IEEE Transactions on Cloud Computing最新文献

{"title":"QoS-Aware, Cost-Efficient Scheduling for Data-Intensive DAGs in Multi-Tier Computing Environment","authors":"Paridhika Kayal;Alberto Leon-Garcia","doi":"10.1109/TCC.2024.3468913","DOIUrl":"https://doi.org/10.1109/TCC.2024.3468913","url":null,"abstract":"In today’s scientific landscape, Directed Acyclic Graphs (DAGs) are pivotal for representing task dependencies in data-intensive applications. Traditionally, two dominant bottom-up DAG scheduling approaches exist: one overlooks communication contention and the other fails to exploit parallelization for improving latency. This study distinguishes itself by advocating a top-down approach prioritizing latency or cost optimization in multi-tier environments to fulfill QoS and SLA requirements. Our strategy effectively mitigates bandwidth contention and facilitates parallel executions, leading to substantial completion time reductions. Our findings suggest that myopic knowledge-based scheduling, emphasizing latency or cost minimization, can yield benefits comparable to its look-ahead counterparts. Through latency-efficient and cost-efficient topological sorting, our \u0000<italic>wDAGSplit</i>\u0000 strategy introduces a two-stage partitioning and scheduling approach. Its simplicity and adaptability extend its usability to DAGs of any scale. Evaluated on over 100,000 real-world DAG applications, \u0000<italic>wDAGSplit</i>\u0000 demonstrates latency enhancements of up to 80x compared to Edge-only scenarios, 15x to Near-Edge-only, and 6x to Cloud-only. In terms of cost, our approach achieves enhancements of up to 60x compared to Edge-only scenarios, 250x to NE-only, and 70x to Cloud-only. Moreover, for DAGs with 50 tasks, we achieve 5x reduced latency compared to previous approaches, along with a complexity reduction of up to 24 times.","PeriodicalId":13202,"journal":{"name":"IEEE Transactions on Cloud Computing","volume":"12 4","pages":"1314-1327"},"PeriodicalIF":5.3,"publicationDate":"2024-09-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142777763","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}

{"title":"Anomaly Transformer Ensemble Model for Cloud Data Anomaly Detection","authors":"Won Sakong;Jongyeop Kwon;Kyungha Min;Suyeon Wang;Wooju Kim","doi":"10.1109/TCC.2024.3466174","DOIUrl":"https://doi.org/10.1109/TCC.2024.3466174","url":null,"abstract":"The stability and user trust in cloud services depends on prompt detection and response to diverse anomalies. This study focuses on an Ensemble-based anomaly detection methodology that integrates log data with computing resource metrics, aiming to overcome the limitations of traditional single-data models. To process the unstructured nature of log data, we use the Drain Parser to transform it into a structured format, and Doc2Vec embeds it. The study adheres to a reconstruction-based approach for anomaly detection, specifically building upon the Anomaly Transformer model. The proposed model leverages the concept of an Anomaly Transformer based on the Attention mechanism. It integrates preprocessed metric data with log data for effective anomaly detection. Experiments were conducted using metric and log data collected from real-world cloud environments. The model’s performance was evaluated based on accuracy, recall, precision, f1 score, and AUROC. The results demonstrate that our proposed Ensemble-based model outperforms traditional models such as LSTM, VAR, and deeplog.","PeriodicalId":13202,"journal":{"name":"IEEE Transactions on Cloud Computing","volume":"12 4","pages":"1305-1313"},"PeriodicalIF":5.3,"publicationDate":"2024-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142777685","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}

