect of the infrared substation equipment image caused by multi-target, small target and occlusion target in complex background, an infrared image recognition method for substation equipment based on CenterNet is proposed. By combining the Adaptive Spatial Feature Fusion(ASFF) module and Feature Pyramid Networks (FPN), a feature fusion network with the structure of ASFF+FPN is constructed, and the cross-scale feature fusion capability of the model for multi-target and small target is enhanced, which excludes background information. To improve the feature capturing ability of occluding targets, the global attention mechanism is introduced to the feature fusion network to enhance target saliency. Additionally, depthwise separable convolution is introduced to reduce parameters number and model inference time, and a lightweight model is achieved. Finally, the problem of poor sensitivity to obscured targets is overcame by using the distribution focal loss function, and the convergence speed and recognition accuracy is improved. Tests are performed on a self-built dataset containing seven infrared substation equipment images. Experimental results demonstrate that the proposed algorithm achieves a recognition accuracy of 95.19%, an improvement of 3.55% compared with the original algorithm, while it only has 32.52M model parameters. Furthermore, the method shows significant advantages in recognition accuracy and algorithm complexity, over four main target recognition algorithms.
To address the challenges of poor recognition effect of the infrared substation equipment image caused by multi-target, small target and occlusion target in complex background, an infrared image recognition method for substation equipment based on CenterNet is proposed. By combining the Adaptive Spatial Feature Fusion(ASFF) module and Feature Pyramid Networks (FPN), a feature fusion network with the structure of ASFF+FPN is constructed, and the cross-scale feature fusion capability of the model for multi-target and small target is enhanced, which excludes background information. To improve the feature capturing ability of occluding targets, the global attention mechanism is introduced to the feature fusion network to enhance target saliency. Additionally, depthwise separable convolution is introduced to reduce parameters number and model inference time, and a lightweight model is achieved. Finally, the problem of poor sensitivity to obscured targets is overcame by using the distribution focal loss function, and the convergence speed and recognition accuracy is improved. Tests are performed on a self-built dataset containing seven infrared substation equipment images. Experimental results demonstrate that the proposed algorithm achieves a recognition accuracy of 95.19%, an improvement of 3.55% compared with the original algorithm, while it only has 32.52M model parameters. Furthermore, the method shows significant advantages in recognition accuracy and algorithm complexity, over four main target recognition algorithms.
Available online , doi: 10.11999/JEIT221203
Abstract:
To overcome the limitation of the Federated Learning (FL) when the data and model of each client are all heterogenous and improve the accuracy, a personalized Federated learning algorithm with Collation game and Knowledge distillation (pFedCK) is proposed. Firstly, each client uploads its soft-predict on public dataset and download the most correlative of the k soft-predict. Then, this method apply the shapley value from collation game to measure the multi-wise influences among clients and quantify their marginal contribution to others on personalized learning performance. Lastly, each client identify it’s optimal coalition and then distill the knowledge to local model and train on private dataset. The results show that compared with the state-of-the-art algorithm, this approach can achieve superior personalized accuracy and can improve by about 10%.
To overcome the limitation of the Federated Learning (FL) when the data and model of each client are all heterogenous and improve the accuracy, a personalized Federated learning algorithm with Collation game and Knowledge distillation (pFedCK) is proposed. Firstly, each client uploads its soft-predict on public dataset and download the most correlative of the k soft-predict. Then, this method apply the shapley value from collation game to measure the multi-wise influences among clients and quantify their marginal contribution to others on personalized learning performance. Lastly, each client identify it’s optimal coalition and then distill the knowledge to local model and train on private dataset. The results show that compared with the state-of-the-art algorithm, this approach can achieve superior personalized accuracy and can improve by about 10%.
Available online , doi: 10.11999/JEIT210265
Abstract:
The application of Frequency Diverse Array and Multiple Input Multiple Output (FDA-MIMO) radar to achieve range-angle estimation of target has attracted more and more attention. The FDA can simultaneously obtain the degree of freedom of transmitting beam pattern in angle and range. However, its performance is degraded due to the periodicity and time-varying of the beam pattern. Therefore, an improved Estimating Signal Parameter via Rotational Invariance Techniques (ESPRIT) algorithm to estimate the target’s parameters based on a new waveform synthesis model of the Time Modulation and Range Compensation FDA-MIMO (TMRC-FDA-MIMO) radar is proposed. Finally, the proposed method is compared with identical frequency increment FDA-MIMO radar system, logarithmically increased frequency offset FDA-MIMO radar system and MUltiple SIgnal Classification (MUSIC) algorithm through the Cramer Rao lower bound and root mean square error of range and angle estimation, and the excellent performance of the proposed method is verified.
