Pub Date : 2024-02-06DOI: 10.1109/TSUSC.2024.3362697
Yiming Wang;Meng Hao;Hui He;Weizhe Zhang;Qiuyuan Tang;Xiaoyang Sun;Zheng Wang
Power and energy consumption is the limiting factor of modern computing systems. As the GPU becomes a mainstream computing device, power management for GPUs becomes increasingly important. Current works focus on GPU kernel-level power management, with challenges in portability due to architecture-specific considerations. We present DRLCap , a general runtime power management framework intended to support power management across various GPU architectures. It periodically monitors system-level information to dynamically detect program phase changes and model the workload and GPU system behavior. This elimination from kernel-specific constraints enhances adaptability and responsiveness. The framework leverages dynamic GPU frequency capping, which is the most widely used power knob, to control the power consumption. DRLCap employs deep reinforcement learning (DRL) to adapt to the changing of program phases by automatically adjusting its power policy through online learning, aiming to reduce the GPU power consumption without significantly compromising the application performance. We evaluate DRLCap on three NVIDIA and one AMD GPU architectures. Experimental results show that DRLCap improves prior GPU power optimization strategies by a large margin. On average, it reduces the GPU energy consumption by 22% with less than 3% performance slowdown on NVIDIA GPUs. This translates to a 20% improvement in the energy efficiency measured by the energy-delay product (EDP) over the NVIDIA default GPU power management strategy. For the AMD GPU architecture, DRLCap saves energy consumption by 10%, on average, with a 4% percentage loss, and improves energy efficiency by 8%.
{"title":"DRLCAP: Runtime GPU Frequency Capping With Deep Reinforcement Learning","authors":"Yiming Wang;Meng Hao;Hui He;Weizhe Zhang;Qiuyuan Tang;Xiaoyang Sun;Zheng Wang","doi":"10.1109/TSUSC.2024.3362697","DOIUrl":"https://doi.org/10.1109/TSUSC.2024.3362697","url":null,"abstract":"Power and energy consumption is the limiting factor of modern computing systems. As the GPU becomes a mainstream computing device, power management for GPUs becomes increasingly important. Current works focus on GPU kernel-level power management, with challenges in portability due to architecture-specific considerations. We present \u0000<sc>DRLCap</small>\u0000, a general runtime power management framework intended to support power management across various GPU architectures. It periodically monitors system-level information to dynamically detect program phase changes and model the workload and GPU system behavior. This elimination from kernel-specific constraints enhances adaptability and responsiveness. The framework leverages dynamic GPU frequency capping, which is the most widely used power knob, to control the power consumption. \u0000<sc>DRLCap</small>\u0000 employs deep reinforcement learning (DRL) to adapt to the changing of program phases by automatically adjusting its power policy through online learning, aiming to reduce the GPU power consumption without significantly compromising the application performance. We evaluate \u0000<sc>DRLCap</small>\u0000 on three NVIDIA and one AMD GPU architectures. Experimental results show that \u0000<sc>DRLCap</small>\u0000 improves prior GPU power optimization strategies by a large margin. On average, it reduces the GPU energy consumption by 22% with less than 3% performance slowdown on NVIDIA GPUs. This translates to a 20% improvement in the energy efficiency measured by the energy-delay product (EDP) over the NVIDIA default GPU power management strategy. For the AMD GPU architecture, \u0000<sc>DRLCap</small>\u0000 saves energy consumption by 10%, on average, with a 4% percentage loss, and improves energy efficiency by 8%.","PeriodicalId":13268,"journal":{"name":"IEEE Transactions on Sustainable Computing","volume":"9 5","pages":"712-726"},"PeriodicalIF":3.0,"publicationDate":"2024-02-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142397260","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 : 2024-02-05DOI: 10.1109/TSUSC.2024.3362074
Ke Gu;XingQiang Wang;Xiong Li
The security of fog computing has been researched and concerned with its development, where malicious attacks pose a greater threat to distributed data storage based on fog computing. Also, the rapid increasing on the number of terminal devices has raised the importance of fog computing-based distributed data storage. In response to this demand, it is essential to establish a secure and privacy-preserving distributed data auditing method that enables security protection of stored data and effective control over identities of auditors. In this paper, we propose a dynamic outsourced data audit scheme for Merkle hash grid-based fog storage with privacy-preserving, where fog servers are used to undertake partial outsourced computation and data storage. Our scheme can provide the function of privacy-preserving for outsourced data by blinding original stored data, and supports data owners to define their auditing access policies by the linear secret-sharing scheme to control the identities of auditors. Further, the construction of Merkle hash grid is used to improve the efficiency of dynamic data operations. Also, a server locating approach is proposed to enable the third-part auditor to identify specific malicious data fog servers within distributed data storage. Under the proposed security model, the security of our scheme can be proved, which can further provide collusion resistance and privacy-preserving for outsourced data. Additionally, both theoretical and experimental evaluations illustrate the efficiency of our proposed scheme.
