Pub Date : 2023-07-01DOI: 10.1109/CSCloud-EdgeCom58631.2023.00021
Zhaohui Ma, Jie Zhang, Mingdong Tang
Software Defined Network (SDN) is a new type of network architecture that realizes network virtualization, with the characteristics of the control and forwarding separation, open programming, centralized control, and its flexibility is more suitable for the current complex and changeable network environment. However, due to its centralized control characteristics, the controller is faced with a huge risk of being subjected to distributed denial of service (DDoS) attacks that will cause the entire network to be paralyzed. Therefore, the detection of DDoS attacks in SDN networks has become the research direction of many scholars. so an algorithm for detecting DDoS attacks in SDN networks using optimizing RFs is proposed. By selecting the appropriate traffic features, creating the traffic dataset in the SDN environment, and using the dataset to optimize the model parameters, the attack detection model is constructed, and the final detection algorithm is as accurate as 99.98% for the collected dataset, which is more accurate and efficient than the common machine learning algorithms such as SVC and KNN.
软件定义网络(Software Defined Network, SDN)是一种实现网络虚拟化的新型网络架构，具有控制与转发分离、开放编程、集中控制等特点，其灵活性更适合当前复杂多变的网络环境。然而，由于其集中控制的特点，控制器面临着遭受分布式拒绝服务攻击的巨大风险，这将导致整个网络瘫痪。因此，SDN网络中DDoS攻击的检测成为众多学者的研究方向。为此，提出了一种基于优化RFs的SDN网络DDoS攻击检测算法。通过选择合适的流量特征，在SDN环境下创建流量数据集，并利用数据集对模型参数进行优化，构建攻击检测模型，最终对采集到的数据集检测算法准确率高达99.98%，比常用的SVC、KNN等机器学习算法准确率更高、效率更高。
{"title":"Optimized Random Forest for DDoS Attack Detection in SDN Environment","authors":"Zhaohui Ma, Jie Zhang, Mingdong Tang","doi":"10.1109/CSCloud-EdgeCom58631.2023.00021","DOIUrl":"https://doi.org/10.1109/CSCloud-EdgeCom58631.2023.00021","url":null,"abstract":"Software Defined Network (SDN) is a new type of network architecture that realizes network virtualization, with the characteristics of the control and forwarding separation, open programming, centralized control, and its flexibility is more suitable for the current complex and changeable network environment. However, due to its centralized control characteristics, the controller is faced with a huge risk of being subjected to distributed denial of service (DDoS) attacks that will cause the entire network to be paralyzed. Therefore, the detection of DDoS attacks in SDN networks has become the research direction of many scholars. so an algorithm for detecting DDoS attacks in SDN networks using optimizing RFs is proposed. By selecting the appropriate traffic features, creating the traffic dataset in the SDN environment, and using the dataset to optimize the model parameters, the attack detection model is constructed, and the final detection algorithm is as accurate as 99.98% for the collected dataset, which is more accurate and efficient than the common machine learning algorithms such as SVC and KNN.","PeriodicalId":56007,"journal":{"name":"Journal of Cloud Computing-Advances Systems and Applications","volume":"14 1","pages":"72-77"},"PeriodicalIF":4.0,"publicationDate":"2023-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87782904","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-07-01DOI: 10.1109/CSCloud-EdgeCom58631.2023.00036
Jing Long, Cuiting Luo, Ruxin Chen
The ubiquitous use of real-time sensors in the Internet-of-Things (IoT) has brought great convenience to data collection. Moreover, sensor anomalies generated by external factors or malicious attacks pose a critical threat to the security of the IoT. Detecting anomalies in multivariate time series has become one of the significant issues in Io T security research. Most existing time series anomaly detection methods, however, merely consider time and space complexity, without taking into account the distance metrics among time series data, which leads inevitably to the model’s insufficient ability to accurately recognize anomalies. This investigation proposes a new hybrid model based on encoder-decoder architecture for time series anomaly detection. This model designs a multi-dimensional feature embedding module to enable utilize more interrelated features. Meanwhile, the relationships between sensors are explicitly learned by using a graph structure and reconstruct the nodes vectors by using a message propagation mechanism with a specific sampling strategy in this model. On this basis, a data fusion method based on the multi-head self-attention mechanism which effectively integrates various information such as time, variables, positional relationships, and distance metrics is designed for capturing global feature information. The experimental results on the dataset SWAT show that, compared with the state-of-the-arts, the proposed approach improves the Recall and F1-score metrics of anomaly detection performance by 8.2% and 5.0% respectively.