The application of Frequency Diverse Array and Multiple Input Multiple Output (FDA-MIMO) radar to achieve range-angle estimation of target has attracted more and more attention. The FDA can simultaneously obtain the degree of freedom of transmitting beam pattern in angle and range. However, its performance is degraded due to the periodicity and time-varying of the beam pattern. Therefore, an improved Estimating Signal Parameter via Rotational Invariance Techniques (ESPRIT) algorithm to estimate the target’s parameters based on a new waveform synthesis model of the Time Modulation and Range Compensation FDA-MIMO (TMRC-FDA-MIMO) radar is proposed. Finally, the proposed method is compared with identical frequency increment FDA-MIMO radar system, logarithmically increased frequency offset FDA-MIMO radar system and MUltiple SIgnal Classification (MUSIC) algorithm through the Cramer Rao lower bound and root mean square error of range and angle estimation, and the excellent performance of the proposed method is verified.
Available online , doi: 10.11999/JEIT230652
Abstract:
In tasks such as precise guidance and obstacle avoidance navigation based on optical compound eyes, the calibration of optical compound eyes plays a crucial role in achieving high accuracy. The classical Zhang's calibration method requires each ommatidium of the optical compound eyes to observe a complete chessboard pattern. However, the complexity of the optical compound eye structure makes it difficult to satisfy this requirement in practical applications. In this paper, a joint internal and external parameters calibration algorithm of optical compound eyes based on a random noise calibration pattern is proposed. This algorithm utilizes the local information captured by the ommatidia when photographing the random noise calibration pattern, enabling simple and fast calibration for optical compound eyes with arbitrary configurations and numbers of ommatidia. To improve the robustness of the calibration, a multi-threshold matching mechanism is introduced to address the issue of sparse feature point quantity in ommatidial visual fields leading to matching failures. Moreover, an error model for the joint internal and external parameters calibration of optical compound eyes is presented to evaluate the accuracy of the proposed algorithm. Experimental comparisons with Zhang’s calibration method demonstrate the robustness of the proposed algorithm. Furthermore, the high accuracy of the proposed joint calibration algorithm is validated in a physical system of optical compound eyes.
In tasks such as precise guidance and obstacle avoidance navigation based on optical compound eyes, the calibration of optical compound eyes plays a crucial role in achieving high accuracy. The classical Zhang's calibration method requires each ommatidium of the optical compound eyes to observe a complete chessboard pattern. However, the complexity of the optical compound eye structure makes it difficult to satisfy this requirement in practical applications. In this paper, a joint internal and external parameters calibration algorithm of optical compound eyes based on a random noise calibration pattern is proposed. This algorithm utilizes the local information captured by the ommatidia when photographing the random noise calibration pattern, enabling simple and fast calibration for optical compound eyes with arbitrary configurations and numbers of ommatidia. To improve the robustness of the calibration, a multi-threshold matching mechanism is introduced to address the issue of sparse feature point quantity in ommatidial visual fields leading to matching failures. Moreover, an error model for the joint internal and external parameters calibration of optical compound eyes is presented to evaluate the accuracy of the proposed algorithm. Experimental comparisons with Zhang’s calibration method demonstrate the robustness of the proposed algorithm. Furthermore, the high accuracy of the proposed joint calibration algorithm is validated in a physical system of optical compound eyes.
