Pub Date : 2021-12-17DOI: 10.1109/ICECE54449.2021.9674283
Jing-song Xiang, Xinhao Lyu
In order to eliminate the impact of atmospheric turbulence on the performance of free-space optical (FSO) communication systems, the multi-aperture receiving technique is wildly recognized as a powerful fading-mitigation technology. As one of the essential technologies in the multi-aperture receiver, digital coherent beam combining relies on the digital phase alignment algorithm to align the different versions of signals in phase. In this paper, the statistical model of combining efficiency for digital phase alignment is derived in multi-aperture FSO receivers by considering the phase alignment errors at each receiving aperture. It can be expressed as a linear function of chi-square distribution by Satterthwaite approximation. Based on this statistical model, we derive the exact expressions of the mean, variance, and probability density function of the combining efficiency. The simulation results show that this model is valuable and practical under the condition of the different number of aperture and signal-to-noise ratio combinations. Combining efficiency is also compared for equal gain combining diversity FSO systems with or without considering aperture selection.
{"title":"Statistical Model of Combining Efficiency for Digital Phase Alignment in Multi-Aperture Free-Space Coherent Optical Receivers","authors":"Jing-song Xiang, Xinhao Lyu","doi":"10.1109/ICECE54449.2021.9674283","DOIUrl":"https://doi.org/10.1109/ICECE54449.2021.9674283","url":null,"abstract":"In order to eliminate the impact of atmospheric turbulence on the performance of free-space optical (FSO) communication systems, the multi-aperture receiving technique is wildly recognized as a powerful fading-mitigation technology. As one of the essential technologies in the multi-aperture receiver, digital coherent beam combining relies on the digital phase alignment algorithm to align the different versions of signals in phase. In this paper, the statistical model of combining efficiency for digital phase alignment is derived in multi-aperture FSO receivers by considering the phase alignment errors at each receiving aperture. It can be expressed as a linear function of chi-square distribution by Satterthwaite approximation. Based on this statistical model, we derive the exact expressions of the mean, variance, and probability density function of the combining efficiency. The simulation results show that this model is valuable and practical under the condition of the different number of aperture and signal-to-noise ratio combinations. Combining efficiency is also compared for equal gain combining diversity FSO systems with or without considering aperture selection.","PeriodicalId":166178,"journal":{"name":"2021 IEEE 4th International Conference on Electronics and Communication Engineering (ICECE)","volume":"360 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133645064","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-12-17DOI: 10.1109/ICECE54449.2021.9674303
Jiahao Qu, Brian Sumali, Y. Mitsukura
Since the outbreak of COVID-19 in Wuhan, China in December 2019, a large number of patients have been seen worldwide, and the number of infections continues to show an increasing trend. The vast majority of COVID-19 patients will have fever, headache, and mild respiratory symptoms, but a small number of severely ill patients will experience respiratory distress and related complications, which seriously endanger their lives. The large number of patients also puts the healthcare system to the test. To maximize the protection of patients’ lives and the effective use of medical resources, this study collected blood data from 313 patients by machine learning, used 7 blood test items as the feature quantity, established an effective linear SVM prediction model for severe/non-severe disease (recall: 93.55%, specificity: 93.22%), and for 3 stages evaluation of the degree of severe level in severe patients was developed for patients with critical illness. The abnormal increase in Ferritin values was also found to be closely related to the development of severity.