{"title":"Multivariate Time Series Anomaly Detection with Improved Encoder-Decoder Based Model","authors":"Jing Long, Cuiting Luo, Ruxin Chen","doi":"10.1109/CSCloud-EdgeCom58631.2023.00036","DOIUrl":"https://doi.org/10.1109/CSCloud-EdgeCom58631.2023.00036","url":null,"abstract":"The ubiquitous use of real-time sensors in the Internet-of-Things (IoT) has brought great convenience to data collection. Moreover, sensor anomalies generated by external factors or malicious attacks pose a critical threat to the security of the IoT. Detecting anomalies in multivariate time series has become one of the significant issues in Io T security research. Most existing time series anomaly detection methods, however, merely consider time and space complexity, without taking into account the distance metrics among time series data, which leads inevitably to the model’s insufficient ability to accurately recognize anomalies. This investigation proposes a new hybrid model based on encoder-decoder architecture for time series anomaly detection. This model designs a multi-dimensional feature embedding module to enable utilize more interrelated features. Meanwhile, the relationships between sensors are explicitly learned by using a graph structure and reconstruct the nodes vectors by using a message propagation mechanism with a specific sampling strategy in this model. On this basis, a data fusion method based on the multi-head self-attention mechanism which effectively integrates various information such as time, variables, positional relationships, and distance metrics is designed for capturing global feature information. The experimental results on the dataset SWAT show that, compared with the state-of-the-arts, the proposed approach improves the Recall and F1-score metrics of anomaly detection performance by 8.2% and 5.0% respectively.","PeriodicalId":56007,"journal":{"name":"Journal of Cloud Computing-Advances Systems and Applications","volume":"4 1","pages":"161-166"},"PeriodicalIF":4.0,"publicationDate":"2023-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75587727","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-07-01DOI: 10.1109/cscloud-edgecom58631.2023.00033
Xiai Yan, Shengkai Ding, Weiqi Shi
To solve the problem of pedestrian scale change, overlap, and occlusion under different cameras, we use the YOLOv5-based pedestrian detection algorithm which is more insensitive to target size, more real-time, and better on the ground. We also use MobileNetV3 as the backbone of YOLOv5 for the problems of weaker robustness and heavier network of YOLOv5. To improve the performance of the pedestrian detection algorithm, we also use a larger and more accurate home-grown dataset, SmartJW, to train and test the algorithm. This paper collects a dataset containing pedestrian images in various real-world scenarios through a dedicated video network in a region of Hunan. The dataset includes images of pedestrians at different times of day, in different weather, locations, and lighting conditions. In addition, to have faster detection speed and higher detection accuracy, we use an efficient and accurate target detection algorithm, YOLOv5. We use YOLOv5’s small network as the base framework of our pedestrian detection algorithm, based on which we make improvements to YOLOv5’s Backbone to obtain the MobileNetV3-YOLOv5 network, replacing the CSPNet with the MobileNetV3-large network. And finally trained and tested on our homemade dataset SmartJW. The results showed that we reached 0.983 for the mAP@0.5 and 0.728 for the mAP@0.5:0.95.
