Pub Date : 2023-12-12DOI: 10.1109/TSUSC.2023.3341836
Yongqi Jiang;Siguang Chen;Xiangwen Bao
In practical applications, federated learning (FL) suffers from slow convergence rate and inferior performance resulting from the statistical heterogeneity of distributed data. Personalized FL (pFL) has been proposed to overcome this problem. However, existing pFL approaches mainly focus on measuring differences between entire model dimensions across clients, ignore the layer-wise differences in convolutional neural networks (CNNs), which may lead to inaccurate personalization. Additionally, two potential threats in FL are that malicious clients may attempt to poison the entire federation by tampering with local labels, and the model information uploaded by clients makes them vulnerable to inference attacks. To tackle these issues, 1) we propose a novel pFL approach in which clients minimize local classification errors and align the local and global prototypes for data from the class that is shared with other clients. This method adopts layer-wise collaborative training to achieve more granular personalization and converts local prototypes to the frequency domain to prevent source data leakage; 2) To prevent the FL model from misclassifying certain test samples as expected by poisoners, we design a robust aggregation method to ensure that benign clients who provide trustworthy model predictions for its local data are weighted far more heavily in the aggregation process than malicious clients. Experiments show that our scheme, especially in the data heterogeneity situation, can produce robust performance and more stable convergence while preserving privacy.
{"title":"Amplitude-Aligned Personalization and Robust Aggregation for Federated Learning","authors":"Yongqi Jiang;Siguang Chen;Xiangwen Bao","doi":"10.1109/TSUSC.2023.3341836","DOIUrl":"https://doi.org/10.1109/TSUSC.2023.3341836","url":null,"abstract":"In practical applications, federated learning (FL) suffers from slow convergence rate and inferior performance resulting from the statistical heterogeneity of distributed data. Personalized FL (pFL) has been proposed to overcome this problem. However, existing pFL approaches mainly focus on measuring differences between entire model dimensions across clients, ignore the layer-wise differences in convolutional neural networks (CNNs), which may lead to inaccurate personalization. Additionally, two potential threats in FL are that malicious clients may attempt to poison the entire federation by tampering with local labels, and the model information uploaded by clients makes them vulnerable to inference attacks. To tackle these issues, 1) we propose a novel pFL approach in which clients minimize local classification errors and align the local and global prototypes for data from the class that is shared with other clients. This method adopts layer-wise collaborative training to achieve more granular personalization and converts local prototypes to the frequency domain to prevent source data leakage; 2) To prevent the FL model from misclassifying certain test samples as expected by poisoners, we design a robust aggregation method to ensure that benign clients who provide trustworthy model predictions for its local data are weighted far more heavily in the aggregation process than malicious clients. Experiments show that our scheme, especially in the data heterogeneity situation, can produce robust performance and more stable convergence while preserving privacy.","PeriodicalId":13268,"journal":{"name":"IEEE Transactions on Sustainable Computing","volume":"9 3","pages":"535-547"},"PeriodicalIF":3.9,"publicationDate":"2023-12-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141264311","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-12-12DOI: 10.1109/TSUSC.2023.3341440
Rui Hao;Xiaohai Dai;Weiqi Dai
Blockchain technology has gained prominence for its potential to address security and privacy challenges in Internet-of-Things (IoT) services and Cyber-Physical Systems (CPS) due to its decentralized, traceable, and immutable nature. However, the considerable energy consumption associated with blockchain, exemplified by Bitcoin, has raised sustainability concerns. This paper introduces BitFT, a consensus protocol that combines the strengths of both lottery-based and voting-based mechanisms to offer a sustainable, comprehensible, and high-performance solution. BitFT dissects the block lifecycle into three phases: dissemination, and commitment phases, which correspond to the Bitcoin framework. It leverages a multiple-round sortition algorithm, a Reliable Broadcast (Rbc) protocol, and a Quorum Certificate (QC) mechanism to facilitate efficient protocol operation. The sortition algorithm functions like a lottery algorithm, while the Rbc protocol and $QC$ mechanism are implemented based on votes. In order to maximize network utilization and enhance system throughput, we further introduce a layered architecture to BitFT, which allows for concurrent commitment of multiple blocks at the same height. We perform a comprehensive analysis to verify the correctness of BitFT and conduct various experiments to demonstrate its high performance.