{"title":"Dynamic Outsourced Data Audit Scheme for Merkle Hash Grid-Based Fog Storage With Privacy-Preserving","authors":"Ke Gu;XingQiang Wang;Xiong Li","doi":"10.1109/TSUSC.2024.3362074","DOIUrl":"https://doi.org/10.1109/TSUSC.2024.3362074","url":null,"abstract":"The security of fog computing has been researched and concerned with its development, where malicious attacks pose a greater threat to distributed data storage based on fog computing. Also, the rapid increasing on the number of terminal devices has raised the importance of fog computing-based distributed data storage. In response to this demand, it is essential to establish a secure and privacy-preserving distributed data auditing method that enables security protection of stored data and effective control over identities of auditors. In this paper, we propose a dynamic outsourced data audit scheme for Merkle hash grid-based fog storage with privacy-preserving, where fog servers are used to undertake partial outsourced computation and data storage. Our scheme can provide the function of privacy-preserving for outsourced data by blinding original stored data, and supports data owners to define their auditing access policies by the linear secret-sharing scheme to control the identities of auditors. Further, the construction of Merkle hash grid is used to improve the efficiency of dynamic data operations. Also, a server locating approach is proposed to enable the third-part auditor to identify specific malicious data fog servers within distributed data storage. Under the proposed security model, the security of our scheme can be proved, which can further provide collusion resistance and privacy-preserving for outsourced data. Additionally, both theoretical and experimental evaluations illustrate the efficiency of our proposed scheme.","PeriodicalId":13268,"journal":{"name":"IEEE Transactions on Sustainable Computing","volume":"9 4","pages":"695-711"},"PeriodicalIF":3.0,"publicationDate":"2024-02-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141965829","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 : 2024-02-01DOI: 10.1109/TSUSC.2024.3360975
Fu Jiang;Yaoxin Xia;Lisen Yan;Weirong Liu;Xiaoyong Zhang;Heng Li;Jun Peng
Battery degradation is a main hinder to extend the persistent lifespan of the portable heterogeneous computing device. Excessive energy consumption and prominent current fluctuations can lead to a sharp decline of battery endurance. To address this issue, a battery-aware workflow scheduling algorithm is proposed to maximize the battery lifetime and release the computing potential of the device fully. First, a dynamic optimal budget strategy is developed to select the highest cost-effectiveness processors to meet the deadline of each task, accelerating the budget optimization by incorporating deep neural network. Second, an integer-programming greedy strategy is utilized to determine the start time of each task, minimizing the fluctuation of the battery supply current to mitigate the battery degradation. Finally, a long-term operation experiment and Monte Carlo experiments are performed on the battery simulator, SLIDE. The experimental results under real operating conditions for more than 1800 hours validate that the proposed scheduling algorithm can effectively extend the battery life by 7.31%-8.23%. The results on various parallel workflows illustrate that the proposed algorithm has comparable performance with speed improvement over the integer programming method.