{"title":"Predicting COVID-19 Severe Patients and Evaluation Method of 3 Stages Severe Level by Machine Learning","authors":"Jiahao Qu, Brian Sumali, Y. Mitsukura","doi":"10.1109/ICECE54449.2021.9674303","DOIUrl":"https://doi.org/10.1109/ICECE54449.2021.9674303","url":null,"abstract":"Since the outbreak of COVID-19 in Wuhan, China in December 2019, a large number of patients have been seen worldwide, and the number of infections continues to show an increasing trend. The vast majority of COVID-19 patients will have fever, headache, and mild respiratory symptoms, but a small number of severely ill patients will experience respiratory distress and related complications, which seriously endanger their lives. The large number of patients also puts the healthcare system to the test. To maximize the protection of patients’ lives and the effective use of medical resources, this study collected blood data from 313 patients by machine learning, used 7 blood test items as the feature quantity, established an effective linear SVM prediction model for severe/non-severe disease (recall: 93.55%, specificity: 93.22%), and for 3 stages evaluation of the degree of severe level in severe patients was developed for patients with critical illness. The abnormal increase in Ferritin values was also found to be closely related to the development of severity.","PeriodicalId":166178,"journal":{"name":"2021 IEEE 4th International Conference on Electronics and Communication Engineering (ICECE)","volume":"85 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114370930","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-12-17DOI: 10.1109/ICECE54449.2021.9674248
Ke-Sheng Huo, Zhuhua Hu, Dake Liu
Based on 3GPP standards, this paper investigates the interleaving and rate matching parts of turbo, convolutional, polar and LDPC codes used in 4G and 5G communication links. For the four different algorithms, this paper optimizes the address mapping formulas and proposes the interleaving storage sharing and module sharing methods, and the structure design of the shared modules. The feasibility of the algorithms is demonstrated through interleaving of codes and simulation verification, and a certain degree of hardware multiplexing of the 4G and 5G communication links is achieved.
{"title":"Rate Matching and Interleaved Hardware Sharing Design","authors":"Ke-Sheng Huo, Zhuhua Hu, Dake Liu","doi":"10.1109/ICECE54449.2021.9674248","DOIUrl":"https://doi.org/10.1109/ICECE54449.2021.9674248","url":null,"abstract":"Based on 3GPP standards, this paper investigates the interleaving and rate matching parts of turbo, convolutional, polar and LDPC codes used in 4G and 5G communication links. For the four different algorithms, this paper optimizes the address mapping formulas and proposes the interleaving storage sharing and module sharing methods, and the structure design of the shared modules. The feasibility of the algorithms is demonstrated through interleaving of codes and simulation verification, and a certain degree of hardware multiplexing of the 4G and 5G communication links is achieved.","PeriodicalId":166178,"journal":{"name":"2021 IEEE 4th International Conference on Electronics and Communication Engineering (ICECE)","volume":"45 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124204055","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-12-17DOI: 10.1109/ICECE54449.2021.9674326
Jue Wang, Xianfu Wan, Liqing Li, Jun Wang
Image segmentation is an effective method to extract the clothing region from the image, which is especially suitable for the analysis and processing of the clothing image with the complex background. At present, the research of image segmentation mainly focuses on the field of deep learning, and image segmentation methods such as DeepLab sequence based on convolutional neural network have been widely used. However, their segmentation results are not good enough when there are the complex deformation and edges in the clothing images. In order to improve the performance of clothing image segmentation, an improved DeepLab model for clothing image segmentation is developed in this paper. Based on the DeepLabV3+ model, the receptive field module and the decoder are redesigned in the new model. For the receptive field module, the ASPP (Atrous Spatial Pyramid Pooling) is changed to an improved RFBs (Receptive Field Block), which performs much better in simulating the human visual perception. For the decoder, the interpolation upsampling is replaced with a transpose convolution one due to it’s deformation adaptability to the edges and corners in the images; the concatenations between the high-level and the low-level features are increased from two-stage to five-stage in order to obtain more low-level features. After training and testing on deepfashion2 dataset, the improved model achieved performance of 97.26% Accuracy, 93.23% mIoU, 90.56% AP75 and 44.80% AP95 which is significantly better compared with DeepLabv3+. It takes 93.806 ms for the improved DeepLab model to complete the inference of one image, which is only slightly slower than that (92. 09S ms) for DeepLabV3+. The improved DeepLab model has a stronger ability to obtain information such as clothing edges, which makes the performance of segmentation better.
