Pub Date : 2024-09-09DOI: 10.23919/cje.2022.00.135
Haowei Meng;Ning Xin;Hao Qin;Di Zhao
Optimization-based radio resource management (RRM) has shown significant performance gains on high-throughput satellites (HTSs). However, as the number of allocable on-board resources increases, traditional RRM is difficult to apply in real satellite systems due to its intense computational complexity. Deep reinforcement learning (DRL) is a promising solution for the resource allocation problem due to its model-free advantages. Nevertheless, the action space faced by DRL increases exponentially with the increase of communication scale, which leads to an excessive exploration cost of the algorithm. In this paper, we propose a recursive frequency resource allocation algorithm based on long-short term memory (LSTM) and proximal policy optimization (PPO), called PPO-RA-LOOP, where RA means resource allocation and LOOP means the algorithm outputs actions in a recursive manner. Specifically, the PPO algorithm uses LSTM network to recursively generate sub-actions about frequency resource allocation for each beam, which significantly cuts down the action space. In addition, the LSTM-based recursive architecture allows PPO to better allocate the next frequency resource by using the generated sub-actions information as a prior knowledge, which reduces the complexity of the neural network. The simulation results show that PPO-RA-LOOP achieved higher spectral efficiency and system satisfaction compared with other frequency allocation algorithms.
{"title":"A Recursive DRL-Based Resource Allocation Method for Multibeam Satellite Communication Systems","authors":"Haowei Meng;Ning Xin;Hao Qin;Di Zhao","doi":"10.23919/cje.2022.00.135","DOIUrl":"https://doi.org/10.23919/cje.2022.00.135","url":null,"abstract":"Optimization-based radio resource management (RRM) has shown significant performance gains on high-throughput satellites (HTSs). However, as the number of allocable on-board resources increases, traditional RRM is difficult to apply in real satellite systems due to its intense computational complexity. Deep reinforcement learning (DRL) is a promising solution for the resource allocation problem due to its model-free advantages. Nevertheless, the action space faced by DRL increases exponentially with the increase of communication scale, which leads to an excessive exploration cost of the algorithm. In this paper, we propose a recursive frequency resource allocation algorithm based on long-short term memory (LSTM) and proximal policy optimization (PPO), called PPO-RA-LOOP, where RA means resource allocation and LOOP means the algorithm outputs actions in a recursive manner. Specifically, the PPO algorithm uses LSTM network to recursively generate sub-actions about frequency resource allocation for each beam, which significantly cuts down the action space. In addition, the LSTM-based recursive architecture allows PPO to better allocate the next frequency resource by using the generated sub-actions information as a prior knowledge, which reduces the complexity of the neural network. The simulation results show that PPO-RA-LOOP achieved higher spectral efficiency and system satisfaction compared with other frequency allocation algorithms.","PeriodicalId":50701,"journal":{"name":"Chinese Journal of Electronics","volume":null,"pages":null},"PeriodicalIF":1.6,"publicationDate":"2024-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10669755","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142159987","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-09DOI: 10.23919/cje.2022.00.407
Yuhao Zhou;Zhenxue He;Jianhui Jiang;Xiaojun Zhao;Fan Zhang;Limin Xiao;Xiang Wang
Area has become one of the main bottlenecks restricting the development of integrated circuits. The area optimization approaches of existing XNOR/OR-based mixed polarity Reed-Muller (MPRM) circuits have poor optimization effect and efficiency. Given that the area optimization of MPRM logic circuits is a combinatorial optimization problem, we propose a whole annealing adaptive bacterial foraging algorithm (WAA-BFA), which includes individual evolution based on Markov chain and Metropolis acceptance criteria, and individual mutation based on adaptive probability. To address the issue of low conversion efficiency in existing polarity conversion approaches, we introduce a fast polarity conversion algorithm (FPCA). Moreover, we present an MPRM circuits area optimization approach that uses the FPCA and WAA-BFA to search for the best polarity corresponding to the minimum circuits area. Experimental results demonstrate that the proposed MPRM circuits area optimization approach is effective and can be used as a promising EDA tool.