{"title":"Battery-Aware Workflow Scheduling for Portable Heterogeneous Computing","authors":"Fu Jiang;Yaoxin Xia;Lisen Yan;Weirong Liu;Xiaoyong Zhang;Heng Li;Jun Peng","doi":"10.1109/TSUSC.2024.3360975","DOIUrl":"https://doi.org/10.1109/TSUSC.2024.3360975","url":null,"abstract":"Battery degradation is a main hinder to extend the persistent lifespan of the portable heterogeneous computing device. Excessive energy consumption and prominent current fluctuations can lead to a sharp decline of battery endurance. To address this issue, a battery-aware workflow scheduling algorithm is proposed to maximize the battery lifetime and release the computing potential of the device fully. First, a dynamic optimal budget strategy is developed to select the highest cost-effectiveness processors to meet the deadline of each task, accelerating the budget optimization by incorporating deep neural network. Second, an integer-programming greedy strategy is utilized to determine the start time of each task, minimizing the fluctuation of the battery supply current to mitigate the battery degradation. Finally, a long-term operation experiment and Monte Carlo experiments are performed on the battery simulator, SLIDE. The experimental results under real operating conditions for more than 1800 hours validate that the proposed scheduling algorithm can effectively extend the battery life by 7.31%-8.23%. The results on various parallel workflows illustrate that the proposed algorithm has comparable performance with speed improvement over the integer programming method.","PeriodicalId":13268,"journal":{"name":"IEEE Transactions on Sustainable Computing","volume":"9 4","pages":"677-694"},"PeriodicalIF":3.0,"publicationDate":"2024-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141965765","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}
Computing servers have played a key role in developing and processing emerging compute-intensive applications in recent years. Consolidating multiple virtual machines (VMs) inside one server to run various applications introduces severe competence for limited resources among VMs. Many techniques such as VM scheduling and resource provisioning are proposed to maximize the cost-efficiency of the computing servers while alleviating the performance inference between VMs. However, these management techniques require accurate performance prediction of the application running inside the VM, which is challenging to get in the public cloud due to the black-box nature of the VMs. From this perspective, this paper proposes a novel machine learning-based performance prediction approach for applications running in the cloud. To achieve high-accuracy predictions for black-box VMs, the proposed method first identifies the running application inside the virtual machine. It then selects highly correlated runtime metrics as the input of the machine learning approach to accurately predict the performance level of the cloud application. Experimental results with state-of-the-art cloud benchmarks demonstrate that our proposed method outperforms existing prediction methods by more than 2× in terms of the worst prediction error. In addition, we successfully tackle the challenge of performance prediction for applications with variable workloads by introducing the performance degradation index, which other comparison methods fail to consider. The workflow versatility of the proposed approach has been verified with different modern servers and VM configurations.