{"title":"BitFT: An Understandable, Performant and Resource-Efficient Blockchain Consensus","authors":"Rui Hao;Xiaohai Dai;Weiqi Dai","doi":"10.1109/TSUSC.2023.3341440","DOIUrl":"https://doi.org/10.1109/TSUSC.2023.3341440","url":null,"abstract":"Blockchain technology has gained prominence for its potential to address security and privacy challenges in Internet-of-Things (IoT) services and Cyber-Physical Systems (CPS) due to its decentralized, traceable, and immutable nature. However, the considerable energy consumption associated with blockchain, exemplified by Bitcoin, has raised sustainability concerns. This paper introduces BitFT, a consensus protocol that combines the strengths of both lottery-based and voting-based mechanisms to offer a sustainable, comprehensible, and high-performance solution. BitFT dissects the block lifecycle into three phases: dissemination, and commitment phases, which correspond to the Bitcoin framework. It leverages a multiple-round sortition algorithm, a Reliable Broadcast (Rbc) protocol, and a Quorum Certificate (QC) mechanism to facilitate efficient protocol operation. The sortition algorithm functions like a lottery algorithm, while the \u0000<small>Rbc</small>\u0000 protocol and \u0000<inline-formula><tex-math>$QC$</tex-math></inline-formula>\u0000 mechanism are implemented based on votes. In order to maximize network utilization and enhance system throughput, we further introduce a layered architecture to BitFT, which allows for concurrent commitment of multiple blocks at the same height. We perform a comprehensive analysis to verify the correctness of BitFT and conduct various experiments to demonstrate its high performance.","PeriodicalId":13268,"journal":{"name":"IEEE Transactions on Sustainable Computing","volume":"9 3","pages":"522-534"},"PeriodicalIF":3.9,"publicationDate":"2023-12-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141264515","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-12-08DOI: 10.1109/TSUSC.2023.3308471
Paul D. Yoo;Zahir Tari
In an increasingly interconnected world, the sophistication of cyber-attacks is on the rise. Cybersecurity research stands as a pivotal factor in shaping the prosperity of nations. To counter threats to network infrastructure and sensitive data, a multitude of security solutions with varying degrees of efficacy have been proposed. However, these solutions have thus far insufficiently accounted for a critical dimension: sustainability. In this context, sustainability entails the continuous support of processes over time by enhancing the computational requisites, scalability, energy efficiency, and resource utilization of defence and security systems. This special issue endeavors to explore recent strides in model development, innovative methodologies, and insightful observations aimed at enhancing cybersecurity, with a particular emphasis on the sustainability of defence and security systems.
{"title":"Editorial Sustainable Defence and Security Systems","authors":"Paul D. Yoo;Zahir Tari","doi":"10.1109/TSUSC.2023.3308471","DOIUrl":"https://doi.org/10.1109/TSUSC.2023.3308471","url":null,"abstract":"In an increasingly interconnected world, the sophistication of cyber-attacks is on the rise. Cybersecurity research stands as a pivotal factor in shaping the prosperity of nations. To counter threats to network infrastructure and sensitive data, a multitude of security solutions with varying degrees of efficacy have been proposed. However, these solutions have thus far insufficiently accounted for a critical dimension: sustainability. In this context, sustainability entails the continuous support of processes over time by enhancing the computational requisites, scalability, energy efficiency, and resource utilization of defence and security systems. This special issue endeavors to explore recent strides in model development, innovative methodologies, and insightful observations aimed at enhancing cybersecurity, with a particular emphasis on the sustainability of defence and security systems.","PeriodicalId":13268,"journal":{"name":"IEEE Transactions on Sustainable Computing","volume":"8 4","pages":"537-539"},"PeriodicalIF":3.9,"publicationDate":"2023-12-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10352168","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138558207","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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}
Pub Date : 2023-11-29DOI: 10.1109/TSUSC.2023.3336691
Kun Cao;Jian Weng
In recent years, serverless edge computing has been widely employed in the deployments of Internet-of-things (IoT) applications. Despite considerable research efforts in this field, existing works fail to jointly consider essential factors such as energy, reliability, personalized user requirements, and stochastic application executions. This oversight results in an inefficient utilization of computation and communication resources within serverless edge computing networks, subsequently diminishing the profit of service providers and degrading the quality-of-experience (QoE) of end users. In this paper, we explore the problem of reliability-ensured personalized function scheduling (REPFS) to jointly optimize the profit of service providers and the holistic QoE of end users in sustainable serverless edge computing. A personality-driven user QoE prediction method is first designed to accurately estimate the QoE of individual end users with differentiated personality types. Afterward, a deterministic function scheduling policy is developed on the problem-specific augmented non-dominated sorting genetic algorithm II (PSA-NSGA-II). Given the inherent uncertainty of application executions, a stochastic function scheduling strategy that can be easily parallelized for modern multicore scheduler platforms is also devised to accelerate solution generation for stochastic applications. Experimental results show that our deterministic function scheduling policy achieves 15% performance enhancement compared with representative multiobjective evolutionary algorithms. Furthermore, our stochastic function scheduling strategy promotes the service profit by 78% and the holistic user QoE by 118% on average compared with the developed deterministic scheduling policy.