{"title":"An Efficient and Fast Area Optimization Approach for Mixed Polarity Reed-Muller Logic Circuits","authors":"Yuhao Zhou;Zhenxue He;Jianhui Jiang;Xiaojun Zhao;Fan Zhang;Limin Xiao;Xiang Wang","doi":"10.23919/cje.2022.00.407","DOIUrl":"https://doi.org/10.23919/cje.2022.00.407","url":null,"abstract":"Area has become one of the main bottlenecks restricting the development of integrated circuits. The area optimization approaches of existing XNOR/OR-based mixed polarity Reed-Muller (MPRM) circuits have poor optimization effect and efficiency. Given that the area optimization of MPRM logic circuits is a combinatorial optimization problem, we propose a whole annealing adaptive bacterial foraging algorithm (WAA-BFA), which includes individual evolution based on Markov chain and Metropolis acceptance criteria, and individual mutation based on adaptive probability. To address the issue of low conversion efficiency in existing polarity conversion approaches, we introduce a fast polarity conversion algorithm (FPCA). Moreover, we present an MPRM circuits area optimization approach that uses the FPCA and WAA-BFA to search for the best polarity corresponding to the minimum circuits area. Experimental results demonstrate that the proposed MPRM circuits area optimization approach is effective and can be used as a promising EDA tool.","PeriodicalId":50701,"journal":{"name":"Chinese Journal of Electronics","volume":null,"pages":null},"PeriodicalIF":1.6,"publicationDate":"2024-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10669742","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142159955","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Object detection is an important task in drone vision. Since the number of objects and their scales always vary greatly in the drone-captured video, small object-oriented feature becomes the bottleneck of model performance, and most existing object detectors tend to underperform in drone-vision scenes. To solve these problems, we propose a novel detector named YOLO-Drone. In the proposed detector, the backbone of YOLO is firstly replaced with ConvNeXt, which is the state-of-the-art one to extract more discriminative features. Then, a novel scale-aware attention (SAA) module is designed in detection head to solve the large disparity scale problem. A scale-sensitive loss (SSL) is also introduced to put more emphasis on object scale to enhance the discriminative ability of the proposed detector. Experimental results on the latest VisDrone 2022 test-challenge dataset (detection track) show that our detector can achieve average precision (AP) of 39.43%, which is tied with the previous state-of-the-art, meanwhile, reducing 39.8% of the computational cost.
{"title":"YOLO-Drone: A Scale-Aware Detector for Drone Vision","authors":"Yutong Li;Miao Ma;Shichang Liu;Chao Yao;Longjiang Guo","doi":"10.23919/cje.2023.00.254","DOIUrl":"10.23919/cje.2023.00.254","url":null,"abstract":"Object detection is an important task in drone vision. Since the number of objects and their scales always vary greatly in the drone-captured video, small object-oriented feature becomes the bottleneck of model performance, and most existing object detectors tend to underperform in drone-vision scenes. To solve these problems, we propose a novel detector named YOLO-Drone. In the proposed detector, the backbone of YOLO is firstly replaced with ConvNeXt, which is the state-of-the-art one to extract more discriminative features. Then, a novel scale-aware attention (SAA) module is designed in detection head to solve the large disparity scale problem. A scale-sensitive loss (SSL) is also introduced to put more emphasis on object scale to enhance the discriminative ability of the proposed detector. Experimental results on the latest VisDrone 2022 test-challenge dataset (detection track) show that our detector can achieve average precision (AP) of 39.43%, which is tied with the previous state-of-the-art, meanwhile, reducing 39.8% of the computational cost.","PeriodicalId":50701,"journal":{"name":"Chinese Journal of Electronics","volume":null,"pages":null},"PeriodicalIF":1.6,"publicationDate":"2024-07-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10606208","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141852824","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Existing high-precision object detection algorithms for UAV (unmanned aerial vehicle) aerial images often have a large number of parameters and heavy weight, which makes it difficult to be applied to mobile devices. We propose three YOLO-based lightweight object detection networks for UAVs, named YOLO-L, YOLO-S, and YOLO-M, respectively. In YOLO-L, we adopt a deconvolution approach to explore suitable upsampling rules during training to improve the detection accuracy. The convolution-batch normalization-SiLU activation function (CBS) structure is replaced with Ghost CBS to reduce the number of parameters and weight, meanwhile Maxpool maximum pooling operation is proposed to replace the CBS structure to avoid generating parameters and weight. YOLO-S greatly reduces the weight of the network by directly introducing CSPGhostNeck residual structures, so that the parameters and weight are respectively decreased by about 15% at the expense of 2.4% mAP. And YOLO-M adopts the CSPGhostNeck residual structure and deconvolution to reduce parameters by 5.6% and weight by 5.7%, while mAP only by 1.8%. The results show that the three lightweight detection networks proposed in this paper have good performance in UAV aerial image object detection task.
