检索结果-内蒙古大学图书馆

23rd International Conference on Computational Science and its Applications (ICCSA)

作者： Bodkhe, Sonali Mahajan, Anjali Sant Gadge Baba Amravati Univ Dept Comp Sci Amravati Maharashtra India Govt Polytech Nagpur Maharashtra India

ISBN: (纸本)9783031368042;9783031368059

Parallel computing and distributed computing are the popular terminologies of scheduling. With advancement in technology, systems have become much more compact and fast and need of parallelization plays a major role for this compaction. Wireless computing is also a common concept associated with each new development. Scheduling of tasks has always been a challenging area and is an NP-complete problem. Moreover, when it comes to wireless distributed computing, reliable scheduling plays an important role in order to complete a task in a wireless distributed system. This work proposes an algorithm to dynamically schedule tasks on heterogeneous processors within a wireless distributed computing system. A lot of heuristics, meta-heuristics & genetics have been used earlier with scheduling strategies. However, most of them haven't taken reliability into account before scheduling. Here a heuristic that deals with reliable scheduling is considered. The scheduler also works within an environment which has dynamically changing resources and adapts itself to changing system resources. The testing was carried out with up to 200 tasks being scheduled while testing in a real time wireless distributed environment. Experiments have shown that the algorithm outperforms the other strategies and can achieve a better reliability along with no increase in make-span, in spite of wireless nodes.

关键词： distributed computing Scheduling Reliability Wireless distributed Systems (WDS)

来源：评论

学校读者我要写书评

暂无评论

Efficient, Failure Resilient Transactions for Parallel and distributed computing

Efficient, Failure Resilient Transactions for Parallel and D...

引用

DISCS The International Workshop on Data-Intensive Scalable computing Systems

作者： Lofstead, Jay Dayal, Jai Jimenez, Ivo Maltzahn, Carlos Sandia Natl Labs Livermore CA 94550 USA Georgia Tech Atlanta GA USA Univ Calif Santa Cruz Santa Cruz CA 95064 USA

ISBN: (纸本)9781479970384

Scientific simulations are moving away from using centralized persistent storage for intermediate data between workflow steps towards an all online model. This shift is motivated by the relatively slow IO bandwidth growth compared with compute speed increases. The challenges presented by this shift to Integrated Application Workflows are motivated by the loss of persistent storage semantics for node-to-node communication. One step towards addressing this semantics gap is using transactions to logically delineate a data set from 100,000s of processes to 1000s of servers as an atomic unit. Our previously demonstrated Doubly distributed Transactions ((DT)-T-2) protocol showed a high-performance solution, but had not explored how to detect and recover from faults. Instead, the focus was on demonstrating high-performance typical case performance. The research presented here addresses fault detection and recovery based on the enhanced protocol design. The total overhead for a full transaction with multiple operations at 65,536 processes is on average 0.055 seconds. Fault detection and recovery mechanisms demonstrate similar performance to the success case with only the addition of appropriate timeouts for the system. This paper explores the challenges in designing a recoverable protocol for doubly distributed transactions, particularly for parallel computing environments.

关键词： distributed computing Parallel Lines Defect detection on-line model RECOVERY MECHANISM PARALLEL PROCESSING (COMPUTERS) Workflow Servers Shift Semantics semiotics protocols MULTIPLE HEAD OPERATIONS

来源：评论

学校读者我要写书评

暂无评论

Africanus II. QUARTICAL: Calibrating radio interferometer data at scale using NUMBA and DASK

引用

ASTRONOMY AND computing 2025年 52卷

作者： Kenyon, J. S. Perkins, S. J. Bester, H. L. Smirnov, O. M. Russeeawon, C. Hugo, B. V. Rhodes Univ Ctr Radio Astron Tech & Technol RATT Dept Phys & Elect Makhanda EC South Africa South African Radio Astron Observ SARAO Cape Town WC South Africa Natl Inst Astrophys INAF IRA Inst Radioastron Bologna Italy