{"title":"CloudProphet: A Machine Learning-Based Performance Prediction for Public Clouds","authors":"Darong Huang;Luis Costero;Ali Pahlevan;Marina Zapater;David Atienza","doi":"10.1109/TSUSC.2024.3359325","DOIUrl":"https://doi.org/10.1109/TSUSC.2024.3359325","url":null,"abstract":"Computing servers have played a key role in developing and processing emerging compute-intensive applications in recent years. Consolidating multiple virtual machines (VMs) inside one server to run various applications introduces severe competence for limited resources among VMs. Many techniques such as VM scheduling and resource provisioning are proposed to maximize the cost-efficiency of the computing servers while alleviating the performance inference between VMs. However, these management techniques require accurate performance prediction of the application running inside the VM, which is challenging to get in the public cloud due to the black-box nature of the VMs. From this perspective, this paper proposes a novel machine learning-based performance prediction approach for applications running in the cloud. To achieve high-accuracy predictions for black-box VMs, the proposed method first identifies the running application inside the virtual machine. It then selects highly correlated runtime metrics as the input of the machine learning approach to accurately predict the performance level of the cloud application. Experimental results with state-of-the-art cloud benchmarks demonstrate that our proposed method outperforms existing prediction methods by more than 2× in terms of the worst prediction error. In addition, we successfully tackle the challenge of performance prediction for applications with variable workloads by introducing the performance degradation index, which other comparison methods fail to consider. The workflow versatility of the proposed approach has been verified with different modern servers and VM configurations.","PeriodicalId":13268,"journal":{"name":"IEEE Transactions on Sustainable Computing","volume":"9 4","pages":"661-676"},"PeriodicalIF":3.0,"publicationDate":"2024-01-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141965766","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 : 2024-01-26DOI: 10.1109/TSUSC.2024.3358915
Aman Mishra;Yash Garg;Om Jee Pandey;Mahendra K. Shukla;Athanasios V. Vasilakos;Rajesh M. Hegde
At present, the centralized learning models, used for IoT applications generating large amount of data, face several challenges such as bandwidth scarcity, more energy consumption, increased uses of computing resources, poor connectivity, high computational complexity, reduced privacy, and large latency towards data transfer. In order to address the aforementioned challenges, Blockchain-Enabled Federated Learning Networks (BFLNs) emerged recently, which deal with trained model parameters only, rather than raw data. BFLNs provide enhanced security along with improved energy-efficiency and Quality-of-Service (QoS). However, BFLNs suffer with the challenges of exponential increased action space in deciding various parameter levels towards training and block generation. Motivated by aforementioned challenges of BFLNs, in this work, we are proposing an actor-critic Reinforcement Learning (RL) method to model the Machine Learning Model Owner (MLMO) in selecting the optimal set of parameter levels, addressing the challenges of exponential grow of action space in BFLNs. Further, due to the implicit entropy exploration, actor-critic RL method balances the exploration-exploitation trade-off and shows better performance than most off-policy methods, on large discrete action spaces. Therefore, in this work, considering the mobile scenario of the devices, MLMO decides the data and energy levels that the mobile devices use for the training and determine the block generation rate. This leads to minimized system latency and reduced overall cost, while achieving the target accuracy. Specifically, we have used Proximal Policy Optimization (PPO) as an on-policy actor-critic method with it's two variants, one based on Monte Carlo (MC) returns and another based on Generalized Advantage Estimate (GAE). We analyzed that PPO has better exploration and sample efficiency, lesser training time, and consistently higher cumulative rewards, when compared to off-policy Deep Q-Network (DQN).
{"title":"A Novel Resource Management Framework for Blockchain-Based Federated Learning in IoT Networks","authors":"Aman Mishra;Yash Garg;Om Jee Pandey;Mahendra K. Shukla;Athanasios V. Vasilakos;Rajesh M. Hegde","doi":"10.1109/TSUSC.2024.3358915","DOIUrl":"https://doi.org/10.1109/TSUSC.2024.3358915","url":null,"abstract":"At present, the centralized learning models, used for IoT applications generating large amount of data, face several challenges such as bandwidth scarcity, more energy consumption, increased uses of computing resources, poor connectivity, high computational complexity, reduced privacy, and large latency towards data transfer. In order to address the aforementioned challenges, Blockchain-Enabled Federated Learning Networks (BFLNs) emerged recently, which deal with trained model parameters only, rather than raw data. BFLNs provide enhanced security along with improved energy-efficiency and Quality-of-Service (QoS). However, BFLNs suffer with the challenges of exponential increased action space in deciding various parameter levels towards training and block generation. Motivated by aforementioned challenges of BFLNs, in this work, we are proposing an actor-critic Reinforcement Learning (RL) method to model the Machine Learning Model Owner (MLMO) in selecting the optimal set of parameter levels, addressing the challenges of exponential grow of action space in BFLNs. Further, due to the implicit entropy exploration, actor-critic RL method balances the exploration-exploitation trade-off and shows better performance