Available online , doi: 10.11999/JEIT231005
Abstract:
To solve the problem of low spectrum utilization of multi-node autonomous frequency decision-making in the dynamic electromagnetic countermeasure environment, the research on intelligent cooperative spectrum allocation technology for non-complete electromagnetic information is carried out, which improves spectrum utilization through multi-node intelligent collaboration. Firstly, the spectrum allocation problem is modelled as an optimization problem to maximize the frequency-using equipment, and secondly, a resource decision-making algorithm based on the multi-node cooperative diversion experience repetition mechanism (Cooperation- Deep double Q-network, Co-DDQN) is proposed. This algorithm evaluates the historical experience data based on the cooperative diversion function and is trained by a hierarchical experience pool, so that each agent can form a lightweight cooperative decision-making ability under self-observation, and solve the problem of inconsistency between the optimization direction of multi-node decision-making and the overall optimization goal under low-visibility conditions. Besides, a hybrid reward function based on confidence allocation is designed, and each node considers itself when the decision is made, which can reduce the emergence of lazy nodes, explore a better overall action strategy, and further improve the system efficiency. Simulation results show that when the number of nodes is 20, the number of accessible devices of the proposed algorithm outperforms the global greedy algorithm and the genetic algorithm, and the difference with the centralized spectrum allocation algorithm with complete information sharing is within 5%, which is more suitable for cooperative spectrum allocation of low-visibility nodes.
To solve the problem of low spectrum utilization of multi-node autonomous frequency decision-making in the dynamic electromagnetic countermeasure environment, the research on intelligent cooperative spectrum allocation technology for non-complete electromagnetic information is carried out, which improves spectrum utilization through multi-node intelligent collaboration. Firstly, the spectrum allocation problem is modelled as an optimization problem to maximize the frequency-using equipment, and secondly, a resource decision-making algorithm based on the multi-node cooperative diversion experience repetition mechanism (Cooperation- Deep double Q-network, Co-DDQN) is proposed. This algorithm evaluates the historical experience data based on the cooperative diversion function and is trained by a hierarchical experience pool, so that each agent can form a lightweight cooperative decision-making ability under self-observation, and solve the problem of inconsistency between the optimization direction of multi-node decision-making and the overall optimization goal under low-visibility conditions. Besides, a hybrid reward function based on confidence allocation is designed, and each node considers itself when the decision is made, which can reduce the emergence of lazy nodes, explore a better overall action strategy, and further improve the system efficiency. Simulation results show that when the number of nodes is 20, the number of accessible devices of the proposed algorithm outperforms the global greedy algorithm and the genetic algorithm, and the difference with the centralized spectrum allocation algorithm with complete information sharing is within 5%, which is more suitable for cooperative spectrum allocation of low-visibility nodes.
Available online , doi: 10.11999/JEIT201066
Abstract:
To solve the problem of Group Repetition Interval (GRI) selection in the construction of the enhanced LORAN (eLORAN) system supplementary transmission station, a screening algorithm based on cross interference rate is proposed mainly from the mathematical point of view. Firstly, this method considers the requirement of second information, and on this basis, conducts a first screening by comparing the mutual Cross Rate Interference (CRI) with the adjacent Loran-C stations in the neighboring countries. Secondly, a second screening is conducted through permutation and pairwise comparison. Finally, the optimal GRI combination scheme is given by considering the requirements of data rate and system specification. Then, in view of the high-precision timing requirements for the new eLORAN system, an optimized selection is made in multiple optimal combinations. The analysis results show that the average interference rate of the optimal combination scheme obtained by this algorithm is comparable to that between the current navigation chains and can take into account the timing requirements, which can provide referential suggestions and theoretical basis for the construction of high-precision ground-based timing system.
To solve the problem of Group Repetition Interval (GRI) selection in the construction of the enhanced LORAN (eLORAN) system supplementary transmission station, a screening algorithm based on cross interference rate is proposed mainly from the mathematical point of view. Firstly, this method considers the requirement of second information, and on this basis, conducts a first screening by comparing the mutual Cross Rate Interference (CRI) with the adjacent Loran-C stations in the neighboring countries. Secondly, a second screening is conducted through permutation and pairwise comparison. Finally, the optimal GRI combination scheme is given by considering the requirements of data rate and system specification. Then, in view of the high-precision timing requirements for the new eLORAN system, an optimized selection is made in multiple optimal combinations. The analysis results show that the average interference rate of the optimal combination scheme obtained by this algorithm is comparable to that between the current navigation chains and can take into account the timing requirements, which can provide referential suggestions and theoretical basis for the construction of high-precision ground-based timing system.