{"title":"REPFS: Reliability-Ensured Personalized Function Scheduling in Sustainable Serverless Edge Computing","authors":"Kun Cao;Jian Weng","doi":"10.1109/TSUSC.2023.3336691","DOIUrl":"https://doi.org/10.1109/TSUSC.2023.3336691","url":null,"abstract":"In recent years, serverless edge computing has been widely employed in the deployments of Internet-of-things (IoT) applications. Despite considerable research efforts in this field, existing works fail to jointly consider essential factors such as energy, reliability, personalized user requirements, and stochastic application executions. This oversight results in an inefficient utilization of computation and communication resources within serverless edge computing networks, subsequently diminishing the profit of service providers and degrading the quality-of-experience (QoE) of end users. In this paper, we explore the problem of reliability-ensured personalized function scheduling (REPFS) to jointly optimize the profit of service providers and the holistic QoE of end users in sustainable serverless edge computing. A personality-driven user QoE prediction method is first designed to accurately estimate the QoE of individual end users with differentiated personality types. Afterward, a deterministic function scheduling policy is developed on the problem-specific augmented non-dominated sorting genetic algorithm II (PSA-NSGA-II). Given the inherent uncertainty of application executions, a stochastic function scheduling strategy that can be easily parallelized for modern multicore scheduler platforms is also devised to accelerate solution generation for stochastic applications. Experimental results show that our deterministic function scheduling policy achieves 15% performance enhancement compared with representative multiobjective evolutionary algorithms. Furthermore, our stochastic function scheduling strategy promotes the service profit by 78% and the holistic user QoE by 118% on average compared with the developed deterministic scheduling policy.","PeriodicalId":13268,"journal":{"name":"IEEE Transactions on Sustainable Computing","volume":"9 3","pages":"494-511"},"PeriodicalIF":3.9,"publicationDate":"2023-11-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10334010","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141264524","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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}
Pub Date : 2023-11-27DOI: 10.1109/TSUSC.2023.3336801
Stephen Clement;Kat Burdett;Nour Rteil;Astrid Wynne;Rich Kenny
As demand for digital services grows, there is need to improve efficiency and reduce the environmental impact of data centers. The largest energy consumer in any data center is the IT, followed by the systems dedicated to cooling. Aiming to improve efficiency, and driven by metrics like PUE, there is a trend towards running data centers hotter to reduce the cooling energy. There is little research investigating the effect this will have on the IT beyond failure rates. To ensure overall efficiency is improving, we must view the data center as a system of systems, taking a holistic view rather than focusing on individual sub-systems. In this paper we use industry standard benchmarks and a wind-tunnel to profile typical enterprise IT. We analyze the effect of environmental conditions on IT efficiency, showing minor increases in temperature or pressure impact the efficiency of servers. Using an idealized, simulated data center case study we show that the interaction between cooling systems, server behavior and local climate are non-trivial and increasing temperatures has potential to worsen efficiency.