than most off-policy methods, on large discrete action spaces. Therefore, in this work, considering the mobile scenario of the devices, MLMO decides the data and energy levels that the mobile devices use for the training and determine the block generation rate. This leads to minimized system latency and reduced overall cost, while achieving the target accuracy. Specifically, we have used Proximal Policy Optimization (PPO) as an on-policy actor-critic method with it's two variants, one based on Monte Carlo (MC) returns and another based on Generalized Advantage Estimate (GAE). We analyzed that PPO has better exploration and sample efficiency, lesser training time, and consistently higher cumulative rewards, when compared to off-policy Deep Q-Network (DQN).","PeriodicalId":13268,"journal":{"name":"IEEE Transactions on Sustainable Computing","volume":"9 4","pages":"648-660"},"PeriodicalIF":3.0,"publicationDate":"2024-01-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141965830","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}
Sea State Estimation (SSE) is essential for Internet of Things (IoT)-enabled autonomous ships, which rely on favorable sea conditions for safe and efficient navigation. Traditional methods, such as wave buoys and radars, are costly, less accurate, and lack real-time capability. Model-driven methods, based on physical models of ship dynamics, are impractical due to wave randomness. Data-driven methods are limited by the data imbalance problem, as some sea states are more frequent and observable than others. To overcome these challenges, we propose a novel data-driven approach for SSE based on ship motion data. Our approach consists of three main components: a data preprocessing module, a parallel convolution feature extractor, and a theoretical-ensured distance-based classifier. The data preprocessing module aims to enhance the data quality and reduce sensor noise. The parallel convolution feature extractor uses a kernel-varying convolutional structure to capture distinctive features. The distance-based classifier learns representative prototypes for each sea state and assigns a sample to the nearest prototype based on a distance metric. The efficiency of our model is validated through experiments on two SSE datasets and the UEA archive, encompassing thirty multivariate time series classification tasks. The results reveal the generalizability and robustness of our approach.
{"title":"A Prototype-Empowered Kernel-Varying Convolutional Model for Imbalanced Sea State Estimation in IoT-Enabled Autonomous Ship","authors":"Mengna Liu;Xu Cheng;Fan Shi;Xiufeng Liu;Hongning Dai;Shengyong Chen","doi":"10.1109/TSUSC.2024.3353183","DOIUrl":"https://doi.org/10.1109/TSUSC.2024.3353183","url":null,"abstract":"Sea State Estimation (SSE) is essential for Internet of Things (IoT)-enabled autonomous ships, which rely on favorable sea conditions for safe and efficient navigation. Traditional methods, such as wave buoys and radars, are costly, less accurate, and lack real-time capability. Model-driven methods, based on physical models of ship dynamics, are impractical due to wave randomness. Data-driven methods are limited by the data imbalance problem, as some sea states are more frequent and observable than others. To overcome these challenges, we propose a novel data-driven approach for SSE based on ship motion data. Our approach consists of three main components: a data preprocessing module, a parallel convolution feature extractor, and a theoretical-ensured distance-based classifier. The data preprocessing module aims to enhance the data quality and reduce sensor noise. The parallel convolution feature extractor uses a kernel-varying convolutional structure to capture distinctive features. The distance-based classifier learns representative prototypes for each sea state and assigns a sample to the nearest prototype based on a distance metric. The efficiency of our model is validated through experiments on two SSE datasets and the UEA archive, encompassing thirty multivariate time series classification tasks. The results reveal the generalizability and robustness of our approach.","PeriodicalId":13268,"journal":{"name":"IEEE Transactions on Sustainable Computing","volume":"9 6","pages":"862-873"},"PeriodicalIF":3.0,"publicationDate":"2024-01-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142810501","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 : 2024-01-12DOI: 10.1109/TSUSC.2024.3353176
Long Cheng;Yan Gu;Qingzhi Liu;Lei Yang;Cheng Liu;Ying Wang
The amalgamation of artificial intelligence with Internet of Things (AIoT) devices have seen a rapid surge in growth, largely due to the effective implementation of deep neural network (DNN) models across various domains. However, the deployment of DNNs on such devices comes with its own set of challenges, primarily related to computational capacity, storage, and energy efficiency. This survey offers an exhaustive review of techniques designed to accelerate DNN inference on AIoT devices, addressing these challenges head-on. We delve into critical model compression techniques designed to adapt to the limitations of devices and hardware optimization strategies that aim to boost efficiency. Furthermore, we examine parallelization methods that leverage parallel computing for swift inference, as well as novel optimization strategies that fine-tune the execution process. This survey also casts a future-forward glance at emerging trends, including advancements in mobile hardware, the co-design of software and hardware, privacy and security considerations, and DNN inference on AIoT devices with constrained resources. All in all, this survey aspires to serve as a holistic guide to advancements in the acceleration of DNN inference on AIoT devices, aiming to provide sustainable computing for upcoming IoT applications driven by artificial intelligence.