{"title":"Lightweight Object Detection Networks for UAV Aerial Images Based on YOLO","authors":"Yanshan Li;Jiarong Wang;Kunhua Zhang;Jiawei Yi;Miaomiao Wei;Lirong Zheng;Weixin Xie","doi":"10.23919/cje.2022.00.300","DOIUrl":"10.23919/cje.2022.00.300","url":null,"abstract":"Existing high-precision object detection algorithms for UAV (unmanned aerial vehicle) aerial images often have a large number of parameters and heavy weight, which makes it difficult to be applied to mobile devices. We propose three YOLO-based lightweight object detection networks for UAVs, named YOLO-L, YOLO-S, and YOLO-M, respectively. In YOLO-L, we adopt a deconvolution approach to explore suitable upsampling rules during training to improve the detection accuracy. The convolution-batch normalization-SiLU activation function (CBS) structure is replaced with Ghost CBS to reduce the number of parameters and weight, meanwhile Maxpool maximum pooling operation is proposed to replace the CBS structure to avoid generating parameters and weight. YOLO-S greatly reduces the weight of the network by directly introducing CSPGhostNeck residual structures, so that the parameters and weight are respectively decreased by about 15% at the expense of 2.4% mAP. And YOLO-M adopts the CSPGhostNeck residual structure and deconvolution to reduce parameters by 5.6% and weight by 5.7%, while mAP only by 1.8%. The results show that the three lightweight detection networks proposed in this paper have good performance in UAV aerial image object detection task.","PeriodicalId":50701,"journal":{"name":"Chinese Journal of Electronics","volume":null,"pages":null},"PeriodicalIF":1.6,"publicationDate":"2024-07-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10606207","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141839371","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-07-22DOI: 10.23919/cje.2023.00.041
Lingshuo Meng;Xueluan Gong;Yanjiao Chen
Backdoor attacks pose great threats to deep neural network models. All existing backdoor attacks are designed for unstructured data (image, voice, and text), but not structured tabular data, which has wide real-world applications, e.g., recommendation systems, fraud detection, and click-through rate prediction. To bridge this research gap, we make the first attempt to design a backdoor attack framework, named BAD-FM, for tabular data prediction models. Unlike images or voice samples composed of homogeneous pixels or signals with continuous values, tabular data samples contain well-defined heterogeneous fields that are usually sparse and discrete. Tabular data prediction models do not solely rely on deep networks but combine shallow components (e.g., factorization machine, FM) with deep components to capture sophisticated feature interactions among fields. To tailor the backdoor attack framework to tabular data models, we carefully design field selection and trigger formation algorithms to intensify the influence of the trigger on the backdoored model. We evaluate BAD-FM with extensive experiments on four datasets, i.e., HUAWEI, Criteo, Avazu, and KDD. The results show that BAD-FM can achieve an attack success rate as high as 100% at a poisoning ratio of 0.001%, outperforming baselines adapted from existing backdoor attacks against unstructured data models. As tabular data prediction models are widely adopted in finance and commerce, our work may raise alarms on the potential risks of these models and spur future research on defenses.