随着数字服务需求的增长，需要提高数据中心的效率并减少对环境的影响。数据中心最大的能源消耗者是 IT，其次是专用于冷却的系统。为了提高效率，在 PUE 等指标的驱动下，数据中心的运行温度有升高的趋势，以减少冷却能耗。除故障率外，几乎没有研究调查这对 IT 的影响。为确保提高整体效率，我们必须将数据中心视为一个由多个系统组成的系统，从整体上而不是只关注单个子系统。在本文中，我们使用行业标准基准和风洞来剖析典型的企业 IT。我们分析了环境条件对 IT 效率的影响，表明温度或压力的微小增加都会影响服务器的效率。通过一个理想化的模拟数据中心案例研究，我们发现冷却系统、服务器行为和当地气候之间的相互作用并不复杂，温度升高有可能导致效率恶化。
{"title":"Is Hot IT a False Economy? An Analysis of Server and Data Center Energy Efficiency as Temperatures Rise","authors":"Stephen Clement;Kat Burdett;Nour Rteil;Astrid Wynne;Rich Kenny","doi":"10.1109/TSUSC.2023.3336801","DOIUrl":"https://doi.org/10.1109/TSUSC.2023.3336801","url":null,"abstract":"As demand for digital services grows, there is need to improve efficiency and reduce the environmental impact of data centers. The largest energy consumer in any data center is the IT, followed by the systems dedicated to cooling. Aiming to improve efficiency, and driven by metrics like PUE, there is a trend towards running data centers hotter to reduce the cooling energy. There is little research investigating the effect this will have on the IT beyond failure rates. To ensure overall efficiency is improving, we must view the data center as a system of systems, taking a holistic view rather than focusing on individual sub-systems. In this paper we use industry standard benchmarks and a wind-tunnel to profile typical enterprise IT. We analyze the effect of environmental conditions on IT efficiency, showing minor increases in temperature or pressure impact the efficiency of servers. Using an idealized, simulated data center case study we show that the interaction between cooling systems, server behavior and local climate are non-trivial and increasing temperatures has potential to worsen efficiency.","PeriodicalId":13268,"journal":{"name":"IEEE Transactions on Sustainable Computing","volume":"9 3","pages":"482-493"},"PeriodicalIF":3.9,"publicationDate":"2023-11-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141264312","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}
In this paper, we study the performance analysis of a cooperative terrestrial-underwater visible light communication (UWVLC) system, considering energy harvesting at a decode and forward (DF) relay on the ship. The transmitter to the ship link is modelled by shadowed Rician fading channel, while the UWVLC link is characterised by mixture exponential generalized gamma (EGG) distribution. The energy constrained relay harvests energy from the received signal based on power splitting (PS) scheme, assuming a part of received power is used for energy harvesting. Energy harvesting at the relay node makes the system sustainable by reducing the carbon footprints. We derive novel closed form expressions for bit error rate (BER), outage probability and ergodic capacity for the considered system at the underwater node, considering the effects of various parameters that include bubble level, temperature and salinity.