Calibration is, and will remain, an integral component of radio interferometric data reduction. However, as larger, more sensitive radio interferometers are conceived and built, the calibration problem grows in both size and difficulty. The increasing size can be attributed to the fact that the data volume scales quadratically with the number of antennas in an array. Additionally, new instruments may have up to two orders of magnitude more channels than their predecessors. Simultaneously, increasing sensitivity is making calibration more challenging: low-level RFI and calibration artefacts (in the resulting images) which would previously have been subsumed by the noise may now limit dynamic range and, ultimately, the derived science. It is against this backdrop that we introduce QUARTICAL: a new Python package implementing radio interferometric calibration routines. QUARTICAL improves upon its predecessor, CUBICAL, in terms of both flexibility and performance. Whilst the same mathematical framework-complex optimization using Wirtinger derivatives-is in use, the approach has been refined to support arbitrary length chains of parameterized gain terms. QUARTICAL utilizes DASK, a library for parallel computing in Python, to express calibration as an embarrassingly parallel task graph. These task graphs can (with some constraints) be mapped onto a number of different hardware configurations, allowing QUARTICAL to scale from running locally on consumer hardware to a distributed, cloud-based cluster. QUARTICAL's qualitative behaviour is demonstrated using MeerKAT observations of PSR J2009-2026. These qualitative results are followed by an analysis of QUARTICAL's performance in terms of wall time and memory footprint for a number of calibration scenarios and hardware configurations.

关键词： Radio astronomy Calibration Software distributed computing Numerical methods

来源：评论

学校读者我要写书评

暂无评论

Framework for the distributed computing of the Application Components 9

Framework for the Distributed Computing of the Application C...

引用

9th International Conference on Dependability and Complex Systems (DepCoS-RELCOMEX)

作者： Aciu, Razvan-Mihai Ciocarlie, Horia Politehn Univ Timisoara Dept Comp & Software Engn Timisoara 300223 Romania

ISBN: (纸本)9783319070131

Writing real world distributed applications is a challenging task. Even if well known models or powerful frameworks such as MapReduce or HADOOP are employed, the complexity of the aspects involved, such as specific programming and data models, deployment scripts or a hard debugging process are enough to require many working hours or even make the entire process unsuitable for practical purposes. For applications which need some of their own components to be computed in a distributed manner, a generic model incurs an unnecessary overhead and makes the whole development slower. We propose a MapReduce framework which automatically handles all the distributed computing tasks such as computing resources abstraction, code deployment, objects serialization, remote invocations and synchronizations with only a minimal coding overhead. With only minimal constructions dependable distributed components can be developed and run on heterogeneous platforms and networks. The presented results confirm the performance of the proposed method.

关键词： distributed computing serialization synchronization invocation

来源：评论

学校读者我要写书评

暂无评论

GHPFL: Advancing Personalized Edge-Based Learning Through Optimized Bandwidth Utilization

引用

IEEE TRANSACTIONS ON CLOUD computing 2025年第2期13卷 473-484页

作者： Mo, Kaiwei Lin, Wei Lu, Jiaxun Xue, Chun Jason Shao, Yunfeng Xu, Hong City Univ Hong Kong Hong Kong Peoples R China Chinese Univ Hong Kong Hong Kong Peoples R China Huawei Technol Beijing 100077 Peoples R China

Federated learning (FL) is increasingly adopted to combine knowledge from clients in training without revealing their private data. In order to improve the performance of different participants, personalized FL has recently been proposed. However, considering the non-independent and identically distributed (non-IID) data and limited bandwidth at clients, the model performance could be compromised. In reality, clients near each other often tend to have similar data distributions. In this work, we train the personalized edge-based model in the client-edge-server FL. While considering the differences in data distribution, we fully utilize the limited bandwidth resources. To make training efficient and accurate at the same time, An intuitive idea is to learn as much useful knowledge as possible from other edges and reduce the accuracy loss incurred by non-IID data. Therefore, we devise Grouping Hierarchical Personalized Federated Learning (GHPFL). In this framework, each edge establishes physical connections with multiple clients, while the server physically connects with edges. It clusters edges into groups and establishes client-edge logical connections for synchronization. This is based on data similarities that the nodes actively identify, as well as the underlying physical topology. We perform a large-scale evaluation to demonstrate GHPFL's benefits over other schemes.