{"title":"BAD-FM: Backdoor Attacks Against Factorization-Machine Based Neural Network for Tabular Data Prediction","authors":"Lingshuo Meng;Xueluan Gong;Yanjiao Chen","doi":"10.23919/cje.2023.00.041","DOIUrl":"10.23919/cje.2023.00.041","url":null,"abstract":"Backdoor attacks pose great threats to deep neural network models. All existing backdoor attacks are designed for unstructured data (image, voice, and text), but not structured tabular data, which has wide real-world applications, e.g., recommendation systems, fraud detection, and click-through rate prediction. To bridge this research gap, we make the first attempt to design a backdoor attack framework, named BAD-FM, for tabular data prediction models. Unlike images or voice samples composed of homogeneous pixels or signals with continuous values, tabular data samples contain well-defined heterogeneous fields that are usually sparse and discrete. Tabular data prediction models do not solely rely on deep networks but combine shallow components (e.g., factorization machine, FM) with deep components to capture sophisticated feature interactions among fields. To tailor the backdoor attack framework to tabular data models, we carefully design field selection and trigger formation algorithms to intensify the influence of the trigger on the backdoored model. We evaluate BAD-FM with extensive experiments on four datasets, i.e., HUAWEI, Criteo, Avazu, and KDD. The results show that BAD-FM can achieve an attack success rate as high as 100% at a poisoning ratio of 0.001%, outperforming baselines adapted from existing backdoor attacks against unstructured data models. As tabular data prediction models are widely adopted in finance and commerce, our work may raise alarms on the potential risks of these models and spur future research on defenses.","PeriodicalId":50701,"journal":{"name":"Chinese Journal of Electronics","volume":null,"pages":null},"PeriodicalIF":1.6,"publicationDate":"2024-07-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10606191","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141852860","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Mobile edge computing (MEC) provides edge services to users in a distributed and on-demand way. Due to the heterogeneity of edge applications, deploying latency and resource-intensive applications on resource-constrained devices is a key challenge for service providers. This is especially true when underlying edge infrastructures are fault and error-prone. In this paper, we propose a fault tolerance approach named DFGP, for enforcing mobile service fault-tolerance in MEC. It synthesizes a generative optimization network (GON) model for predicting resource failure and a deep deterministic policy gradient (DDPG) model for yielding preemptive migration decisions. We show through extensive simulation experiments that DFGP is more effective in fault detection and guaranteeing quality of service, in terms of fault detection accuracy, migration efficiency, task migration time, task scheduling time, and energy consumption than other existing methods.
{"title":"A Deep Deterministic Policy Gradient-Based Method for Enforcing Service Fault-Tolerance in MEC","authors":"Tingyan Long;Peng Chen;Yunni Xia;Yong Ma;Xiaoning Sun;Jiale Zhao;Yifei Lyu","doi":"10.23919/cje.2023.00.105","DOIUrl":"10.23919/cje.2023.00.105","url":null,"abstract":"Mobile edge computing (MEC) provides edge services to users in a distributed and on-demand way. Due to the heterogeneity of edge applications, deploying latency and resource-intensive applications on resource-constrained devices is a key challenge for service providers. This is especially true when underlying edge infrastructures are fault and error-prone. In this paper, we propose a fault tolerance approach named DFGP, for enforcing mobile service fault-tolerance in MEC. It synthesizes a generative optimization network (GON) model for predicting resource failure and a deep deterministic policy gradient (DDPG) model for yielding preemptive migration decisions. We show through extensive simulation experiments that DFGP is more effective in fault detection and guaranteeing quality of service, in terms of fault detection accuracy, migration efficiency, task migration time, task scheduling time, and energy consumption than other existing methods.","PeriodicalId":50701,"journal":{"name":"Chinese Journal of Electronics","volume":null,"pages":null},"PeriodicalIF":1.6,"publicationDate":"2024-07-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10606213","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141843834","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-07-22DOI: 10.23919/cje.2022.00.412
Ying Chen;Jintao Hu;Jie Zhao;Geyong Min
Low earth orbit (LEO) satellite edge computing can overcome communication difficulties in harsh environments, which lack the support of terrestrial communication infrastructure. It is an indispensable option for achieving worldwide wireless communication coverage in the future. To improve the quality-of-service (QoS) for Internet-of-things (IoT) devices, we combine LEO satellite edge computing and ground communication systems to provide network services for IoT devices in harsh environments. We study the QoS-aware computation offloading (QCO) problem for IoT devices in LEO satellite edge computing. Then we investigate the computation offloading strategy for IoT devices that can minimize the total QoS cost of all devices while satisfying multiple constraints, such as the computing resource constraint, delay constraint, and energy consumption constraint. We formulate the QoS-aware computation offloading problem as a game model named QCO game based on the non-cooperative competition game among IoT devices. We analyze the finite improvement property of the QCO game and prove that there is a Nash equilibrium for the QCO game. We propose a distributed QoS-aware computation offloading (DQCO) algorithm for the QCO game. Experimental results show that the DQCO algorithm can effectively reduce the total QoS cost of IoT devices.