{"title":"WorkloadDiff: Conditional Denoising Diffusion Probabilistic Models for Cloud Workload Prediction","authors":"Weiping Zheng;Zongxiao Chen;Kaiyuan Zheng;Weijian Zheng;Yiqi Chen;Xiaomao Fan","doi":"10.1109/TCC.2024.3461649","DOIUrl":"10.1109/TCC.2024.3461649","url":null,"abstract":"Accurate workload forecasting plays a crucial role in optimizing resource allocation, enhancing performance, and reducing energy consumption in cloud data centers. Deep learning-based methods have emerged as the dominant approach in this field, exhibiting exceptional performance. However, most existing methods lack the ability to quantify confidence, limiting their practical decision-making utility. To address this limitation, we propose a novel denoising diffusion probabilistic model (DDPM)-based method, termed WorkloadDiff, for multivariate probabilistic workload prediction. WorkloadDiff leverages both original and noisy signals from input conditions using a two-path neural network. Additionally, we introduce a multi-scale feature extraction method and an adaptive fusion approach to capture diverse temporal patterns within the workload. To enhance consistency between conditions and predicted values, we incorporate a resampling strategy into the inference of WorkloadDiff. Extensive experiments conducted on four public datasets demonstrate the superior performance of WorkloadDiff over all baseline models, establishing it as a robust tool for resource management in cloud data centers.","PeriodicalId":13202,"journal":{"name":"IEEE Transactions on Cloud Computing","volume":"12 4","pages":"1291-1304"},"PeriodicalIF":5.3,"publicationDate":"2024-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142260182","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}

{"title":"A Lightweight Privacy-Preserving Ciphertext Retrieval Scheme Based on Edge Computing","authors":"Na Wang;Wen Zhou;Qingyun Han;Jianwei Liu;Weilue Liao;Junsong Fu","doi":"10.1109/TCC.2024.3461732","DOIUrl":"10.1109/TCC.2024.3461732","url":null,"abstract":"With the rapid development of cloud computing and Internet of Things (IoT) technologies, large amounts of data collected from IoT devices are encrypted and outsourced to cloud servers for storage and sharing. However, traditional ciphertext retrieval schemes impose high computation and storage overhead on end users. Meanwhile, IoT devices with limited resources are difficult to adapt to large amounts of data computation and transmission, which leads to transmission delay and poor user experience. In this article, we propose a lightweight privacy-preserving ciphertext retrieval scheme based on edge computing (LPCR) by extending searchable encryption (SE) and ciphertext policy attribute-based encryption (CP-ABE) techniques. First, to avoid network delay and paralysis, we introduce edge servers into LPCR and design a collaboration mechanism between the user side and the edge servers. The user side only needs to accomplish lightweight computation and storage tasks, which greatly reduces their resource consumption. Second, we extend the basic ciphertext policy attribute-based keyword search (CP-ABKS) technique and design the Linear Secret Sharing Scheme (LSSS) access control algorithm with attribute values to hide access policies and attributes. In addition, to improve the retrieval accuracy, the document indexes and query trapdoors are set up by conjunctive keywords to help the cloud server locate exactly the data that the user wishes to query. Formal security analysis verifies that LPCR can achieve the security of chosen plaintext attack (CPA) and chosen keyword attack (CKA), and resist collusion attack. Simulation experiments prove that LPCR is lightweight and feasible.","PeriodicalId":13202,"journal":{"name":"IEEE Transactions on Cloud Computing","volume":"12 4","pages":"1273-1290"},"PeriodicalIF":5.3,"publicationDate":"2024-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142260183","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}

{"title":"Generative Adversarial Privacy for Multimedia Analytics Across the IoT-Edge Continuum","authors":"Xin Wang;Jianhui Lv;Byung-Gyu Kim;Carsten Maple;B. D. Parameshachari;Adam Slowik;Keqin Li","doi":"10.1109/TCC.2024.3459789","DOIUrl":"10.1109/TCC.2024.3459789","url":null,"abstract":"The proliferation of multimedia-enabled IoT devices and edge computing enables a new class of data-intensive applications. However, analyzing the massive volumes of multimedia data presents significant privacy challenges. We propose a novel framework called generative adversarial privacy (GAP) that leverages generative adversarial networks (GANs) to synthesize privacy-preserving surrogate data for multimedia analytics across the IoT-Edge continuum. GAP carefully perturbs the GAN's training process to provide rigorous differential privacy guarantees without compromising utility. Moreover, we present optimization strategies, including dynamic privacy budget allocation, adaptive gradient clipping, and weight clustering to improve convergence and data quality under a constrained privacy budget. Theoretical analysis proves that GAP provides rigorous privacy protections while enabling high-fidelity analytics. Extensive experiments on real-world multimedia datasets demonstrate that GAP outperforms existing methods, producing high-quality synthetic data for privacy-preserving multimedia processing in diverse IoT-Edge applications.","PeriodicalId":13202,"journal":{"name":"IEEE Transactions on Cloud Computing","volume":"12 4","pages":"1260-1272"},"PeriodicalIF":5.3,"publicationDate":"2024-09-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142218868","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}