图像分割是从图像中提取服装区域的一种有效方法，特别适用于具有复杂背景的服装图像的分析和处理。目前，对图像分割的研究主要集中在深度学习领域，基于卷积神经网络的DeepLab序列等图像分割方法已得到广泛应用。然而，当服装图像中存在复杂的变形和边缘时，它们的分割效果不够好。为了提高服装图像分割的性能，本文提出了一种改进的DeepLab服装图像分割模型。在DeepLabV3+模型的基础上，对接收野模块和解码器进行了重新设计。在感受野模块中，将astrous空间金字塔池(ASPP)改为改进的RFBs (receptive field Block)，可以更好地模拟人类的视觉感受。解码器利用其对图像边角形变的适应性，将插值上采样替换为转置卷积上采样;为了获得更多的低级特征，将高级特征和低级特征之间的连接从两级增加到五级。经过在deepfashion2数据集上的训练和测试，改进后的模型准确率为97.26%，mIoU为93.23%，AP75为90.56%，AP95为44.80%，明显优于DeepLabv3+。改进的DeepLab模型完成一幅图像的推理需要93.806 ms，只比(92)稍慢。09S ms)的DeepLabV3+。改进后的DeepLab模型对服装边缘等信息的获取能力更强，使得分割的性能更好。
{"title":"An Improved DeepLab Model for Clothing Image Segmentation","authors":"Jue Wang, Xianfu Wan, Liqing Li, Jun Wang","doi":"10.1109/ICECE54449.2021.9674326","DOIUrl":"https://doi.org/10.1109/ICECE54449.2021.9674326","url":null,"abstract":"Image segmentation is an effective method to extract the clothing region from the image, which is especially suitable for the analysis and processing of the clothing image with the complex background. At present, the research of image segmentation mainly focuses on the field of deep learning, and image segmentation methods such as DeepLab sequence based on convolutional neural network have been widely used. However, their segmentation results are not good enough when there are the complex deformation and edges in the clothing images. In order to improve the performance of clothing image segmentation, an improved DeepLab model for clothing image segmentation is developed in this paper. Based on the DeepLabV3+ model, the receptive field module and the decoder are redesigned in the new model. For the receptive field module, the ASPP (Atrous Spatial Pyramid Pooling) is changed to an improved RFBs (Receptive Field Block), which performs much better in simulating the human visual perception. For the decoder, the interpolation upsampling is replaced with a transpose convolution one due to it’s deformation adaptability to the edges and corners in the images; the concatenations between the high-level and the low-level features are increased from two-stage to five-stage in order to obtain more low-level features. After training and testing on deepfashion2 dataset, the improved model achieved performance of 97.26% Accuracy, 93.23% mIoU, 90.56% AP75 and 44.80% AP95 which is significantly better compared with DeepLabv3+. It takes 93.806 ms for the improved DeepLab model to complete the inference of one image, which is only slightly slower than that (92. 09S ms) for DeepLabV3+. The improved DeepLab model has a stronger ability to obtain information such as clothing edges, which makes the performance of segmentation better.","PeriodicalId":166178,"journal":{"name":"2021 IEEE 4th International Conference on Electronics and Communication Engineering (ICECE)","volume":"168 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122565046","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-12-17DOI: 10.1109/ICECE54449.2021.9674524
Yibo Huang, Di Wu, Pengfei Wu
When a laser is transmitted in a foggy environment, the droplets floating in the air will attenuate the laser transmission. Based on Mie scattering theory, the Monte Carlo method and five common empirical models are used to analyze the transmission of a laser in a foggy environment. The transmission characteristics of the laser in a real foggy environment are measured. The atmospheric transmittance is compared with the predicted model results. The results show that the total energy attenuation of the droplet particles is approximately twice the attenuation of the scattering cross-section, and the laser attenuation in the advection fog is greater than that in the radiation fog under the same conditions. By comparing and analyzing the measured transmittance data in foggy weather and the values from the empirical prediction model, it is found that when the laser is transmitted in a foggy environment, each laser link has a corresponding empirical prediction model.
