Pub Date : 2025-07-10DOI: 10.1109/TCC.2025.3586878
Meng Tian;Zhicheng Liu;Chenxuan Hou;Chao Qiu;Xiaofei Wang;Dusit Niyato;Victor C. M. Leung
While diffusion models have demonstrated remarkable success in computer vision tasks, their deployment in Internet of Things environments remains challenging. Edge devices face significant constraints in computational resources and must adapt to dynamic operating conditions. To address these limitations, we propose a novel system that accelerates AI-generated content (AIGC) collaborative inference in dynamic edge networks. The proposed system introduces a multi-exit vision transformer-based U-Net architecture that enables efficient processing through adaptive exit point selection during the diffusion process, optimizing the trade-off between inference accuracy and computational efficiency. To optimize device-level operations, we develop an innovative generative AI-assisted reinforcement learning framework that determines optimal exit selection and offloading strategies to maximize generation quality and inference speed. Furthermore, we design a fine-tuning approach with policy reuse mechanisms that facilitates rapid reinforcement learning algorithm deployment across diverse environments. Extensive experimental evaluations demonstrate that our system outperforms existing algorithms in terms of balancing inference latency and generation quality, while also exhibiting improved adaptability to environmental variations.
{"title":"Accelerating AI-Generated Content Collaborative Inference Via Transfer Reinforcement Learning in Dynamic Edge Networks","authors":"Meng Tian;Zhicheng Liu;Chenxuan Hou;Chao Qiu;Xiaofei Wang;Dusit Niyato;Victor C. M. Leung","doi":"10.1109/TCC.2025.3586878","DOIUrl":"https://doi.org/10.1109/TCC.2025.3586878","url":null,"abstract":"While diffusion models have demonstrated remarkable success in computer vision tasks, their deployment in Internet of Things environments remains challenging. Edge devices face significant constraints in computational resources and must adapt to dynamic operating conditions. To address these limitations, we propose a novel system that accelerates AI-generated content (AIGC) collaborative inference in dynamic edge networks. The proposed system introduces a multi-exit vision transformer-based U-Net architecture that enables efficient processing through adaptive exit point selection during the diffusion process, optimizing the trade-off between inference accuracy and computational efficiency. To optimize device-level operations, we develop an innovative generative AI-assisted reinforcement learning framework that determines optimal exit selection and offloading strategies to maximize generation quality and inference speed. Furthermore, we design a fine-tuning approach with policy reuse mechanisms that facilitates rapid reinforcement learning algorithm deployment across diverse environments. Extensive experimental evaluations demonstrate that our system outperforms existing algorithms in terms of balancing inference latency and generation quality, while also exhibiting improved adaptability to environmental variations.","PeriodicalId":13202,"journal":{"name":"IEEE Transactions on Cloud Computing","volume":"13 3","pages":"1011-1025"},"PeriodicalIF":5.0,"publicationDate":"2025-07-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144997132","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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 proliferation of Internet of Things (IoT) devices has led to a surge in data generation within smart home environments. This data explosion has raised significant privacy concerns and highlighted a lack of user-friendly controls. Consequently, there is a pressing need for a robust privacy-enhancing mechanism tailored for smart homes, safeguarding sensitive data from a user-centric perspective. In this article, we introduce the Fog-enhanced Personalized Differential Privacy (FEPDP) model, which utilizes the distributed nature of fog computing to improve data processing efficiency and security in smart homes. Specifically, the personalization, as a key feature of FEPDP, is manifested through an array of user-driven policy specifications, enabling home users to specify secret and privacy specifications for their personal data. These specifications not only enhance control over personal data but also align with the heterogeneous nature of smart home environments. Subsequently, aligned with fog-based smart home architecture, we propose two policy-driven partitioning mechanisms that utilize threshold partitioning based on dynamic programming to effectively implement FEPDP. Finally, comprehensive theoretical analysis and experimental validation across various statistical analysis tasks and datasets confirm that FEPDP achieves a superior privacy-utility trade-off for smart home data by leveraging non-sensitive data and fog-based partitioning.
