Guibin Wang, Xuejun Yang, Y. Zhang, T. Tang, Xudong Fang
Graphic Processing Unit (GPU), with many light-weight data-parallel cores, can provide substantial parallel computational power to accelerate general purpose applications. But the powerful computing capacity could not be fully utilized for memory-intensive applications, which are limited by off-chip memory bandwidth and latency. Stencil computation has abundant parallelism and low computational intensity which make it a useful architectural evaluation benchmark. In this paper, we propose some memory optimizations for a stencil based application mgrid from SPEC 2K benchmarks. Through exploiting data locality in 3-level memory hierarchies and tuning the thread granularity, we reduce the pressure on the off-chip memory bandwidth. To hide the long off-chip memory access latency, we further prefetch data during computation through double-buffer. In order to fully exploit the CPU-GPU heterogeneous system, we redistribute the computation between these two computing resource. Through all these optimizations, we gain 24.2x speedup compared to the simple mapping version, and get as high as 34.3x speedup when compared with a CPU implementation.
{"title":"Program Optimization of Stencil Based Application on the GPU-Accelerated System","authors":"Guibin Wang, Xuejun Yang, Y. Zhang, T. Tang, Xudong Fang","doi":"10.1109/ISPA.2009.70","DOIUrl":"https://doi.org/10.1109/ISPA.2009.70","url":null,"abstract":"Graphic Processing Unit (GPU), with many light-weight data-parallel cores, can provide substantial parallel computational power to accelerate general purpose applications. But the powerful computing capacity could not be fully utilized for memory-intensive applications, which are limited by off-chip memory bandwidth and latency. Stencil computation has abundant parallelism and low computational intensity which make it a useful architectural evaluation benchmark. In this paper, we propose some memory optimizations for a stencil based application mgrid from SPEC 2K benchmarks. Through exploiting data locality in 3-level memory hierarchies and tuning the thread granularity, we reduce the pressure on the off-chip memory bandwidth. To hide the long off-chip memory access latency, we further prefetch data during computation through double-buffer. In order to fully exploit the CPU-GPU heterogeneous system, we redistribute the computation between these two computing resource. Through all these optimizations, we gain 24.2x speedup compared to the simple mapping version, and get as high as 34.3x speedup when compared with a CPU implementation.","PeriodicalId":346815,"journal":{"name":"2009 IEEE International Symposium on Parallel and Distributed Processing with Applications","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-08-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128073552","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Load balancing is an important problem for parallel applications. Recently, many super computers are built on multi-core processors which are usually sharing the last level cache. On one hand different accesses from different cores conflict each other, on the other hand different cores have different work loads resulting in load unbalancing. In this paper, we present a novel technique for balancing parallel applications for multi-core processors based on cache partitioning which can allocate different part of shared caches to different cores exclusively. Our intuitive idea is partitioning shared cache to different cores based on their workloads. That is to say, a heavy load core will get more shared caches than a light load core, so the heavy load core runs faster. We give 2 algorithms in this paper, initial cache partitioning algorithm (ICP) and dynamical cache partitioning algorithm (DCP). ICP is used to determine the best partition when application starting while DCP is used to adjust the initial partition based on the changes of load balancing. Our experiment results show that the running time can be reduced by 7% on average when our load balancing mechanism based on cache partitioning is used.
{"title":"Balancing Parallel Applications on Multi-core Processors Based on Cache Partitioning","authors":"Guang Suo, Xuejun Yang","doi":"10.1109/ISPA.2009.37","DOIUrl":"https://doi.org/10.1109/ISPA.2009.37","url":null,"abstract":"Load balancing is an important problem for parallel applications. Recently, many super computers are built on multi-core processors which are usually sharing the last level cache. On one hand different accesses from different cores conflict each other, on the other hand different cores have different work loads resulting in load unbalancing. In this paper, we present a novel technique for balancing parallel applications for multi-core processors based on cache partitioning which can allocate different part of shared caches to different cores exclusively. Our intuitive idea is partitioning shared cache to different cores based on their workloads. That is to say, a heavy load core will get more shared caches than a light load core, so the heavy load core runs faster. We give 2 algorithms in this paper, initial cache partitioning algorithm (ICP) and dynamical cache partitioning algorithm (DCP). ICP is used to determine the best partition when application starting while DCP is used to adjust the initial partition based on the changes of load balancing. Our experiment results show that the running time can be reduced by 7% on average when our load balancing mechanism based on cache partitioning is used.","PeriodicalId":346815,"journal":{"name":"2009 IEEE International Symposium on Parallel and Distributed Processing with Applications","volume":"35 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-08-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128011713","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The security of inter-domain routing system greatly depends on the trustworthiness of routing information and routing behavior of autonomous system (AS). Many researches on e-commerce, grid, and p2p have proven that reputation mechanism is helpful to inhibit the spread of false route and the occurrence of malicious routing behavior. To increase AS resistance to malicious routing attack, we designs an alliance based reputation model for AS routing behavior trust evaluation. Our approach calculates AS reputation with the Bayesian probability model and manages AS reputation with AS alliance. Compared with the fully distributed reputation model, our model has lower storage and communication overhead. This reputation model is incremental deployment and easy to implement. It can be employed for securing AS routing and assisting malicious behavior detection.