{"title":"Experimental Study of the Attenuation Effect of a Laser in a Foggy Environment in an FSO System","authors":"Yibo Huang, Di Wu, Pengfei Wu","doi":"10.1109/ICECE54449.2021.9674524","DOIUrl":"https://doi.org/10.1109/ICECE54449.2021.9674524","url":null,"abstract":"When a laser is transmitted in a foggy environment, the droplets floating in the air will attenuate the laser transmission. Based on Mie scattering theory, the Monte Carlo method and five common empirical models are used to analyze the transmission of a laser in a foggy environment. The transmission characteristics of the laser in a real foggy environment are measured. The atmospheric transmittance is compared with the predicted model results. The results show that the total energy attenuation of the droplet particles is approximately twice the attenuation of the scattering cross-section, and the laser attenuation in the advection fog is greater than that in the radiation fog under the same conditions. By comparing and analyzing the measured transmittance data in foggy weather and the values from the empirical prediction model, it is found that when the laser is transmitted in a foggy environment, each laser link has a corresponding empirical prediction model.","PeriodicalId":166178,"journal":{"name":"2021 IEEE 4th International Conference on Electronics and Communication Engineering (ICECE)","volume":"185 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131481508","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-12-17DOI: 10.1109/ICECE54449.2021.9674366
Rui Zhang, Li Gao
In order to improve the fusion accuracy of CT and MRI images, a new medical image fusion method with parameter-adaptive pulse-coupled neural network in nonsubsampled shearlet transform domain is proposed. In the proposed method, the NSST decomposition is first performed on the source images to obtain one low-frequency sub-band and a series of high-frequency sub-bands. Secondly, the high-frequency sub-bands are fused by a PAPCNN model, in which all the PCNN parameters can be adaptively estimated by the input sub-bands. The low-frequency sub-bands adopt an energy attribute-based fusion strategy, which is more conducive to preserving the complete basic information. Finally, the fused image is reconstructed by performing inverse NSST on the fused high-frequency and low-frequency sub-bands. The experimental results demonstrate that the fused images obtained in this paper have clear contours, high contrast, better preservation of detailed texture, and better results have been achieved in objective indicators such as average gradient, entropy, and peak signal-to-noise ratio.
{"title":"Medical Image Fusion Based on NonSubsampled Shearlet Transform and Parameter-Adaptive Pulse-Coupled Neural Network","authors":"Rui Zhang, Li Gao","doi":"10.1109/ICECE54449.2021.9674366","DOIUrl":"https://doi.org/10.1109/ICECE54449.2021.9674366","url":null,"abstract":"In order to improve the fusion accuracy of CT and MRI images, a new medical image fusion method with parameter-adaptive pulse-coupled neural network in nonsubsampled shearlet transform domain is proposed. In the proposed method, the NSST decomposition is first performed on the source images to obtain one low-frequency sub-band and a series of high-frequency sub-bands. Secondly, the high-frequency sub-bands are fused by a PAPCNN model, in which all the PCNN parameters can be adaptively estimated by the input sub-bands. The low-frequency sub-bands adopt an energy attribute-based fusion strategy, which is more conducive to preserving the complete basic information. Finally, the fused image is reconstructed by performing inverse NSST on the fused high-frequency and low-frequency sub-bands. The experimental results demonstrate that the fused images obtained in this paper have clear contours, high contrast, better preservation of detailed texture, and better results have been achieved in objective indicators such as average gradient, entropy, and peak signal-to-noise ratio.","PeriodicalId":166178,"journal":{"name":"2021 IEEE 4th International Conference on Electronics and Communication Engineering (ICECE)","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116910730","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-12-17DOI: 10.1109/ICECE54449.2021.9674707
Seungjun Hong, Seung-Jin Lee, Inhun Choi, E. Huh
As the innovation of computing infrastructure evolves to edge computing via cloud computing, intelligent devices such as robots, drones, and autonomous vehicles, which are mobile edge nodes, also surged. Since the edge nodes have limited resources, artificial intelligence services are provided based on lightweight containers. In addition, as intelligent edge node users increase and the categories of users become vast, in order to provide artificial intelligence services according to the situations of all users, data on each situation is collected, and it is necessary to continuously update the learning model. However, if the service is being provided, downtime is inevitable for the updated model to be applied to the service. Therefore, in this paper, we propose a micro-renewal method that minimizes the interruption of the service provided to users in real time when the learning model in the service is updated.