{"title":"Advancements in Accelerating Deep Neural Network Inference on AIoT Devices: A Survey","authors":"Long Cheng;Yan Gu;Qingzhi Liu;Lei Yang;Cheng Liu;Ying Wang","doi":"10.1109/TSUSC.2024.3353176","DOIUrl":"https://doi.org/10.1109/TSUSC.2024.3353176","url":null,"abstract":"The amalgamation of artificial intelligence with Internet of Things (AIoT) devices have seen a rapid surge in growth, largely due to the effective implementation of deep neural network (DNN) models across various domains. However, the deployment of DNNs on such devices comes with its own set of challenges, primarily related to computational capacity, storage, and energy efficiency. This survey offers an exhaustive review of techniques designed to accelerate DNN inference on AIoT devices, addressing these challenges head-on. We delve into critical model compression techniques designed to adapt to the limitations of devices and hardware optimization strategies that aim to boost efficiency. Furthermore, we examine parallelization methods that leverage parallel computing for swift inference, as well as novel optimization strategies that fine-tune the execution process. This survey also casts a future-forward glance at emerging trends, including advancements in mobile hardware, the co-design of software and hardware, privacy and security considerations, and DNN inference on AIoT devices with constrained resources. All in all, this survey aspires to serve as a holistic guide to advancements in the acceleration of DNN inference on AIoT devices, aiming to provide sustainable computing for upcoming IoT applications driven by artificial intelligence.","PeriodicalId":13268,"journal":{"name":"IEEE Transactions on Sustainable Computing","volume":"9 6","pages":"830-847"},"PeriodicalIF":3.0,"publicationDate":"2024-01-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142810543","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 : 2024-01-09DOI: 10.1109/TSUSC.2024.3351684
Jie Li;Yuhui Deng;Zhifeng Fan;Zijie Zhong;Geyong Min
The explosion of large-scale data has increased the scale and capacity of storage clusters in data centers, leading to huge power consumption issues. Cloud providers can effectively promote the energy efficiency of data centers by employing energy-aware data placement techniques, which primarily encompass storage cluster's power and cooling power. Traditional data placement approaches do not diminish the overall power consumption of the data center due to the heat recirculation effect between storage nodes. To fill this gap, we build an elaborate thermal-aware data center model. Then we propose two energy-efficient thermal-aware data placement strategies, ETDP-I and ETDP-II, to reduce the overall power consumption of the data center. The principle of our proposed algorithm is to utilize a greedy algorithm to calculate the optimal disk sequence at the minimum total power of the data center and then place the data into the optimal disk sequence. We implement these two strategies in a cloud computing simulation platform based on CloudSim. Experimental results unveil that ETDA-I and ETDP-II outperform MinTin-G and MinTout-G in terms of the supplied temperature of CRAC, storage nodes power, cooling cost, and total power consumption of the data center. In particular, ETDP-I and ETDP-II algorithms can save about 9.46 $%$ -38.93 $%$ of the overall power consumption compared to MinTout-G and MinTin-G algorithms.