关键词： Federated learning Servers Training Data models Accuracy Bandwidth Cloud computing Adaptation models Load modeling Interpolation distributed algorithms distributed computing federated learning

来源：评论

学校读者我要写书评

暂无评论

DBP: distributed Power Budgeting for Many-Core Systems in Dark Silicon

引用

IEEE TRANSACTIONS ON COMPUTER-AIDED DESIGN OF INTEGRATED CIRCUITS AND SYSTEMS 2022年第12期41卷 5727-5731页

作者： Wang, Hai He, Wenjun Yang, Qinhui Peng, Xizhu Tang, He Univ Elect Sci & Technol China Sch Elect Sci & Engn Chengdu 610054 Peoples R China Univ Elect Sci & Technol China Yangtze Delta Reg Inst Huzhou Huzhou 313000 Peoples R China

Power budget is an important power constraint provided to guarantee the thermal reliability of an integrated system. In this work, we present DBP, a distributed power budgeting method, for dark silicon many-core systems. In DBP, there are two new techniques proposed to bring accurate and optimized power budgets in a distributed way. First, a distributed active core locating technique is developed to find an active core distribution that leads to a high-power budget. Second, a distributed power budget computing technique is introduced which computes the power budget for each active core accurately. Experiments show DBP outperforms the state-of-the-art power budgeting methods' thermal safe power (TSP) and greedy dynamic power (GDP) on many-core dark silicon systems by providing a high and accurate power budget with low overhead and good scalability.

关键词： distributed computing many-core system power budget thermal modeling

来源：评论

学校读者我要写书评

暂无评论

A Decentralized Authorization and Security Framework for distributed Research Workflows 47

A Decentralized Authorization and Security Framework for Dis...

引用

47th IEEE-Computer-Society Annual International Conference on Computers, Software, and Applications (COMPSAC)

作者： Cardone, Richard Padhy, Smruti Black, Steve Stubbs, Joe Cleveland, Sean Univ Texas Austin Texas Adv Comp Ctr Austin TX 78712 USA Univ Hawaii Syst Informat Technol Serv Honolulu HI USA

ISBN: (纸本)9798350326970

Research challenges such as climate change and the search for habitable planets increasingly use academic and commercial computing resources distributed across different institutions and physical sites. Furthermore, such analyses often require a level of automation that precludes direct human interaction, and securing these workflows involves adherence to security policies across institutions. In this paper, we present a decentralized authorization and security framework that enables researchers to utilize resources across different sites while allowing service providers to maintain autonomy over their secrets and authorization policies. We describe this framework as part of the Tapis platform, a web-based, hosted API used by researchers from multiple institutions.

关键词： distributed computing authentication authorization HPC microservices middleware

来源：评论

学校读者我要写书评

暂无评论

Split Process Cluster: A distributed computing Platform for Edge Extraction of Massive Remote Sensing Images

Split Process Cluster: A Distributed Computing Platform for ...

引用

International Conference on Advances in Materials Science and Information Technologies in Industry (AMSITI)

作者： Cheng, Fuchao Miao, Fang Yang, Wenhui Chengdu Univ Technol Coll Geophys Chengdu Peoples R China Chengdu Univ Technol Sch Informat Sci & Technol Chengdu Peoples R China

ISBN: (纸本)9783038350125

In existed distributed edge extraction method based on MapReduce, the inappropriate dataset split algorithms leaded to the loss problem of image features in result. We presented a distributed computing platform called Split Process Cluster (SPC) to resolve this problem. In SPC, the images are partitioned with the resilient image pyramid model (RIP), a multi-layer and redundant data structure we presented earlier, to ensure the integrity of original image features. And SPC packages the image data to the form of Key-Value pairs, which could be processed through Hadoop, and reduces the results with density-based spatial clustering of applications with noise (DBSCAN) algorithm. Compared to traditional method, the extraction rate of image feature by using SPC has been improved, which indicates that using SPC is an efficient way to improve the extraction rate of distributed edge extraction.