{"title":"QoS-Aware Computation Offloading in LEO Satellite Edge Computing for IoT: A Game-Theoretical Approach","authors":"Ying Chen;Jintao Hu;Jie Zhao;Geyong Min","doi":"10.23919/cje.2022.00.412","DOIUrl":"10.23919/cje.2022.00.412","url":null,"abstract":"Low earth orbit (LEO) satellite edge computing can overcome communication difficulties in harsh environments, which lack the support of terrestrial communication infrastructure. It is an indispensable option for achieving worldwide wireless communication coverage in the future. To improve the quality-of-service (QoS) for Internet-of-things (IoT) devices, we combine LEO satellite edge computing and ground communication systems to provide network services for IoT devices in harsh environments. We study the QoS-aware computation offloading (QCO) problem for IoT devices in LEO satellite edge computing. Then we investigate the computation offloading strategy for IoT devices that can minimize the total QoS cost of all devices while satisfying multiple constraints, such as the computing resource constraint, delay constraint, and energy consumption constraint. We formulate the QoS-aware computation offloading problem as a game model named QCO game based on the non-cooperative competition game among IoT devices. We analyze the finite improvement property of the QCO game and prove that there is a Nash equilibrium for the QCO game. We propose a distributed QoS-aware computation offloading (DQCO) algorithm for the QCO game. Experimental results show that the DQCO algorithm can effectively reduce the total QoS cost of IoT devices.","PeriodicalId":50701,"journal":{"name":"Chinese Journal of Electronics","volume":null,"pages":null},"PeriodicalIF":1.6,"publicationDate":"2024-07-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10606190","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141847454","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-07-22DOI: 10.23919/cje.2022.00.452
Hao Li;Yi Zhang;Jinwei Wang;Weiming Zhang;Xiangyang Luo
Existing deep learning-based steganography detection methods utilize convolution to automatically capture and learn steganographic features, yielding higher detection efficiency compared to manually designed steganography detection methods. Detection methods based on convolutional neural network frameworks can extract global features by increasing the network's depth and width. These frameworks are not highly sensitive to global features and can lead to significant resource consumption. This manuscript proposes a lightweight steganography detection method based on multiple residual structures and Transformer (ResFormer). A multi-residuals block based on channel rearrangement is designed in the preprocessing layer. Multiple residuals are used to enrich the residual features and channel shuffle is used to enhance the feature representation capability. A lightweight convolutional and Transformer feature extraction backbone is constructed, which reduces the computational and parameter complexity of the network by employing depth-wise separable convolutions. This backbone integrates local and global image features through the fusion of convolutional layers and Transformer, enhancing the network's ability to learn global features and effectively enriching feature diversity. An effective weighted loss function is introduced for learning both local and global features, BiasLoss loss function is used to give full play to the role of feature diversity in classification, and cross-entropy loss function and contrast loss function are organically combined to enhance the expression ability of features. Based on BossBase-1.01, BOWS2 and ALASKA#2, extensive experiments are conducted on the stego images generated by spatial and JPEG domain adaptive steganographic algorithms, employing both classical and state-of-the-art steganalysis techniques. The experimental results demonstrate that compared to the SRM, SRNet, SiaStegNet, CSANet, LWENet, and SiaIRNet methods, the proposed ResFormer method achieves the highest reduction in the parameter, up to 91.82%. It achieves the highest improvement in detection accuracy, up to 5.10%. Compared to the SRNet and EWNet methods, the proposed ResFormer method achieves an improvement in detection accuracy for the J-UNIWARD algorithm by 5.78% and 6.24%, respectively.