{"title":"Performance Analysis of Hybrid RF/VLC Energy Harvested Terrestrial-Underwater System","authors":"Krati Mittal;Akash Gupta;Nikhil Sharma;Manan Jani;Parul Garg","doi":"10.1109/TSUSC.2023.3336374","DOIUrl":"https://doi.org/10.1109/TSUSC.2023.3336374","url":null,"abstract":"In this paper, we study the performance analysis of a cooperative terrestrial-underwater visible light communication (UWVLC) system, considering energy harvesting at a decode and forward (DF) relay on the ship. The transmitter to the ship link is modelled by shadowed Rician fading channel, while the UWVLC link is characterised by mixture exponential generalized gamma (EGG) distribution. The energy constrained relay harvests energy from the received signal based on power splitting (PS) scheme, assuming a part of received power is used for energy harvesting. Energy harvesting at the relay node makes the system sustainable by reducing the carbon footprints. We derive novel closed form expressions for bit error rate (BER), outage probability and ergodic capacity for the considered system at the underwater node, considering the effects of various parameters that include bubble level, temperature and salinity.","PeriodicalId":13268,"journal":{"name":"IEEE Transactions on Sustainable Computing","volume":"9 3","pages":"512-521"},"PeriodicalIF":3.9,"publicationDate":"2023-11-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141264335","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-11-13DOI: 10.1109/TSUSC.2023.3331179
Fan Yang;Xiaofeng Liao;Xinyu Lei;Nankun Mu;Di Zhang
Data trading is an effective way for commercial companies to obtain massive personal data to develop their data-driven businesses. However, when data owners may want to sell their data without revealing privacy, data consumers also face the dilemma of high purchase costs due to purchasing too much invalid data. Therefore, there is an urgent need for a data trading scheme that can protect personal privacy and save expenses simultaneously. In this paper, we design a priv AC y-preserving and pra C tical aggr E gate S tati S tic trading scheme (named as ACCESS). Technically, we focus on the group-level pricing strategy to make ACCESS easier to implement. The differential privacy technique is applied to protect the data owners’ privacy, and the sampling algorithm is adopted to reduce the data consumers’ costs. Specifically, to provide a maximum tolerant privacy loss guarantee for the data owners, we design a decision algorithm to detect whether a conflict occurs between the consumer-specified accuracy level and the maximum tolerable privacy loss budget. Besides, to minimize the purchase cost for the data brokers, we develop a sampling-based aggregation method consisting of two sampling algorithms (called as BUSA and BKSA, respectively). BUSA enables reducing purchase costs with no additional background knowledge. Once the data broker knows the data boundary, BKSA can significantly reduce the amount of data that needs to be purchased, thereby the purchase cost is reduced. Rigorous theoretical analysis and extensive experiments (over four real-world and public datasets) further demonstrate the practicability of ACCESS.
{"title":"Towards Privacy-Preserving and Practical Data Trading for Aggregate Statistic","authors":"Fan Yang;Xiaofeng Liao;Xinyu Lei;Nankun Mu;Di Zhang","doi":"10.1109/TSUSC.2023.3331179","DOIUrl":"10.1109/TSUSC.2023.3331179","url":null,"abstract":"Data trading is an effective way for commercial companies to obtain massive personal data to develop their data-driven businesses. However, when data owners may want to sell their data without revealing privacy, data consumers also face the dilemma of high purchase costs due to purchasing too much invalid data. Therefore, there is an urgent need for a data trading scheme that can protect personal privacy and save expenses simultaneously. In this paper, we design a priv\u0000<bold>AC</b>\u0000y-preserving and pra\u0000<bold>C</b>\u0000tical aggr\u0000<bold>E</b>\u0000gate \u0000<bold>S</b>\u0000tati\u0000<bold>S</b>\u0000tic trading scheme (named as ACCESS). Technically, we focus on the group-level pricing strategy to make ACCESS easier to implement. The differential privacy technique is applied to protect the data owners’ privacy, and the sampling algorithm is adopted to reduce the data consumers’ costs. Specifically, to provide a maximum tolerant privacy loss guarantee for the data owners, we design a decision algorithm to detect whether a conflict occurs between the consumer-specified accuracy level and the maximum tolerable privacy loss budget. Besides, to minimize the purchase cost for the data brokers, we develop a sampling-based aggregation method consisting of two sampling algorithms (called as BUSA and BKSA, respectively). BUSA enables reducing purchase costs with no additional background knowledge. Once the data broker knows the data boundary, BKSA can significantly reduce the amount of data that needs to be purchased, thereby the purchase cost is reduced. Rigorous theoretical analysis and extensive experiments (over four real-world and public datasets) further demonstrate the practicability of ACCESS.","PeriodicalId":13268,"journal":{"name":"IEEE Transactions on Sustainable Computing","volume":"9 3","pages":"452-463"},"PeriodicalIF":3.9,"publicationDate":"2023-11-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135610928","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-11-06DOI: 10.1109/TSUSC.2023.3330671
José Miguel Aragón-Jurado;Juan Carlos de la Torre;Patricia Ruiz;Pedro L. Galindo;Albert Y. Zomaya;Bernabé Dorronsoro
Efficient green software solutions require being aware of the characteristics of both the software and the hardware where it is executed. Separately optimizing them leads to inefficient results, and there is a need for a perfect synergy between software and hardware for optimal outcomes. We present a novel combinatorial optimization problem for the minimization of the software execution time on a specific hardware, taking into account the existing uncertainty in the system. A solution to the problem is a sequence of LLVM code transformations, and a cellular genetic algorithm is used to find it. Assuming that hardware does not change, reducing the software runtime typically leads to a greener version with lower consumption. To cope with the uncertainty, two novel approaches relying on bootstrap method to compute confident intervals of the software runtime at negligible cost are proposed and compared to three other techniques and −O3 Clang compilation flag over four hardware architectures. Results show how the proposed approach effectively copes with the uncertainty, providing more robust solutions with respect to the compared methods. The execution time of the raw program is reduced from 28.1% to up to 63.2%, outperforming −O3 flag by 13.9% to 26.3%, for the different architectures.