关键词： edge extraction image pyramid distributed computing integrity mapreduce

来源：评论

学校读者我要写书评

暂无评论

On Secure distributed Linearly Separable Computation

引用

IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS 2022年第3期40卷 912-926页

作者： Wan, Kai Sun, Hua Ji, Mingyue Caire, Giuseppe Tech Univ Berlin Elect Engn & Comp Sci Dept D-10587 Berlin Germany Univ North Texas Dept Elect Engn Denton TX 76203 USA Univ Utah Dept Elect & Comp Engn Salt Lake City UT 84112 USA

distributed linearly separable computation, where a user asks some distributed servers to compute a linearly separable function, was recently formulated by the same authors and aims to alleviate the bottlenecks of stragglers and communication cost in distributed computation. The data center assigns a subset of input datasets to each server in an uncoded manner, and each server computes some coded packets on the assigned datasets, which are then sent to the user. The user should recover the task function from the answers of a subset of servers, such that the effect of stragglers could be tolerated. In this paper, we formulate a novel secure framework for this distributed linearly separable computation, where we aim to let the user only retrieve the desired task function without obtaining any other information about the input datasets, even if it receives the answers of all servers. In order to preserve the security of the input datasets, some common randomness variable independent of the datasets should be introduced into the transmission. We show that any non-secure linear-coding based computing scheme for the original distributed linearly separable computation problem, can be made secure without increasing the communication cost (number of symbols the user should receive). Then we focus on the case where the computation cost of each server (number of datasets assigned to each server) is minimum and aim to minimize the size of the randomness variable (i.e., randomness size) introduced in the system while achieving the optimal communication cost. We first propose an information theoretic converse bound on the randomness size. We then propose secure computing schemes based on two well-known data assignments, namely fractional repetition assignment and cyclic assignment. These schemes are optimal subject to using these assignments. Motivated by the observation of the general limitation of these two schemes on the randomness size, we propose a computing scheme with nove

关键词： Servers Costs Task analysis Security Encoding Data centers Sun distributed computing security straggler mitigation

来源：评论

学校读者我要写书评

暂无评论

Infusing Cognition into distributed computing A new approach to distributed datacenters with self-managing services on commodity hardware (Virtualized or Not) 23

Infusing Cognition into Distributed Computing A new approach...

引用

23rd IEEE International Conference on Enabling Technologies, Infrastructure for Collaborative Enterprises (WETICE)

作者： Mikkilineni, Rao Morana, Giovanni C3 DNA Inc Santa Clara CA 95054 USA C3 DNA Inc Catania Italy

ISBN: (纸本)9781479942497

The advent of virtualization technologies and cloud computing has improved application provisioning speed, resource utilization, fault-management, availability using auto-failover and performance optimization using auto-scaling in distributed computing environments. However, heterogeneous virtualization technologies offered by different service providers with disparate infrastructure and their orchestration and management systems have also increased the complexity of managing distributed applications and vendor lock-in. In this paper we utilize new computing, management and programming models introduced using the DIME network architecture to provide end to end service visibility and control across distributed physical or virtual infrastructure. Resulting decoupling of application and service transaction management from myriad distributed infrastructure management systems at run-time enables policy based, secure, service mobility across physical servers or virtual machines deployed in datacenters or public clouds. Using the new architecture we have implemented service self-repair, auto-scaling, live-migration and end-to-end service transaction security independent of server and network security mechanisms.

关键词： business processes service management distributed computing cloud computing cognition DIME network architecture (DNA) Cognitive Container

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：