{"title":"Fog-Enhanced Personalized Privacy-Preserving Data Analysis for Smart Homes","authors":"Jiajun Chen;Chunqiang Hu;Weihong Sheng;Hui Xia;Pengfei Hu;Jiguo Yu","doi":"10.1109/TCC.2025.3586052","DOIUrl":"https://doi.org/10.1109/TCC.2025.3586052","url":null,"abstract":"The proliferation of Internet of Things (IoT) devices has led to a surge in data generation within smart home environments. This data explosion has raised significant privacy concerns and highlighted a lack of user-friendly controls. Consequently, there is a pressing need for a robust privacy-enhancing mechanism tailored for smart homes, safeguarding sensitive data from a user-centric perspective. In this article, we introduce the Fog-enhanced Personalized Differential Privacy (FEPDP) model, which utilizes the distributed nature of fog computing to improve data processing efficiency and security in smart homes. Specifically, the personalization, as a key feature of FEPDP, is manifested through an array of user-driven policy specifications, enabling home users to specify secret and privacy specifications for their personal data. These specifications not only enhance control over personal data but also align with the heterogeneous nature of smart home environments. Subsequently, aligned with fog-based smart home architecture, we propose two policy-driven partitioning mechanisms that utilize threshold partitioning based on dynamic programming to effectively implement FEPDP. Finally, comprehensive theoretical analysis and experimental validation across various statistical analysis tasks and datasets confirm that FEPDP achieves a superior privacy-utility trade-off for smart home data by leveraging non-sensitive data and fog-based partitioning.","PeriodicalId":13202,"journal":{"name":"IEEE Transactions on Cloud Computing","volume":"13 3","pages":"995-1010"},"PeriodicalIF":5.0,"publicationDate":"2025-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144998331","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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 : 2025-06-24DOI: 10.1109/TCC.2025.3582645
Boli Hu;Kai Zhang;Junqing Gong;Haifeng Qian
Outsourcing personal data to a convenient and affordable cloud platform has become a popular practice. Considering the risk of privacy leakage, users usually encrypt their data before uploading it to the cloud server. Searchable encryption (SE) allows cloud servers to manage and search data in encrypted form based on user-specified requests. However, coercion attacks are rarely considered, where users may be forced to open search records and results. Therefore, deniable SE solutions against coercion attacks are presented, but they suffer from large storage overhead or fail to consider the dual coercion situation towards both sides of data owners and data users. In this paper, we roughly combine oblivious cross-tags protocol (OXT) and deniable encryption to propose a deniable SE (deniable cross-tag, DXT) scheme, which supports boolean queries and resists dual coercion attacks. Technically, we formalize a new primitive called updatable deniable encryption, and combine it with OXT in a non-trivial manner. In addition, we give formal system model, security model, and security proof of DXT. By employing the HUAWEI cloud platform, we conduct sufficient comparative experiments between DXT and state-of-the-art solutions based on a public dataset. The experimental results demonstrate that DXT outperforms higher search efficiency while achieving better features.