{"title":"MobileNetV3-YOLOv5-based Network Model for Pedestrian Detection","authors":"Xiai Yan, Shengkai Ding, Weiqi Shi","doi":"10.1109/cscloud-edgecom58631.2023.00033","DOIUrl":"https://doi.org/10.1109/cscloud-edgecom58631.2023.00033","url":null,"abstract":"To solve the problem of pedestrian scale change, overlap, and occlusion under different cameras, we use the YOLOv5-based pedestrian detection algorithm which is more insensitive to target size, more real-time, and better on the ground. We also use MobileNetV3 as the backbone of YOLOv5 for the problems of weaker robustness and heavier network of YOLOv5. To improve the performance of the pedestrian detection algorithm, we also use a larger and more accurate home-grown dataset, SmartJW, to train and test the algorithm. This paper collects a dataset containing pedestrian images in various real-world scenarios through a dedicated video network in a region of Hunan. The dataset includes images of pedestrians at different times of day, in different weather, locations, and lighting conditions. In addition, to have faster detection speed and higher detection accuracy, we use an efficient and accurate target detection algorithm, YOLOv5. We use YOLOv5’s small network as the base framework of our pedestrian detection algorithm, based on which we make improvements to YOLOv5’s Backbone to obtain the MobileNetV3-YOLOv5 network, replacing the CSPNet with the MobileNetV3-large network. And finally trained and tested on our homemade dataset SmartJW. The results showed that we reached 0.983 for the mAP@0.5 and 0.728 for the mAP@0.5:0.95.","PeriodicalId":56007,"journal":{"name":"Journal of Cloud Computing-Advances Systems and Applications","volume":"31 1","pages":"144-149"},"PeriodicalIF":4.0,"publicationDate":"2023-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"72559031","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-07-01DOI: 10.1109/CSCloud-EdgeCom58631.2023.00056
Yin'e Zhang, Xiaowen Ye, Qi Zhou
In recent years, more and more researchers use deep learning to process inpainting tasks. Among them, the use of generation countermeasure network to process inpainting tasks has become more and more popular and has achieved good results. However, there are still issues with blurry repair results and unsmooth structure. In this paper, we propose a method of inpainting based on u-net structure for generation adversarial network, the first two layers of our encoder use multi-scale shallow feature extraction modules (MSFEM) to extract lowdimensional texture and structural information. We introduce multi-scale spatial attention module (MSAM) into skip connections to obtain more shallow features and improve repair performance. The decoder uses improved dense convolutional blocks to fully utilize and extract feature information. The experiment used two datasets, CelebA and Palace2, through experiments, the repair effect of our proposed method is better than the state-of-the-art image inpainting approaches.
{"title":"Research on Application of Generative Adversarial Neural Network in Image Restoration","authors":"Yin'e Zhang, Xiaowen Ye, Qi Zhou","doi":"10.1109/CSCloud-EdgeCom58631.2023.00056","DOIUrl":"https://doi.org/10.1109/CSCloud-EdgeCom58631.2023.00056","url":null,"abstract":"In recent years, more and more researchers use deep learning to process inpainting tasks. Among them, the use of generation countermeasure network to process inpainting tasks has become more and more popular and has achieved good results. However, there are still issues with blurry repair results and unsmooth structure. In this paper, we propose a method of inpainting based on u-net structure for generation adversarial network, the first two layers of our encoder use multi-scale shallow feature extraction modules (MSFEM) to extract lowdimensional texture and structural information. We introduce multi-scale spatial attention module (MSAM) into skip connections to obtain more shallow features and improve repair performance. The decoder uses improved dense convolutional blocks to fully utilize and extract feature information. The experiment used two datasets, CelebA and Palace2, through experiments, the repair effect of our proposed method is better than the state-of-the-art image inpainting approaches.","PeriodicalId":56007,"journal":{"name":"Journal of Cloud Computing-Advances Systems and Applications","volume":"79 1","pages":"287-291"},"PeriodicalIF":4.0,"publicationDate":"2023-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"77663978","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-07-01DOI: 10.1109/CSCloud-EdgeCom58631.2023.00011
Yuchen Nie, Xiangling Ding, Wenyi Zhu, Yulin Zhao
The widespread use of image editing software and the development of image processing technology have made manipulated images very easy. Image forgery, especially in remote sensing satellite images, may bring incalculable and serious consequences, which makes researchers focus on how to verify the integrity of remote sensing images. This paper introduces remote sensing images and manipulating operations in remote sensing images, mainly splicing operations. After that, we discuss methods in depth for detecting and locating manipulation in remote sensing images in recent years. The generative model reflects its unparalleled advantages in these methods. Finally, the future development direction is prospected.