{"title":"Lightweight Steganography Detection Method Based on Multiple Residual Structures and Transformer","authors":"Hao Li;Yi Zhang;Jinwei Wang;Weiming Zhang;Xiangyang Luo","doi":"10.23919/cje.2022.00.452","DOIUrl":"10.23919/cje.2022.00.452","url":null,"abstract":"Existing deep learning-based steganography detection methods utilize convolution to automatically capture and learn steganographic features, yielding higher detection efficiency compared to manually designed steganography detection methods. Detection methods based on convolutional neural network frameworks can extract global features by increasing the network's depth and width. These frameworks are not highly sensitive to global features and can lead to significant resource consumption. This manuscript proposes a lightweight steganography detection method based on multiple residual structures and Transformer (ResFormer). A multi-residuals block based on channel rearrangement is designed in the preprocessing layer. Multiple residuals are used to enrich the residual features and channel shuffle is used to enhance the feature representation capability. A lightweight convolutional and Transformer feature extraction backbone is constructed, which reduces the computational and parameter complexity of the network by employing depth-wise separable convolutions. This backbone integrates local and global image features through the fusion of convolutional layers and Transformer, enhancing the network's ability to learn global features and effectively enriching feature diversity. An effective weighted loss function is introduced for learning both local and global features, BiasLoss loss function is used to give full play to the role of feature diversity in classification, and cross-entropy loss function and contrast loss function are organically combined to enhance the expression ability of features. Based on BossBase-1.01, BOWS2 and ALASKA#2, extensive experiments are conducted on the stego images generated by spatial and JPEG domain adaptive steganographic algorithms, employing both classical and state-of-the-art steganalysis techniques. The experimental results demonstrate that compared to the SRM, SRNet, SiaStegNet, CSANet, LWENet, and SiaIRNet methods, the proposed ResFormer method achieves the highest reduction in the parameter, up to 91.82%. It achieves the highest improvement in detection accuracy, up to 5.10%. Compared to the SRNet and EWNet methods, the proposed ResFormer method achieves an improvement in detection accuracy for the J-UNIWARD algorithm by 5.78% and 6.24%, respectively.","PeriodicalId":50701,"journal":{"name":"Chinese Journal of Electronics","volume":null,"pages":null},"PeriodicalIF":1.6,"publicationDate":"2024-07-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10606201","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141847662","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-07-22DOI: 10.23919/cje.2022.00.451
Chenhao Lin;Xingliang Zhang;Chao Shen
With the increasing deployment of deep learning-based systems in various scenes, it is becoming important to conduct sufficient testing and evaluation of deep learning models to improve their interpretability and robustness. Recent studies have proposed different criteria and strategies for deep neural network (DNN) testing. However, they rarely conduct effective testing on the robustness of DNN models and lack interpretability. This paper proposes a new priority testing criterion, called DeepLogic, to analyze the robustness of the DNN models from the perspective of model interpretability. We first define the neural units in DNN with the highest average activation probability as “interpretable logic units”. We analyze the changes in these units to evaluate the model's robustness by conducting adversarial attacks. After that, the interpretable logic units of the inputs are taken as context attributes, and the probability distribution of the softmax layer in the model is taken as internal attributes to establish a comprehensive test prioritization framework. The weight fusion of context and internal factors is carried out, and the test cases are sorted according to this priority. The experimental results on four popular DNN models using eight testing metrics show that our DeepLogic significantly outperforms existing state-of-the-art methods.