{"title":"Refrain From Inquiring About My Scalable Storage and Boolean Queries for Secure Cloud","authors":"Boli Hu;Kai Zhang;Junqing Gong;Haifeng Qian","doi":"10.1109/TCC.2025.3582645","DOIUrl":"https://doi.org/10.1109/TCC.2025.3582645","url":null,"abstract":"Outsourcing personal data to a convenient and affordable cloud platform has become a popular practice. Considering the risk of privacy leakage, users usually encrypt their data before uploading it to the cloud server. Searchable encryption (SE) allows cloud servers to manage and search data in encrypted form based on user-specified requests. However, coercion attacks are rarely considered, where users may be forced to open search records and results. Therefore, deniable SE solutions against coercion attacks are presented, but they suffer from large storage overhead or fail to consider the dual coercion situation towards both sides of data owners and data users. In this paper, we roughly combine oblivious cross-tags protocol (OXT) and deniable encryption to propose a deniable SE (deniable cross-tag, DXT) scheme, which supports boolean queries and resists dual coercion attacks. Technically, we formalize a new primitive called updatable deniable encryption, and combine it with OXT in a non-trivial manner. In addition, we give formal system model, security model, and security proof of DXT. By employing the HUAWEI cloud platform, we conduct sufficient comparative experiments between DXT and state-of-the-art solutions based on a public dataset. The experimental results demonstrate that DXT outperforms higher search efficiency while achieving better features.","PeriodicalId":13202,"journal":{"name":"IEEE Transactions on Cloud Computing","volume":"13 3","pages":"969-982"},"PeriodicalIF":5.0,"publicationDate":"2025-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144998092","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Diversity of workload demands lays a critical impact on efficient resource allocation and management of cloud services. The existing literature has either weakly considered or overlooked the heterogeneous feature of job requests received from wide range of internet services users. To address this context, the proposed approach named Reliable and Energy Efficient Traffic Management (REE-TM) has exploited the diversity of internet traffic in terms of variation in resource demands and expected complexity. Specifically, REE-TM incorporates categorization of heterogeneous job requests and executes them by selecting the most admissible virtual node (a software-defined instance such as a virtual machine or container) and physical node (an actual hardware server or compute host) within the cloud infrastructure. To deal with resource-contention-based resource failures and performance degradation, a novel workload estimator ‘Toffoli Gate-based Quantum Neural Network’ (TG-QNN) is proposed, wherein learning process or interconnection weights optimization is achieved using Quantum version of BlackHole (QBHO) algorithm. The proactively estimated workload is used to compute entropy of the upcoming internet traffic with various traffic states analysis for detection of probable resource-congestion. REE-TM is extensively evaluated through simulations using a benchmark dataset and compared with optimal and without REE-TM versions. The performance evaluation and comparison of REE-TM with measured significant metrics reveal its effectiveness in assuring higher reliability by up to 30.25% and energy-efficiency by up to 23% as compared without REE-TM.
{"title":"REE-TM: Reliable and Energy-Efficient Traffic Management Model for Diverse Cloud Workloads","authors":"Ashutosh Kumar Singh;Deepika Saxena;Volker Lindenstruth","doi":"10.1109/TCC.2025.3581697","DOIUrl":"https://doi.org/10.1109/TCC.2025.3581697","url":null,"abstract":"Diversity of workload demands lays a critical impact on efficient resource allocation and management of cloud services. The existing literature has either weakly considered or overlooked the heterogeneous feature of job requests received from wide range of internet services users. To address this context, the proposed approach named <bold>R</b>eliable and <bold>E</b>nergy <bold>E</b>fficient <bold>T</b>raffic <bold>M</b>anagement (<bold>REE-TM</b>) has exploited the diversity of internet traffic in terms of variation in resource demands and expected complexity. Specifically, REE-TM incorporates categorization of heterogeneous job requests and executes them by selecting the most admissible <italic>virtual node</i> (a software-defined instance such as a virtual machine or container) and <italic>physical node</i> (an actual hardware server or compute host) within the cloud infrastructure. To deal with resource-contention-based resource failures and performance degradation, a novel workload estimator ‘Toffoli Gate-based Quantum Neural Network’ (TG-QNN) is proposed, wherein learning process or interconnection weights optimization is achieved using Quantum version of BlackHole (QBHO) algorithm. The proactively estimated workload is used to compute entropy of the upcoming internet traffic with various traffic states analysis for detection of probable resource-congestion. REE-TM is extensively evaluated through simulations using a benchmark dataset and compared with optimal and without REE-TM versions. The performance evaluation and comparison of REE-TM with measured significant metrics reveal its effectiveness in assuring higher reliability by up to 30.25% and energy-efficiency by up to 23% as compared without REE-TM.","PeriodicalId":13202,"journal":{"name":"IEEE Transactions on Cloud Computing","volume":"13 3","pages":"953-968"},"PeriodicalIF":5.0,"publicationDate":"2025-06-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144997133","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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 : 2025-06-10DOI: 10.1109/TCC.2025.3578557
William Pourmajidi;Lei Zhang;John Steinbacher;Tony Erwin;Andriy Miranskyy
The evolution of cloud computing has given rise to Cloud Native Applications (CNAs), presenting new challenges in governance, particularly when faced with strict compliance requirements. This work explores the unique characteristics of CNAs and their impact on governance. We introduce a comprehensive reference architecture designed to streamline governance across CNAs, along with a sample implementation, offering insights for both single and multi-cloud environments. Our architecture seamlessly integrates governance within the CNA framework, adhering to a “battery-included” philosophy. Tailored for both expansive and compact CNA deployments across various industries, this design enables cloud practitioners to prioritize product development by alleviating the complexities associated with governance. In addition, it provides a building block for academic exploration of generic CNA frameworks, highlighting their relevance in the evolving cloud computing landscape.