{"title":"Downtime Minimization for Real-time AI Service on Intelligent Edge Nodes: Micro-Renewal Method","authors":"Seungjun Hong, Seung-Jin Lee, Inhun Choi, E. Huh","doi":"10.1109/ICECE54449.2021.9674707","DOIUrl":"https://doi.org/10.1109/ICECE54449.2021.9674707","url":null,"abstract":"As the innovation of computing infrastructure evolves to edge computing via cloud computing, intelligent devices such as robots, drones, and autonomous vehicles, which are mobile edge nodes, also surged. Since the edge nodes have limited resources, artificial intelligence services are provided based on lightweight containers. In addition, as intelligent edge node users increase and the categories of users become vast, in order to provide artificial intelligence services according to the situations of all users, data on each situation is collected, and it is necessary to continuously update the learning model. However, if the service is being provided, downtime is inevitable for the updated model to be applied to the service. Therefore, in this paper, we propose a micro-renewal method that minimizes the interruption of the service provided to users in real time when the learning model in the service is updated.","PeriodicalId":166178,"journal":{"name":"2021 IEEE 4th International Conference on Electronics and Communication Engineering (ICECE)","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128461908","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-12-17DOI: 10.1109/ICECE54449.2021.9674574
Xiaoqing Xu, Hong Tang, Juan Wu, Liuyihui Qian, Han Zeng
The improvement of network performance needs to optimize network topology (that is adding network link) and expand link capacity. This paper comprehensively considers delay constraints, demand constraints and link capacity constraints and researches the joint optimization problem of network topology and link capacity under complex constraints. A new heuristic algorithm is proposed, called greedy-mutation genetic algorithm. Our method, based on a conventional genetic algorithm, conducts mutations on original solutions based on various constraints and the greedy algorithm, therefore, it can find better optimized solutions and fulfill all the constraints better. We applied the greedy-mutation genetic algorithm into two public backbone networks’ topology optimization and link expansion cases. Our results show that the proposed algorithm can effectively decrease the total cost of network optimization. The proposed method can also be applied in the network design and planning of wide area networks under complicated constraints, which is helpful to network operators.
{"title":"Joint Optimization of Network Topology and Link Capacity Expansion Based on a Greedy-Mutation Genetic Algorithm","authors":"Xiaoqing Xu, Hong Tang, Juan Wu, Liuyihui Qian, Han Zeng","doi":"10.1109/ICECE54449.2021.9674574","DOIUrl":"https://doi.org/10.1109/ICECE54449.2021.9674574","url":null,"abstract":"The improvement of network performance needs to optimize network topology (that is adding network link) and expand link capacity. This paper comprehensively considers delay constraints, demand constraints and link capacity constraints and researches the joint optimization problem of network topology and link capacity under complex constraints. A new heuristic algorithm is proposed, called greedy-mutation genetic algorithm. Our method, based on a conventional genetic algorithm, conducts mutations on original solutions based on various constraints and the greedy algorithm, therefore, it can find better optimized solutions and fulfill all the constraints better. We applied the greedy-mutation genetic algorithm into two public backbone networks’ topology optimization and link expansion cases. Our results show that the proposed algorithm can effectively decrease the total cost of network optimization. The proposed method can also be applied in the network design and planning of wide area networks under complicated constraints, which is helpful to network operators.","PeriodicalId":166178,"journal":{"name":"2021 IEEE 4th International Conference on Electronics and Communication Engineering (ICECE)","volume":"85 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116750706","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-12-17DOI: 10.1109/ICECE54449.2021.9674552
Kun Xu, Yan Liu
Code clone refers to a pair of semantically similar but syntactically similar or different code fragments that exist in code base. Excessive code clones in software system will cause a negative impact on system development and maintenance. In recent years, as deep learning has become a hot research area of machine learning, researchers have tried to apply deep learning techniques to code clone detection tasks. They have proposed a series of detection techniques using including unstructured (code in the form of sequential tokens) and structured (code in the form of abstract syntax trees and control-flow graphs) information to detect semantically similar but syntactically different code clone, which is the most difficult-to-detect clone type. However, although these methods have achieved an important improvement in the precision of semantic code clone detection, the corresponding false positive rate(FPR) is also at a very high level, making these methods unable to be effectively applied to real-world code bases. This paper proposed SCCD-GAN, an enhanced semantic code clone detection model which based on a graph representation form of programs and uses Graph Attention Network to measure the similarity of code pairs and achieved a lower detection FPR than existing methods. We built the graph representation of the code by expanding the control flow and data flow information to the original abstract syntax tree, and equipped with an attention mechanism to our model that focuses on the most important code parts and features which contribute much to the final detection precision.We implemented and evaluated our proposed method based on the benchmark dataset in the field of code clone detection-BigCloneBench2 and Google Code Jam. SCCD-GAN performed better than the existing state-of-the-art methods in terms of precision and false positive rate.