{"title":"Corrections to “DNN Surgery: Accelerating DNN Inference on the Edge through Layer Partitioning”","authors":"Huanghuang Liang;Qianlong Sang;Chuang Hu;Dazhao Cheng;Xiaobo Zhou;Dan Wang;Wei Bao;Yu Wang","doi":"10.1109/TCC.2024.3404548","DOIUrl":"https://doi.org/10.1109/TCC.2024.3404548","url":null,"abstract":"In this paper, we reference the previous conference version and complete the grant number mentioned in the acknowledgments of the conference version.","PeriodicalId":13202,"journal":{"name":"IEEE Transactions on Cloud Computing","volume":"12 3","pages":"966-966"},"PeriodicalIF":5.3,"publicationDate":"2024-09-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10666943","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142143671","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"FedPAW: Federated Learning With Personalized Aggregation Weights for Urban Vehicle Speed Prediction","authors":"Yuepeng He;Pengzhan Zhou;Yijun Zhai;Fang Qu;Zhida Qin;Mingyan Li;Songtao Guo","doi":"10.1109/TCC.2024.3452696","DOIUrl":"10.1109/TCC.2024.3452696","url":null,"abstract":"Vehicle speed prediction is crucial for intelligent transportation systems, promoting more reliable autonomous driving by accurately predicting future vehicle conditions. Due to variations in drivers’ driving styles and vehicle types, speed predictions for different target vehicles may significantly differ. Existing methods may not realize personalized vehicle speed prediction while protecting drivers’ data privacy. We propose a Federated learning framework with Personalized Aggregation Weights (FedPAW) to overcome these challenges. This method captures client-specific information by measuring the weighted mean squared error between the parameters of local models and global models. The server sends tailored aggregated models to clients instead of a single global model, without incurring additional computational and communication overhead for clients. To evaluate the effectiveness of FedPAW, we collected driving data in urban scenarios using the autonomous driving simulator CARLA, employing an LSTM-based Seq2Seq model with a multi-head attention mechanism to predict the future speed of target vehicles. The results demonstrate that our proposed FedPAW ranks lowest in prediction error within the time horizon of 10 seconds, with a 0.8% reduction in test MAE, compared to eleven representative benchmark baselines.","PeriodicalId":13202,"journal":{"name":"IEEE Transactions on Cloud Computing","volume":"12 4","pages":"1248-1259"},"PeriodicalIF":5.3,"publicationDate":"2024-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142218871","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}

{"title":"Large-Scale Measurements and Optimizations on Latency in Edge Clouds","authors":"Heng Zhang;Shaoyuan Huang;Mengwei Xu;Deke Guo;Xiaofei Wang;Xin Wang;Victor C. M. Leung;Wenyu Wang","doi":"10.1109/TCC.2024.3452094","DOIUrl":"10.1109/TCC.2024.3452094","url":null,"abstract":"The emergence of next-generation latency-critical applications places strict requirements on network latency and stability. Edge cloud, an instantiated paradigm for edge computing, is gaining more and more attention due to its benefits of low latency. In this work, we make an in-depth investigation into the network QoS, especially end-to-end latency, at both spatial and temporal dimensions on a nationwide edge computing platform. Through the measurements, we collect a multi-variable large-scale real-world dataset on latency. We then quantify how the spatial-temporal factors affect the end-to-end latency, and verify the predictability of end-to-end latency. The results reveal the limitation of centralized clouds and illustrate how could edge clouds provide low and stable latency. Our results also point out that existing edge clouds merely increase the density of servers and ignore spatial-temporal factors, so they still suffer from high latency and fluctuations. Based on a quantified latency impact factor, we have proposed several optimization strategies for edge cloud latency and validated their effectiveness. We also propose a robust prototype edge cloud model based on lessons we learn from the measurement and evaluate its performance in the production environment. Evaluation result shows that edge clouds achieve 84.1% latency reduction with 0.5 ms latency fluctuation and 73.3% QoS improvement compared with the centralized clouds.","PeriodicalId":13202,"journal":{"name":"IEEE Transactions on Cloud Computing","volume":"12 4","pages":"1218-1231"},"PeriodicalIF":5.3,"publicationDate":"2024-08-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142218872","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}