{"title":"A Reference Architecture for Governance of Cloud Native Applications","authors":"William Pourmajidi;Lei Zhang;John Steinbacher;Tony Erwin;Andriy Miranskyy","doi":"10.1109/TCC.2025.3578557","DOIUrl":"https://doi.org/10.1109/TCC.2025.3578557","url":null,"abstract":"The evolution of cloud computing has given rise to Cloud Native Applications (CNAs), presenting new challenges in governance, particularly when faced with strict compliance requirements. This work explores the unique characteristics of CNAs and their impact on governance. We introduce a comprehensive reference architecture designed to streamline governance across CNAs, along with a sample implementation, offering insights for both single and multi-cloud environments. Our architecture seamlessly integrates governance within the CNA framework, adhering to a “battery-included” philosophy. Tailored for both expansive and compact CNA deployments across various industries, this design enables cloud practitioners to prioritize product development by alleviating the complexities associated with governance. In addition, it provides a building block for academic exploration of generic CNA frameworks, highlighting their relevance in the evolving cloud computing landscape.","PeriodicalId":13202,"journal":{"name":"IEEE Transactions on Cloud Computing","volume":"13 3","pages":"935-952"},"PeriodicalIF":5.0,"publicationDate":"2025-06-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144998014","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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 : 2025-06-06DOI: 10.1109/TCC.2025.3577211
Jinfeng Wen;Haodi Ping
Serverless computing is a burgeoning cloud computing paradigm that allows developers to implement applications at the function level, known as serverless applications. Amazon Web Services (AWS), the leading provider in this field, offers Serverless Application Model (AWS SAM), a widely adopted configuration schema for configuring functions and managing resources. However, misconfigurations pose a major challenge during serverless application development, and existing methods are not applicable. To our knowledge, the configuration characteristics and misconfiguration detection for serverless applications have not been well explored. To address this gap, we collect and analyze 733 real-world serverless application configuration files using AWS SAM to understand their characteristics and challenges. Based on the insights, we design PHOENIX, a misconfiguration detection approach for serverless computing. PHOENIX learns configuration patterns from uniform representations of configurations and identifies potential misconfigurations that deviate from these patterns. To evaluate PHOENIX, we construct a dataset comprising 35 injected misconfigurations and 70 real-world misconfigurations with confirmed causes. Our results show that PHOENIX detects 100% of the injected misconfigurations and identifies 97.14% of real-world misconfigurations, significantly outperforming the state-of-the-art tool.