{"title":"Towards Energy-Efficient and Thermal-Aware Data Placement for Storage Clusters","authors":"Jie Li;Yuhui Deng;Zhifeng Fan;Zijie Zhong;Geyong Min","doi":"10.1109/TSUSC.2024.3351684","DOIUrl":"https://doi.org/10.1109/TSUSC.2024.3351684","url":null,"abstract":"The explosion of large-scale data has increased the scale and capacity of storage clusters in data centers, leading to huge power consumption issues. Cloud providers can effectively promote the energy efficiency of data centers by employing energy-aware data placement techniques, which primarily encompass storage cluster's power and cooling power. Traditional data placement approaches do not diminish the overall power consumption of the data center due to the heat recirculation effect between storage nodes. To fill this gap, we build an elaborate thermal-aware data center model. Then we propose two energy-efficient thermal-aware data placement strategies, ETDP-I and ETDP-II, to reduce the overall power consumption of the data center. The principle of our proposed algorithm is to utilize a greedy algorithm to calculate the optimal disk sequence at the minimum total power of the data center and then place the data into the optimal disk sequence. We implement these two strategies in a cloud computing simulation platform based on CloudSim. Experimental results unveil that ETDA-I and ETDP-II outperform MinTin-G and MinTout-G in terms of the supplied temperature of CRAC, storage nodes power, cooling cost, and total power consumption of the data center. In particular, ETDP-I and ETDP-II algorithms can save about 9.46\u0000<inline-formula><tex-math>$%$</tex-math></inline-formula>\u0000-38.93\u0000<inline-formula><tex-math>$%$</tex-math></inline-formula>\u0000 of the overall power consumption compared to MinTout-G and MinTin-G algorithms.","PeriodicalId":13268,"journal":{"name":"IEEE Transactions on Sustainable Computing","volume":"9 4","pages":"631-647"},"PeriodicalIF":3.0,"publicationDate":"2024-01-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141965832","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 : 2024-01-09DOI: 10.1109/TSUSC.2024.3351282
Tao Liu;Peng Li;Zhou Su;Mianxiong Dong
Graph neural networks (GNNs) have attracted significant research attention because of their impressive capability in dealing with graph-structure data, such as energy networks, that are crucial for sustainable computing. We find that the communication of data loading from main memory to GPUs is the main bottleneck of GNN inference because of redundant data loading. In this paper, we propose RAIN, an efficient GNN inference system for graph learning. There are two key designs. First, we explore the opportunity of conducting similar inference batches sequentially and reusing repeated nodes among adjacent batches to reduce redundant data loading. This method requires reordering the batches based on their similarity. However, comparing the similarity across a large number of inference batches is a difficult task with a high computational cost. Thus, we propose a local sensitive hash (LSH)-based clustering scheme to group similar batches together quickly without pair-wise comparison. Second, RAIN contains an efficient adaptive sampling strategy, allowing users to sample target nodes’ neighbors according to their degree. The number of sampled neighbors is proportional to the size of the node's degree. We conduct extensive experiments with various baselines. RAIN can achieve up to 6.8X acceleration, and the accuracy decrease is smaller than 0.1%.