{"title":"A Systematic Review on Detection of Manipulated Satellite Images","authors":"Yuchen Nie, Xiangling Ding, Wenyi Zhu, Yulin Zhao","doi":"10.1109/CSCloud-EdgeCom58631.2023.00011","DOIUrl":"https://doi.org/10.1109/CSCloud-EdgeCom58631.2023.00011","url":null,"abstract":"The widespread use of image editing software and the development of image processing technology have made manipulated images very easy. Image forgery, especially in remote sensing satellite images, may bring incalculable and serious consequences, which makes researchers focus on how to verify the integrity of remote sensing images. This paper introduces remote sensing images and manipulating operations in remote sensing images, mainly splicing operations. After that, we discuss methods in depth for detecting and locating manipulation in remote sensing images in recent years. The generative model reflects its unparalleled advantages in these methods. Finally, the future development direction is prospected.","PeriodicalId":56007,"journal":{"name":"Journal of Cloud Computing-Advances Systems and Applications","volume":"18 1","pages":"6-11"},"PeriodicalIF":4.0,"publicationDate":"2023-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75340339","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-07-01DOI: 10.1109/CSCloud-EdgeCom58631.2023.00017
Marc Titus Trifan, Alexandru Nicolau, A. Veidenbaum
The increased popularity of Machine Learning as a Service (MLaaS) makes the privacy of user data and network weights a critical concern. Using Torus FHE (TFHE) [1] offers a solution for privacy-preserving computation in a cloud environment by allowing computation directly over encrypted data. However, software TFHE implementations of cyphertext-cyphertext multiplication needed when both input data and weights are encrypted are either lacking or are too slow. This paper proposes a new way to improve the performance of such multiplication by applying carry save addition. Its theoretical speedup is proportional to the bit width of the plaintext integer operands. This also speeds up multi-operand summation. A speedup of 15x is obtained for 16-bit multiplication on a 64-core processor, when compared to previous results. Multiplication also becomes more than twice as fast on a GPU if our approach is utilized. This leads to much faster dot product and convolution computations, which combine multiplications and a multi-operand sum. A 45x speedup is achieved for a 16-bit, 32element dot product and $mathrm{a}sim 30mathrm{x}$ speedup for a convolution with a 32x32 filter size.
{"title":"Enhancing the Privacy of Machine Learning via faster arithmetic over Torus FHE","authors":"Marc Titus Trifan, Alexandru Nicolau, A. Veidenbaum","doi":"10.1109/CSCloud-EdgeCom58631.2023.00017","DOIUrl":"https://doi.org/10.1109/CSCloud-EdgeCom58631.2023.00017","url":null,"abstract":"The increased popularity of Machine Learning as a Service (MLaaS) makes the privacy of user data and network weights a critical concern. Using Torus FHE (TFHE) [1] offers a solution for privacy-preserving computation in a cloud environment by allowing computation directly over encrypted data. However, software TFHE implementations of cyphertext-cyphertext multiplication needed when both input data and weights are encrypted are either lacking or are too slow. This paper proposes a new way to improve the performance of such multiplication by applying carry save addition. Its theoretical speedup is proportional to the bit width of the plaintext integer operands. This also speeds up multi-operand summation. A speedup of 15x is obtained for 16-bit multiplication on a 64-core processor, when compared to previous results. Multiplication also becomes more than twice as fast on a GPU if our approach is utilized. This leads to much faster dot product and convolution computations, which combine multiplications and a multi-operand sum. A 45x speedup is achieved for a 16-bit, 32element dot product and $mathrm{a}sim 30mathrm{x}$ speedup for a convolution with a 32x32 filter size.","PeriodicalId":56007,"journal":{"name":"Journal of Cloud Computing-Advances Systems and Applications","volume":"39 1","pages":"46-52"},"PeriodicalIF":4.0,"publicationDate":"2023-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"78815795","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-07-01DOI: 10.1109/CSCloud-EdgeCom58631.2023.00041
Zuoting Ning, Zihua Ouyang, Xiangcheng Deng
Apriori algorithm has the advantages of simplicity and easy realization, which can realize the search for some association between data. However, it also has unavoidable flaws. The traditional serial Apriori algorithm has the disadvantages of frequent scanning database, generating large amount of candidate item-set data and consuming a lot of memory resources. This paper proposes some technologies based on association principle, and proposes an improved Apriori algorithm to overcome the shortcomings of Apriori algorithm. When dealing with transactions, the algorithm filters out transactions that cannot generate frequent item-set, which can greatly reduce the amount of data processing. Theoretical analysis and experimental results show that the proposed scheme has the best performance in efficiency and stability.