无服务器计算是一种新兴的云计算范式，它允许开发人员在功能级别实现应用程序，称为无服务器应用程序。该领域的领先提供商Amazon Web Services （AWS）提供无服务器应用程序模型（AWS SAM），这是一种广泛采用的配置模式，用于配置功能和管理资源。然而，在无服务器应用程序开发过程中，错误配置是一个主要的挑战，现有的方法不适用。据我们所知，无服务器应用程序的配置特征和错误配置检测还没有得到很好的研究。为了解决这一差距，我们使用AWS SAM收集并分析了733个实际的无服务器应用程序配置文件，以了解它们的特征和挑战。基于这些见解，我们设计了PHOENIX，一种用于无服务器计算的错误配置检测方法。PHOENIX从配置的统一表示中学习配置模式，并识别偏离这些模式的潜在错误配置。为了评估PHOENIX，我们构建了一个包含35个注入错误配置和70个真实错误配置的数据集，这些错误配置具有确定的原因。结果表明，PHOENIX可以检测到100%的注入错误配置，并识别出97.14%的实际错误配置，明显优于最先进的工具。
{"title":"PHOENIX: Misconfiguration Detection for AWS Serverless Computing","authors":"Jinfeng Wen;Haodi Ping","doi":"10.1109/TCC.2025.3577211","DOIUrl":"https://doi.org/10.1109/TCC.2025.3577211","url":null,"abstract":"Serverless computing is a burgeoning cloud computing paradigm that allows developers to implement applications at the function level, known as serverless applications. Amazon Web Services (AWS), the leading provider in this field, offers Serverless Application Model (AWS SAM), a widely adopted configuration schema for configuring functions and managing resources. However, misconfigurations pose a major challenge during serverless application development, and existing methods are not applicable. To our knowledge, the configuration characteristics and misconfiguration detection for serverless applications have not been well explored. To address this gap, we collect and analyze 733 real-world serverless application configuration files using AWS SAM to understand their characteristics and challenges. Based on the insights, we design <italic>PHOENIX</i>, a misconfiguration detection approach for serverless computing. <italic>PHOENIX</i> learns configuration patterns from uniform representations of configurations and identifies potential misconfigurations that deviate from these patterns. To evaluate <italic>PHOENIX</i>, we construct a dataset comprising 35 injected misconfigurations and 70 real-world misconfigurations with confirmed causes. Our results show that <italic>PHOENIX</i> detects 100% of the injected misconfigurations and identifies 97.14% of real-world misconfigurations, significantly outperforming the state-of-the-art tool.","PeriodicalId":13202,"journal":{"name":"IEEE Transactions on Cloud Computing","volume":"13 3","pages":"922-934"},"PeriodicalIF":5.0,"publicationDate":"2025-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144998012","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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 : 2025-06-02DOI: 10.1109/TCC.2025.3575790
Ruichao Mo;Weiwei Lin;Haocheng Zhong;Minxian Xu;Keqin Li
Nowadays, machine learning (ML)-based power prediction models for servers have shown remarkable performance, leveraging large volumes of labeled data for training. However, collecting extensive labeled power data from servers in cloud data centers incurs substantial costs. Additionally, varying resource demands across different workloads (e.g., CPU-intensive, memory-intensive, and I/O-intensive) lead to significant differences in power consumption behaviors, known as domain shift. Consequently, power data collected from one type of workload cannot effectively train power prediction models for other workloads, limiting the exploration of the collected power data. To tackle these challenges, we propose TGCP, a cross-workload power prediction method based on multi-source transfer Gaussian process regression. TGCP transfers knowledge from abundant power data across multiple source workloads to a target workload with limited power data. Furthermore, Continuous normalizing flows adjust the posterior prediction distribution of Gaussian process, making it locally non-Gaussian, enhancing TGCP’s ability to handle real-world power data distribution. This method enhances prediction accuracy for the target workload while reducing the expense of acquiring power data for real cloud data centers. Experimental results on a realistic power consumption dataset demonstrate that TGCP surpasses four traditional ML methods and three transfer learning methods in cross-workload power prediction.