{"title":"DeepLogic: Priority Testing of Deep Learning Through Interpretable Logic Units","authors":"Chenhao Lin;Xingliang Zhang;Chao Shen","doi":"10.23919/cje.2022.00.451","DOIUrl":"https://doi.org/10.23919/cje.2022.00.451","url":null,"abstract":"With the increasing deployment of deep learning-based systems in various scenes, it is becoming important to conduct sufficient testing and evaluation of deep learning models to improve their interpretability and robustness. Recent studies have proposed different criteria and strategies for deep neural network (DNN) testing. However, they rarely conduct effective testing on the robustness of DNN models and lack interpretability. This paper proposes a new priority testing criterion, called DeepLogic, to analyze the robustness of the DNN models from the perspective of model interpretability. We first define the neural units in DNN with the highest average activation probability as “interpretable logic units”. We analyze the changes in these units to evaluate the model's robustness by conducting adversarial attacks. After that, the interpretable logic units of the inputs are taken as context attributes, and the probability distribution of the softmax layer in the model is taken as internal attributes to establish a comprehensive test prioritization framework. The weight fusion of context and internal factors is carried out, and the test cases are sorted according to this priority. The experimental results on four popular DNN models using eight testing metrics show that our DeepLogic significantly outperforms existing state-of-the-art methods.","PeriodicalId":50701,"journal":{"name":"Chinese Journal of Electronics","volume":null,"pages":null},"PeriodicalIF":1.6,"publicationDate":"2024-07-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10606210","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141964672","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-07-22DOI: 10.23919/cje.2022.00.340
Yibo Li;Junhui Zhao;Jieyu Liao;Fajin Hu
Communication and sensing are basically required in intelligent transportation. The combination of two functions can provide a viable way in alleviating concerns about resource limitations. To achieve this, we propose an integrated sensing and communication (ISAC) system based on cellular vehicle-to-everything (C-V2X). We first analyze the feasibility of new radio (NR) waveform for ISAC system. We discuss the possibility of reusing NR waveform for sensing based on current NR-V2X standards. Ambiguity function is calculated to investigate the sensing performance limitation of NR waveform. A C-V2X-based ISAC system is then designed to realize the two tasks in vehicular network simultaneously. We formulate an integrated framework of vehicular communication and automotive sensing using the already-existing NR-V2X network. Based on the proposed ISAC framework, we develop a receiver algorithm for target detection/estimation and communication with minor modifications. We evaluate the performance of the proposed ISAC system with communication throughput, detection probability, and range/velocity estimation accuracy. Simulations show that the proposed system achieves high reliability communication with 99.9999% throughput and high accuracy sensing with errors below 1 m and 1 m/s in vehicle scenarios.
{"title":"Cellular V2X-Based Integrated Sensing and Communication System: Feasibility and Performance Analysis","authors":"Yibo Li;Junhui Zhao;Jieyu Liao;Fajin Hu","doi":"10.23919/cje.2022.00.340","DOIUrl":"10.23919/cje.2022.00.340","url":null,"abstract":"Communication and sensing are basically required in intelligent transportation. The combination of two functions can provide a viable way in alleviating concerns about resource limitations. To achieve this, we propose an integrated sensing and communication (ISAC) system based on cellular vehicle-to-everything (C-V2X). We first analyze the feasibility of new radio (NR) waveform for ISAC system. We discuss the possibility of reusing NR waveform for sensing based on current NR-V2X standards. Ambiguity function is calculated to investigate the sensing performance limitation of NR waveform. A C-V2X-based ISAC system is then designed to realize the two tasks in vehicular network simultaneously. We formulate an integrated framework of vehicular communication and automotive sensing using the already-existing NR-V2X network. Based on the proposed ISAC framework, we develop a receiver algorithm for target detection/estimation and communication with minor modifications. We evaluate the performance of the proposed ISAC system with communication throughput, detection probability, and range/velocity estimation accuracy. Simulations show that the proposed system achieves high reliability communication with 99.9999% throughput and high accuracy sensing with errors below 1 m and 1 m/s in vehicle scenarios.","PeriodicalId":50701,"journal":{"name":"Chinese Journal of Electronics","volume":null,"pages":null},"PeriodicalIF":1.6,"publicationDate":"2024-07-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10606214","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141851789","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：