{"title":"Automatic Software Tailoring for Optimal Performance","authors":"José Miguel Aragón-Jurado;Juan Carlos de la Torre;Patricia Ruiz;Pedro L. Galindo;Albert Y. Zomaya;Bernabé Dorronsoro","doi":"10.1109/TSUSC.2023.3330671","DOIUrl":"10.1109/TSUSC.2023.3330671","url":null,"abstract":"Efficient green software solutions require being aware of the characteristics of both the software and the hardware where it is executed. Separately optimizing them leads to inefficient results, and there is a need for a perfect synergy between software and hardware for optimal outcomes. We present a novel combinatorial optimization problem for the minimization of the software execution time on a specific hardware, taking into account the existing uncertainty in the system. A solution to the problem is a sequence of LLVM code transformations, and a cellular genetic algorithm is used to find it. Assuming that hardware does not change, reducing the software runtime typically leads to a greener version with lower consumption. To cope with the uncertainty, two novel approaches relying on bootstrap method to compute confident intervals of the software runtime at negligible cost are proposed and compared to three other techniques and −O3 Clang compilation flag over four hardware architectures. Results show how the proposed approach effectively copes with the uncertainty, providing more robust solutions with respect to the compared methods. The execution time of the raw program is reduced from 28.1% to up to 63.2%, outperforming −O3 flag by 13.9% to 26.3%, for the different architectures.","PeriodicalId":13268,"journal":{"name":"IEEE Transactions on Sustainable Computing","volume":"9 3","pages":"464-481"},"PeriodicalIF":3.9,"publicationDate":"2023-11-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135501605","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}
This paper introduces a novel parallel trajectory mechanism that combines Levenberg-Marquardt and Forward Accumulation Through Time algorithms to train a recurrent neural network controller in a closed-loop control system by distributing the calculation of trajectories across Central Processing Unit (CPU) cores/workers depending on the computing platforms, computing program languages, and software packages available. Without loss of generality, the recurrent neural network controller of a grid-connected converter for solar integration to a power system was selected as the benchmark test closed-loop control system. Two software packages were developed in Matlab and C++ to verify and demonstrate the efficiency of the proposed parallel training method. The training of the deep neural network controller was migrated from a single workstation to both cloud computing platforms and High-Performance Computing clusters. The training results show excellent speed-up performance, which significantly reduces the training time for a large number of trajectories with high sampling frequency, and further demonstrates the effectiveness and scalability of the proposed parallel mechanism.