{"title":"SCCD-GAN: An Enhanced Semantic Code Clone Detection Model Using GAN","authors":"Kun Xu, Yan Liu","doi":"10.1109/ICECE54449.2021.9674552","DOIUrl":"https://doi.org/10.1109/ICECE54449.2021.9674552","url":null,"abstract":"Code clone refers to a pair of semantically similar but syntactically similar or different code fragments that exist in code base. Excessive code clones in software system will cause a negative impact on system development and maintenance. In recent years, as deep learning has become a hot research area of machine learning, researchers have tried to apply deep learning techniques to code clone detection tasks. They have proposed a series of detection techniques using including unstructured (code in the form of sequential tokens) and structured (code in the form of abstract syntax trees and control-flow graphs) information to detect semantically similar but syntactically different code clone, which is the most difficult-to-detect clone type. However, although these methods have achieved an important improvement in the precision of semantic code clone detection, the corresponding false positive rate(FPR) is also at a very high level, making these methods unable to be effectively applied to real-world code bases. This paper proposed SCCD-GAN, an enhanced semantic code clone detection model which based on a graph representation form of programs and uses Graph Attention Network to measure the similarity of code pairs and achieved a lower detection FPR than existing methods. We built the graph representation of the code by expanding the control flow and data flow information to the original abstract syntax tree, and equipped with an attention mechanism to our model that focuses on the most important code parts and features which contribute much to the final detection precision.We implemented and evaluated our proposed method based on the benchmark dataset in the field of code clone detection-BigCloneBench2 and Google Code Jam. SCCD-GAN performed better than the existing state-of-the-art methods in terms of precision and false positive rate.","PeriodicalId":166178,"journal":{"name":"2021 IEEE 4th International Conference on Electronics and Communication Engineering (ICECE)","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116761925","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In current computer communication networks, the increasing packet loss and delay caused by the increasing traffic becomes the bottleneck for the desired Quality of Service (QoS). Weighted Fair Queueing can be used to provide differentiated services according to the Service Level Agreement (SLA) associated with each packet. However, due to inaccurate measurements of queue usage, drop rate and delay in real routers, and the intrinsic property of a real network system that there will always be some unpredictable traffic patterns, current methods for WFQ updating can be improved and extended further. In this work, an uncertainty-aware soft actor-critic agent is introduced. First, the learned weights updating strategy is a maximum entropy policy, which is robust under estimation and model error. Second, the technique of model uncertainty estimation is adopted into the agent so that it is capable of detecting novel states that are unseen during the training period, which facilitates a strategy switching framework. The proposed algorithm shows the potential of using reinforcement learning for WFQ weights updating and is compatible with existing techniques by monitoring the model uncertainty, which makes a more robust and stable system. The benefits of applying the proposed algorithm is validated through the simulation studies, showing a promising direction for further exploration.
{"title":"Uncertainty-aware Weighted Fair Queueing for Routers Based on Deep Reinforcement Learning","authors":"Pengyue Wang, Zhaoyu Jiang, Meiyu Qi, Longfei Dai, Huiying Xu","doi":"10.1109/ICECE54449.2021.9674580","DOIUrl":"https://doi.org/10.1109/ICECE54449.2021.9674580","url":null,"abstract":"In current computer communication networks, the increasing packet loss and delay caused by the increasing traffic becomes the bottleneck for the desired Quality of Service (QoS). Weighted Fair Queueing can be used to provide differentiated services according to the Service Level Agreement (SLA) associated with each packet. However, due to inaccurate measurements of queue usage, drop rate and delay in real routers, and the intrinsic property of a real network system that there will always be some unpredictable traffic patterns, current methods for WFQ updating can be improved and extended further. In this work, an uncertainty-aware soft actor-critic agent is introduced. First, the learned weights updating strategy is a maximum entropy policy, which is robust under estimation and model error. Second, the technique of model uncertainty estimation is adopted into the agent so that it is capable of detecting novel states that are unseen during the training period, which facilitates a strategy switching framework. The proposed algorithm shows the potential of using reinforcement learning for WFQ weights updating and is compatible with existing techniques by monitoring the model uncertainty, which makes a more robust and stable system. The benefits of applying the proposed algorithm is validated through the simulation studies, showing a promising direction for further exploration.","PeriodicalId":166178,"journal":{"name":"2021 IEEE 4th International Conference on Electronics and Communication Engineering (ICECE)","volume":"34 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132956350","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","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：