{"title":"A Cross-Workload Power Prediction Method Based on Transfer Gaussian Process Regression in Cloud Data Centers","authors":"Ruichao Mo;Weiwei Lin;Haocheng Zhong;Minxian Xu;Keqin Li","doi":"10.1109/TCC.2025.3575790","DOIUrl":"https://doi.org/10.1109/TCC.2025.3575790","url":null,"abstract":"Nowadays, machine learning (ML)-based power prediction models for servers have shown remarkable performance, leveraging large volumes of labeled data for training. However, collecting extensive labeled power data from servers in cloud data centers incurs substantial costs. Additionally, varying resource demands across different workloads (e.g., CPU-intensive, memory-intensive, and I/O-intensive) lead to significant differences in power consumption behaviors, known as domain shift. Consequently, power data collected from one type of workload cannot effectively train power prediction models for other workloads, limiting the exploration of the collected power data. To tackle these challenges, we propose <italic>TGCP</i>, a cross-workload power prediction method based on multi-source transfer Gaussian process regression. <italic>TGCP</i> transfers knowledge from abundant power data across multiple source workloads to a target workload with limited power data. Furthermore, Continuous normalizing flows adjust the posterior prediction distribution of Gaussian process, making it locally non-Gaussian, enhancing <italic>TGCP</i>’s ability to handle real-world power data distribution. This method enhances prediction accuracy for the target workload while reducing the expense of acquiring power data for real cloud data centers. Experimental results on a realistic power consumption dataset demonstrate that <italic>TGCP</i> surpasses four traditional ML methods and three transfer learning methods in cross-workload power prediction.","PeriodicalId":13202,"journal":{"name":"IEEE Transactions on Cloud Computing","volume":"13 3","pages":"910-921"},"PeriodicalIF":5.0,"publicationDate":"2025-06-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144997970","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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 : 2025-04-24DOI: 10.1109/TCC.2025.3563848
Shusmoy Chowdhury;Ajay Katangur
The triumph of cloud computing hinges upon the adept instantiation of infrastructure and the judicious utilization of available resources. Load balancing, a pivotal facet, substantiates the fulfillment of these imperatives, thereby augmenting the performance of the cloud environment for its users. Our research introduces a load balancing algorithm grounded in threshold principles devised to ensure equitable distribution of workloads among nodes. The main objective of the algorithm is to preclude the overburdening of virtual machines (VMs) within the cloud with tasks or their idleness due to task allocation deficiencies in the presence of active tasks. The threshold values embedded in our algorithm ascertain the judicious deployment of VMs, forestalling both task overload and idle states arising from task allocation inadequacies. Simulation outcomes manifest that our threshold-based algorithm markedly enhances response time for tasks/requests and data processing duration within datacenters, outperforming extant algorithms such as First Come First Serve, Round Robin, and the Equally Spread Current Execution Load Balancing algorithm. Our threshold algorithm attains superior results to alternative load balancing algorithms when coupled with an optimized response time service broker policy.
{"title":"Optimizing Cloud Computing Performance Through Integration of a Threshold-Based Load Balancing Algorithm With Multiple Service Broker Policies","authors":"Shusmoy Chowdhury;Ajay Katangur","doi":"10.1109/TCC.2025.3563848","DOIUrl":"https://doi.org/10.1109/TCC.2025.3563848","url":null,"abstract":"The triumph of cloud computing hinges upon the adept instantiation of infrastructure and the judicious utilization of available resources. Load balancing, a pivotal facet, substantiates the fulfillment of these imperatives, thereby augmenting the performance of the cloud environment for its users. Our research introduces a load balancing algorithm grounded in threshold principles devised to ensure equitable distribution of workloads among nodes. The main objective of the algorithm is to preclude the overburdening of virtual machines (VMs) within the cloud with tasks or their idleness due to task allocation deficiencies in the presence of active tasks. The threshold values embedded in our algorithm ascertain the judicious deployment of VMs, forestalling both task overload and idle states arising from task allocation inadequacies. Simulation outcomes manifest that our threshold-based algorithm markedly enhances response time for tasks/requests and data processing duration within datacenters, outperforming extant algorithms such as First Come First Serve, Round Robin, and the Equally Spread Current Execution Load Balancing algorithm. Our threshold algorithm attains superior results to alternative load balancing algorithms when coupled with an optimized response time service broker policy.","PeriodicalId":13202,"journal":{"name":"IEEE Transactions on Cloud Computing","volume":"13 2","pages":"751-768"},"PeriodicalIF":5.3,"publicationDate":"2025-04-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144232134","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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 : 2025-04-16DOI: 10.1109/TCC.2025.3561281
Shuai Zhang;Baojun Chen;Weiqiang Sun;Weisheng Hu
When optical switching is deployed in Data Center Networks (DCNs), the reconfiguration of the optical switching matrix leads to substantially longer overheads, posing a significant impact on the system performance. Despite the extensive studies on the scheduling algorithms based on demand matrix decomposition (DMD), the stateful and irregular nature of the scheduling processes hinders the development of quantitative models, thereby limiting our understanding of resource scheduling in optical switching DCNs based on DMD. In this article, we model the DMD based resource scheduling process under a bursty and skewed traffic pattern and derive closed-form equations for the burst completion time. Our study shows that an increased reconfiguration delay will lead to an approximate linear increase in the burst completion time. Our study also demonstrates that the size of the slot and the maximum allowed duration of one match are approximately inversely proportional to the burst completion time, with diminishing marginal returns.