本文介绍了一种新颖的并行轨迹机制，该机制结合了 Levenberg-Marquardt 算法和时间前向累加算法，可在闭环控制系统中训练递归神经网络控制器，方法是根据计算平台、计算程序语言和可用软件包的不同，将轨迹计算分配到中央处理器（CPU）内核/处理器上。在不失一般性的前提下，我们选择了用于太阳能并入电力系统的并网转换器的递归神经网络控制器作为闭环控制系统的基准测试。为了验证和展示所提出的并行训练方法的效率，我们用 Matlab 和 C++ 开发了两个软件包。深度神经网络控制器的训练从单个工作站迁移到云计算平台和高性能计算集群。训练结果显示了优异的加速性能，在大量高采样频率轨迹的情况下显著缩短了训练时间，进一步证明了所提出的并行机制的有效性和可扩展性。
{"title":"Parallel Trajectory Training of Recurrent Neural Network Controllers With Levenberg–Marquardt and Forward Accumulation Through Time in Closed-Loop Control Systems","authors":"Xingang Fu;Jordan Sturtz;Eduardo Alonso;Rajab Challoo;Letu Qingge","doi":"10.1109/TSUSC.2023.3330573","DOIUrl":"10.1109/TSUSC.2023.3330573","url":null,"abstract":"This paper introduces a novel parallel trajectory mechanism that combines Levenberg-Marquardt and Forward Accumulation Through Time algorithms to train a recurrent neural network controller in a closed-loop control system by distributing the calculation of trajectories across Central Processing Unit (CPU) cores/workers depending on the computing platforms, computing program languages, and software packages available. Without loss of generality, the recurrent neural network controller of a grid-connected converter for solar integration to a power system was selected as the benchmark test closed-loop control system. Two software packages were developed in Matlab and C++ to verify and demonstrate the efficiency of the proposed parallel training method. The training of the deep neural network controller was migrated from a single workstation to both cloud computing platforms and High-Performance Computing clusters. The training results show excellent speed-up performance, which significantly reduces the training time for a large number of trajectories with high sampling frequency, and further demonstrates the effectiveness and scalability of the proposed parallel mechanism.","PeriodicalId":13268,"journal":{"name":"IEEE Transactions on Sustainable Computing","volume":"9 2","pages":"222-229"},"PeriodicalIF":3.9,"publicationDate":"2023-11-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135501371","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-11-03DOI: 10.1109/TSUSC.2023.3329995
Bin Cai;Weihong Sheng;Jiajun Chen;Chunqiang Hu;Jiguo Yu
The growing prevalence of graphs representations in our society has led to a corresponding rise in the publishing of graphs by researchers and organizations. To protect the privacy, it is important to ensure that graphs including sensitive data are not disclosed. Since the weight of edges could be utilized to infer confidential information, the graph should be privately published to avoid ethical and legal issues. In this paper, we propose a novel method for privately publishing shortest paths while preserving the privacy of sensitive edge weights in graph. Specifically, we divide the edge weights into internal and external edges based on their edge betweenness centrality. Then, we give two different differentially private algorithms to perturb edge weights based on the distinction between internal and external edges, respectively. To reduce the error ratios between differentially private shortest paths and real shortest paths, we employ edge betweenness centrality to search for the shortest path, which is closest to the true one. Our experimental results show that our mechanisms can effectively reduce the error in the average shortest path distance by 1.1% for large graphs, while for the shortest path change rate, our mechanisms can reduce it by 8.3%.
{"title":"Shortest Paths Publishing With Differential Privacy","authors":"Bin Cai;Weihong Sheng;Jiajun Chen;Chunqiang Hu;Jiguo Yu","doi":"10.1109/TSUSC.2023.3329995","DOIUrl":"10.1109/TSUSC.2023.3329995","url":null,"abstract":"The growing prevalence of graphs representations in our society has led to a corresponding rise in the publishing of graphs by researchers and organizations. To protect the privacy, it is important to ensure that graphs including sensitive data are not disclosed. Since the weight of edges could be utilized to infer confidential information, the graph should be privately published to avoid ethical and legal issues. In this paper, we propose a novel method for privately publishing shortest paths while preserving the privacy of sensitive edge weights in graph. Specifically, we divide the edge weights into internal and external edges based on their edge betweenness centrality. Then, we give two different differentially private algorithms to perturb edge weights based on the distinction between internal and external edges, respectively. To reduce the error ratios between differentially private shortest paths and real shortest paths, we employ edge betweenness centrality to search for the shortest path, which is closest to the true one. Our experimental results show that our mechanisms can effectively reduce the error in the average shortest path distance by 1.1% for large graphs, while for the shortest path change rate, our mechanisms can reduce it by 8.3%.","PeriodicalId":13268,"journal":{"name":"IEEE Transactions on Sustainable Computing","volume":"9 2","pages":"209-221"},"PeriodicalIF":3.9,"publicationDate":"2023-11-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134982272","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：