{"title":"Efficient Inference of Graph Neural Networks Using Local Sensitive Hash","authors":"Tao Liu;Peng Li;Zhou Su;Mianxiong Dong","doi":"10.1109/TSUSC.2024.3351282","DOIUrl":"https://doi.org/10.1109/TSUSC.2024.3351282","url":null,"abstract":"Graph neural networks (GNNs) have attracted significant research attention because of their impressive capability in dealing with graph-structure data, such as energy networks, that are crucial for sustainable computing. We find that the communication of data loading from main memory to GPUs is the main bottleneck of GNN inference because of redundant data loading. In this paper, we propose RAIN, an efficient GNN inference system for graph learning. There are two key designs. First, we explore the opportunity of conducting similar inference batches sequentially and reusing repeated nodes among adjacent batches to reduce redundant data loading. This method requires reordering the batches based on their similarity. However, comparing the similarity across a large number of inference batches is a difficult task with a high computational cost. Thus, we propose a local sensitive hash (LSH)-based clustering scheme to group similar batches together quickly without pair-wise comparison. Second, RAIN contains an efficient adaptive sampling strategy, allowing users to sample target nodes’ neighbors according to their degree. The number of sampled neighbors is proportional to the size of the node's degree. We conduct extensive experiments with various baselines. RAIN can achieve up to 6.8X acceleration, and the accuracy decrease is smaller than 0.1%.","PeriodicalId":13268,"journal":{"name":"IEEE Transactions on Sustainable Computing","volume":"9 3","pages":"548-558"},"PeriodicalIF":3.9,"publicationDate":"2024-01-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141264414","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 : 2024-01-05DOI: 10.1109/TSUSC.2024.3350343
Yuan Su;Yuheng Wang;Jiliang Li;Zhou Su;Witold Pedrycz;Qinnan Hu
The Public Key Infrastructure (PKI) system is the cornerstone of today’s security communications. All users in the service domain covered by the same PKI system are able to authenticate each other before exchanging messages. However, there is identity isolation in different domains, making the identity of users in different domains cannot be recognized by PKI systems in other domains. To achieve cross-domain authentication, the consortium blockchain system is leveraged in the existing schemes. Unfortunately, the consortium blockchain-based authentication schemes have the following challenges: high cost, privacy concerns, scalability and economic unsustainability. To solve these challenges, we propose a scalable and privacy-preserving cross-domain authentication scheme called Bifrost-Auth. Firstly, Bifrost-Auth is designed to use a decentralized oracle to directly interact with blockchains in different domains instead of maintaining a consortium blockchain and enables mutual authentication for users lying in different domains. Secondly, users can succinctly authenticate their membership of the domain by the accumulator technique, where the membership proof is turned into zero knowledge to protect users’ privacy. Finally, Bifrost-Auth is proven to be secure against various attacks, and thorough experiments are carried out and demonstrate the security and efficiency of Bifrost-Auth.
{"title":"Oracle Based Privacy-Preserving Cross-Domain Authentication Scheme","authors":"Yuan Su;Yuheng Wang;Jiliang Li;Zhou Su;Witold Pedrycz;Qinnan Hu","doi":"10.1109/TSUSC.2024.3350343","DOIUrl":"https://doi.org/10.1109/TSUSC.2024.3350343","url":null,"abstract":"The Public Key Infrastructure (PKI) system is the cornerstone of today’s security communications. All users in the service domain covered by the same PKI system are able to authenticate each other before exchanging messages. However, there is identity isolation in different domains, making the identity of users in different domains cannot be recognized by PKI systems in other domains. To achieve cross-domain authentication, the consortium blockchain system is leveraged in the existing schemes. Unfortunately, the consortium blockchain-based authentication schemes have the following challenges: high cost, privacy concerns, scalability and economic unsustainability. To solve these challenges, we propose a scalable and privacy-preserving cross-domain authentication scheme called Bifrost-Auth. Firstly, Bifrost-Auth is designed to use a decentralized oracle to directly interact with blockchains in different domains instead of maintaining a consortium blockchain and enables mutual authentication for users lying in different domains. Secondly, users can succinctly authenticate their membership of the domain by the accumulator technique, where the membership proof is turned into zero knowledge to protect users’ privacy. Finally, Bifrost-Auth is proven to be secure against various attacks, and thorough experiments are carried out and demonstrate the security and efficiency of Bifrost-Auth.","PeriodicalId":13268,"journal":{"name":"IEEE Transactions on Sustainable Computing","volume":"9 4","pages":"602-614"},"PeriodicalIF":3.0,"publicationDate":"2024-01-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141965768","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：