{"title":"Modeling Resource Scheduling in Optical Switching DCNs Under Bursty and Skewed Traffic","authors":"Shuai Zhang;Baojun Chen;Weiqiang Sun;Weisheng Hu","doi":"10.1109/TCC.2025.3561281","DOIUrl":"https://doi.org/10.1109/TCC.2025.3561281","url":null,"abstract":"When optical switching is deployed in Data Center Networks (DCNs), the reconfiguration of the optical switching matrix leads to substantially longer overheads, posing a significant impact on the system performance. Despite the extensive studies on the scheduling algorithms based on demand matrix decomposition (DMD), the stateful and irregular nature of the scheduling processes hinders the development of quantitative models, thereby limiting our understanding of resource scheduling in optical switching DCNs based on DMD. In this article, we model the DMD based resource scheduling process under a bursty and skewed traffic pattern and derive closed-form equations for the burst completion time. Our study shows that an increased reconfiguration delay will lead to an approximate linear increase in the burst completion time. Our study also demonstrates that the size of the slot and the maximum allowed duration of one match are approximately inversely proportional to the burst completion time, with diminishing marginal returns.","PeriodicalId":13202,"journal":{"name":"IEEE Transactions on Cloud Computing","volume":"13 2","pages":"737-750"},"PeriodicalIF":5.3,"publicationDate":"2025-04-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144230587","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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 : 2025-04-16DOI: 10.1109/TCC.2025.3561586
Yuuya Fukuchi;Sota Hashimoto;Kazuya Sakai;Satoshi Fukumoto;Min-Te Sun;Wei-Shinn Ku
Privacy-preserving k-nearest neighbor (PPkNN) classification for multiple clouds enables categorizing queried data into a class in keeping with data privacy, where the database and key servers jointly perform cryptographic operations. The existing solutions, unfortunately, take a long time and incur a large amount of traffic between the database and key servers. Therefore, in this article, we propose a fast and secure kNN classification protocol, namely FSkNN, over distributed databases deployed in multiple clouds under the semi-honest model. Particularly, we focus on optimizing the network-related operations during kNN classification. That is, the proposed cryptographic protocol reduces the number of interactions between the servers by using a fully homomorphic encryption scheme and eliminates unnecessary traffic by applying mathematical techniques. In addition, the indistinguishability-based security of FSkNN is proven. We implemented FSkNN with C++ and the testbed experiments demonstrate that the proposed scheme significantly facilitates the query response time and reduces the communication cost.
{"title":"Secure kNN for Distributed Cloud Environment Using Fully Homomorphic Encryption","authors":"Yuuya Fukuchi;Sota Hashimoto;Kazuya Sakai;Satoshi Fukumoto;Min-Te Sun;Wei-Shinn Ku","doi":"10.1109/TCC.2025.3561586","DOIUrl":"https://doi.org/10.1109/TCC.2025.3561586","url":null,"abstract":"Privacy-preserving k-nearest neighbor (PPkNN) classification for multiple clouds enables categorizing queried data into a class in keeping with data privacy, where the database and key servers jointly perform cryptographic operations. The existing solutions, unfortunately, take a long time and incur a large amount of traffic between the database and key servers. Therefore, in this article, we propose a fast and secure kNN classification protocol, namely FSkNN, over distributed databases deployed in multiple clouds under the semi-honest model. Particularly, we focus on optimizing the network-related operations during kNN classification. That is, the proposed cryptographic protocol reduces the number of interactions between the servers by using a fully homomorphic encryption scheme and eliminates unnecessary traffic by applying mathematical techniques. In addition, the indistinguishability-based security of FSkNN is proven. We implemented FSkNN with C++ and the testbed experiments demonstrate that the proposed scheme significantly facilitates the query response time and reduces the communication cost.","PeriodicalId":13202,"journal":{"name":"IEEE Transactions on Cloud Computing","volume":"13 2","pages":"721-736"},"PeriodicalIF":5.3,"publicationDate":"2025-04-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144230588","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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：