{"title":"Attribute-Based Hierarchical Keyword Auditing With Batch Fault Localization Assisted by Smart Contracts","authors":"Jingting Xue;Shuqin Luo;Fagen Li;Wenzheng Zhang;Liang Liu;Yu Zhou;Xiaojun Zhang","doi":"10.1109/TCC.2024.3452324","DOIUrl":"10.1109/TCC.2024.3452324","url":null,"abstract":"Keyword-based auditing (KA) provides a means for users to verify the integrity of only the outsourced data they are interested in. Existing KA schemes employ relation authentication labels to conduct targeted audits with keywords, which significantly improves the cost-effectiveness. However, such schemes typically support only a single-challenge scenario, which may not always be practical. To overcome this constraint, we introduce a hierarchical challenge mechanism grounded in user attributes. This mechanism leverages inequality and affiliation relationships to comply with a predefined tree structure for access policies. Incorporated during the challenge-response phase of the auditing model, it permits users to initiate cross-challenges. Expanding upon this hierarchical mechanism, we propose an attribute-based hierarchical keyword auditing scheme, abbreviated as \u0000<inline-formula><tex-math>$mathcal{AHKA}$</tex-math></inline-formula>\u0000. \u0000<inline-formula><tex-math>$mathcal{AHKA}$</tex-math></inline-formula>\u0000 combines searchable encryption to conduct cross-targeted audits and benefits from the hash collision mapping of Bloom filters to safeguard against keyword guessing attacks. Moreover, we design a fault localization algorithm based on a variant of the binary search technique. It locates in batch the faulty cloud servers and damaged data blocks after an audit failure. As an integral part of \u0000<inline-formula><tex-math>$mathcal{AHKA}$</tex-math></inline-formula>\u0000, the algorithm significantly enhances our scheme's practicability. Security analyses indicate that \u0000<inline-formula><tex-math>$mathcal{AHKA}$</tex-math></inline-formula>\u0000 can effectively withstand both forgery and replace attacks on audit proofs. The smart contract component ensures that our scheme's processes can be monitored and regulated. Experimental data corroborate that deploying \u0000<inline-formula><tex-math>$mathcal{AHKA}$</tex-math></inline-formula>\u0000 on the client side and on the blockchain is both efficient and feasible.","PeriodicalId":13202,"journal":{"name":"IEEE Transactions on Cloud Computing","volume":"12 4","pages":"1232-1247"},"PeriodicalIF":5.3,"publicationDate":"2024-08-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142218870","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}

{"title":"Online Pricing and Resource Scheduling for Profit Maximization of Cloud Storage Providers","authors":"Kyungtae Lee;Yeongjin Kim","doi":"10.1109/TCC.2024.3450876","DOIUrl":"10.1109/TCC.2024.3450876","url":null,"abstract":"There is increasing competition among cloud object storage service (COSS) providers as the demand for COSSs grows. However, existing pricing models offered by commercial COSS providers fail to effectively adapt to changing client demand and resource supply. Consequently, many COSS providers are still grappling with operational challenges in maximizing their profits, such as pricing policy, load balancing, server scheduling, and energy management. In this paper, we propose a novel approach called time-dependent pricing and scheduling (\u0000<italic>TD-PnS</i>\u0000), which is based on the Lyapunov-drift-minus-profit technique. To maximize the profits of COSS providers, \u0000<italic>TD-PnS</i>\u0000 enables joint and dynamic decision-making across several key factors that have been dealt with separately so far: \u0000<italic>(i)</i>\u0000 service pricing, \u0000<italic>(ii)</i>\u0000 CPU clock scaling and encoding scheduling, \u0000<italic>(iii)</i>\u0000 network scheduling, and \u0000<italic>(iv)</i>\u0000 energy storage management. We propose an enhanced version of \u0000<italic>TD-PnS</i>\u0000, called \u0000<italic>TD-PnS-Adv</i>\u0000, further to improve other aspects, such as system stabilization. Finally, through trace-driven simulations utilizing a real dataset, we demonstrate the superior performance of the proposed algorithms compared to existing algorithms and pricing models in terms of profit maximization.","PeriodicalId":13202,"journal":{"name":"IEEE Transactions on Cloud Computing","volume":"12 4","pages":"1186-1199"},"PeriodicalIF":5.3,"publicationDate":"2024-08-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142218873","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}