{"title":"An Improved Apriori Algorithm Based on Transaction Sequence Counting","authors":"Zuoting Ning, Zihua Ouyang, Xiangcheng Deng","doi":"10.1109/CSCloud-EdgeCom58631.2023.00041","DOIUrl":"https://doi.org/10.1109/CSCloud-EdgeCom58631.2023.00041","url":null,"abstract":"Apriori algorithm has the advantages of simplicity and easy realization, which can realize the search for some association between data. However, it also has unavoidable flaws. The traditional serial Apriori algorithm has the disadvantages of frequent scanning database, generating large amount of candidate item-set data and consuming a lot of memory resources. This paper proposes some technologies based on association principle, and proposes an improved Apriori algorithm to overcome the shortcomings of Apriori algorithm. When dealing with transactions, the algorithm filters out transactions that cannot generate frequent item-set, which can greatly reduce the amount of data processing. Theoretical analysis and experimental results show that the proposed scheme has the best performance in efficiency and stability.","PeriodicalId":56007,"journal":{"name":"Journal of Cloud Computing-Advances Systems and Applications","volume":"269 1","pages":"192-196"},"PeriodicalIF":4.0,"publicationDate":"2023-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79865056","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Wireless networks are vulnerable to various network attacks due to easy access to the nodes. The development of technologies for network intrusion detection, including those based on deep learning, is expected to bring ultimate solutions to this problem. Nevertheless, existing intrusion detection models based on deep learning have low detection accuracy and cannot effectively detect several new types of attacks. Aimed at such, this article proposes IFLV, an intrusion detection model for wireless networks, by integrating Fully Convolutional Network (FCN), Long Short-Term Memory (LSTM), and Vision Transformer (ViT). IFLV can extract the local and global features of traffic data and learn its temporal and spatial features, to improve the accuracy of network traffic classification. Based on the improvements of the traditional ViT model to overcome the poor classification effect in small and medium-sized datasets, IFLV can achieve expressive results even with fewer training resources. Experimental results show that IFLV has a high accuracy of network traffic intrusion detection with an accuracy of 99.973% in the AWID dataset and significantly superior performance compared to existing models.
{"title":"IFLV: Wireless network intrusion detection model integrating FCN, LSTM, and ViT","authors":"Wenmin Zeng, Dezhi Han, Mingming Cui, Zhongdai Wu, Bing Han, Hongxu Zhou","doi":"10.1109/CSCloud-EdgeCom58631.2023.00086","DOIUrl":"https://doi.org/10.1109/CSCloud-EdgeCom58631.2023.00086","url":null,"abstract":"Wireless networks are vulnerable to various network attacks due to easy access to the nodes. The development of technologies for network intrusion detection, including those based on deep learning, is expected to bring ultimate solutions to this problem. Nevertheless, existing intrusion detection models based on deep learning have low detection accuracy and cannot effectively detect several new types of attacks. Aimed at such, this article proposes IFLV, an intrusion detection model for wireless networks, by integrating Fully Convolutional Network (FCN), Long Short-Term Memory (LSTM), and Vision Transformer (ViT). IFLV can extract the local and global features of traffic data and learn its temporal and spatial features, to improve the accuracy of network traffic classification. Based on the improvements of the traditional ViT model to overcome the poor classification effect in small and medium-sized datasets, IFLV can achieve expressive results even with fewer training resources. Experimental results show that IFLV has a high accuracy of network traffic intrusion detection with an accuracy of 99.973% in the AWID dataset and significantly superior performance compared to existing models.","PeriodicalId":56007,"journal":{"name":"Journal of Cloud Computing-Advances Systems and Applications","volume":"145 1","pages":"470-475"},"PeriodicalIF":4.0,"publicationDate":"2023-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"77320613","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-07-01DOI: 10.1109/CSCloud-EdgeCom58631.2023.00022
Zhixue Li, Shiwen Zhang, N. Xiong, Wei Liang
In recent years, as a novel perceptual paradigm, Mobile Crowd Sensing (MCS) has gradually become one of the most popular research contents. It utilizes mobile devices carried by users to collect various sensing data about social events and phenomena. To improve the credibility of the data, it is critical to recruit mobile users, but it leads to the privacy leakage of mobile users. Therefore, how to achieve efficient task allocation while protecting user data privacy is a challenging problem in MCS. In this paper, we propose an efficient and secure task allocation scheme (ESTA). In ESTA, the service provider enables to forecast the spatial distribution of sensing users and select high quality sensing data according to their trust levels without invading user privacy. By utilizing the advantage of federated learning (FL) that does not centrally collect the user data to prevent privacy leakage. Finally, we show the security properties of ESTA and demonstrate its efficiency in terms of task finished ratio and task allocation ratio.
{"title":"Achieving Efficient and Secure Task Allocation Scheme in Mobile Crowd Sensing","authors":"Zhixue Li, Shiwen Zhang, N. Xiong, Wei Liang","doi":"10.1109/CSCloud-EdgeCom58631.2023.00022","DOIUrl":"https://doi.org/10.1109/CSCloud-EdgeCom58631.2023.00022","url":null,"abstract":"In recent years, as a novel perceptual paradigm, Mobile Crowd Sensing (MCS) has gradually become one of the most popular research contents. It utilizes mobile devices carried by users to collect various sensing data about social events and phenomena. To improve the credibility of the data, it is critical to recruit mobile users, but it leads to the privacy leakage of mobile users. Therefore, how to achieve efficient task allocation while protecting user data privacy is a challenging problem in MCS. In this paper, we propose an efficient and secure task allocation scheme (ESTA). In ESTA, the service provider enables to forecast the spatial distribution of sensing users and select high quality sensing data according to their trust levels without invading user privacy. By utilizing the advantage of federated learning (FL) that does not centrally collect the user data to prevent privacy leakage. Finally, we show the security properties of ESTA and demonstrate its efficiency in terms of task finished ratio and task allocation ratio.","PeriodicalId":56007,"journal":{"name":"Journal of Cloud Computing-Advances Systems and Applications","volume":"11 1","pages":"78-84"},"PeriodicalIF":4.0,"publicationDate":"2023-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89071879","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The automotive industry has always been seeking innovative solutions to improve car performance, safety, and cost savings. Lightweight design technology has become one of the solutions. This article summarizes the modeling and optimization methods of automotive lightweight design, as well as key technologies based on intelligent optimization learning. First, this article outlines the basic concepts of automotive lightweight design, as well as the needs and challenges of the industry for lightweight design. Then, the modeling methods of lightweight design are introduced in detail, including geometric modeling, topology optimization, structural optimization, and multidisciplinary optimization. At the same time, commonly used materials, manufacturing processes, and testing methods in lightweight design are introduced, as well as relevant design guidelines and standards. This article also introduces some algorithms and their applicable scenarios. Additionally, this article summarizes the application prospects and future development directions of key technologies for automotive lightweight design modeling and intelligent optimization learning. We emphasize the opportunities and challenges in this field and propose how to continue promoting the development of lightweight design technology and responding to increasingly complex market demands. This article provides a systematic review of key technologies for automotive lightweight design modeling and intelligent optimization learning, which helps researchers and practitioners to deepen their understanding of the technical development and application trends in this field.
{"title":"Automotive Lightweight Design Modeling And Intelligent Optimization Learn Key Technologies","authors":"Gejing Xu, Wei Liang, Jiahong Cai, Jiahong Xiao, Xingyu Chen, Yinyan Gong","doi":"10.1109/CSCloud-EdgeCom58631.2023.00071","DOIUrl":"https://doi.org/10.1109/CSCloud-EdgeCom58631.2023.00071","url":null,"abstract":"The automotive industry has always been seeking innovative solutions to improve car performance, safety, and cost savings. Lightweight design technology has become one of the solutions. This article summarizes the modeling and optimization methods of automotive lightweight design, as well as key technologies based on intelligent optimization learning. First, this article outlines the basic concepts of automotive lightweight design, as well as the needs and challenges of the industry for lightweight design. Then, the modeling methods of lightweight design are introduced in detail, including geometric modeling, topology optimization, structural optimization, and multidisciplinary optimization. At the same time, commonly used materials, manufacturing processes, and testing methods in lightweight design are introduced, as well as relevant design guidelines and standards. This article also introduces some algorithms and their applicable scenarios. Additionally, this article summarizes the application prospects and future development directions of key technologies for automotive lightweight design modeling and intelligent optimization learning. We emphasize the opportunities and challenges in this field and propose how to continue promoting the development of lightweight design technology and responding to increasingly complex market demands. This article provides a systematic review of key technologies for automotive lightweight design modeling and intelligent optimization learning, which helps researchers and practitioners to deepen their understanding of the technical development and application trends in this field.","PeriodicalId":56007,"journal":{"name":"Journal of Cloud Computing-Advances Systems and Applications","volume":"7 1","pages":"381-386"},"PeriodicalIF":4.0,"publicationDate":"2023-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82849971","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","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：