{"title":"An Alliance Based Reputation Model for Internet Autonomous System Trust Evaluation","authors":"Ning Hu, Peidong Zhu, Peng Zou","doi":"10.1109/ISPA.2009.104","DOIUrl":"https://doi.org/10.1109/ISPA.2009.104","url":null,"abstract":"The security of inter-domain routing system greatly depends on the trustworthiness of routing information and routing behavior of autonomous system (AS). Many researches on e-commerce, grid, and p2p have proven that reputation mechanism is helpful to inhibit the spread of false route and the occurrence of malicious routing behavior. To increase AS resistance to malicious routing attack, we designs an alliance based reputation model for AS routing behavior trust evaluation. Our approach calculates AS reputation with the Bayesian probability model and manages AS reputation with AS alliance. Compared with the fully distributed reputation model, our model has lower storage and communication overhead. This reputation model is incremental deployment and easy to implement. It can be employed for securing AS routing and assisting malicious behavior detection.","PeriodicalId":346815,"journal":{"name":"2009 IEEE International Symposium on Parallel and Distributed Processing with Applications","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-08-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132926097","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
With the development of intrusion technologies, dynamic forensics is becoming more and more important. Dynamic forensics using IDS or honeypot are all based on a common hypothesis that the system is still in a reliable working situation and collected evidences are believable even if the system is suffered from intrusion. In fact, the system has already transferred into an insecurity and unreliable state, it is uncertain that whether the intrusion detectors and investigators could run as normal and whether the obtained evidences are credible. Although intrusion tolerance has been applied in many areas of security for years, few researches are referred to network forensics. The work presented in this paper is based on an idea to integrate Intrusion tolerance into dynamic forensics to make the system under control, ensure the reliability of evidences and aim to gather more useful evidences for investigation. A mechanism of dynamic forensics based on intrusion forensics is proposed. This paper introduces the architecture of the model which uses IDS as tolerance and forensics trigger and honeypot as shadow server, the finite state machine model is described to specify the mechanism, and then two cases are analyzed to illuminate the mechanism.
{"title":"Dynamic Forensics Based on Intrusion Tolerance","authors":"Lin Chen, Zhitang Li, C. Gao, Lan Liu","doi":"10.1109/ISPA.2009.66","DOIUrl":"https://doi.org/10.1109/ISPA.2009.66","url":null,"abstract":"With the development of intrusion technologies, dynamic forensics is becoming more and more important. Dynamic forensics using IDS or honeypot are all based on a common hypothesis that the system is still in a reliable working situation and collected evidences are believable even if the system is suffered from intrusion. In fact, the system has already transferred into an insecurity and unreliable state, it is uncertain that whether the intrusion detectors and investigators could run as normal and whether the obtained evidences are credible. Although intrusion tolerance has been applied in many areas of security for years, few researches are referred to network forensics. The work presented in this paper is based on an idea to integrate Intrusion tolerance into dynamic forensics to make the system under control, ensure the reliability of evidences and aim to gather more useful evidences for investigation. A mechanism of dynamic forensics based on intrusion forensics is proposed. This paper introduces the architecture of the model which uses IDS as tolerance and forensics trigger and honeypot as shadow server, the finite state machine model is described to specify the mechanism, and then two cases are analyzed to illuminate the mechanism.","PeriodicalId":346815,"journal":{"name":"2009 IEEE International Symposium on Parallel and Distributed Processing with Applications","volume":"48 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-08-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130380111","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Xiang Zhang, Jie Ma, Yanchao Miao, Qingwei Meng, Dan Meng
The overhead caused by virtualization makes it difficult to apply VM in the applications which require high degrees of both performance isolation and efficiency, such as the high performance computing. In this paper, we present a lightweight virtual machine, named Solo. It simplifies the design of VMM greatly by making most privileged instructions bypass the VMM, except the I/O operations. Solo allows VM running directly on hardware with the highest privileges, therefore greatly reduces the overhead caused by virtualization. Our evaluation shows that Solo not only guarantees the VM performance isolation, but also improves VM performance to the level of traditional OS, and thus meets the requirements of the high performance applications without special hardware support.
{"title":"Solo: A Lightweight Virtual Machine","authors":"Xiang Zhang, Jie Ma, Yanchao Miao, Qingwei Meng, Dan Meng","doi":"10.1109/ISPA.2009.7","DOIUrl":"https://doi.org/10.1109/ISPA.2009.7","url":null,"abstract":"The overhead caused by virtualization makes it difficult to apply VM in the applications which require high degrees of both performance isolation and efficiency, such as the high performance computing. In this paper, we present a lightweight virtual machine, named Solo. It simplifies the design of VMM greatly by making most privileged instructions bypass the VMM, except the I/O operations. Solo allows VM running directly on hardware with the highest privileges, therefore greatly reduces the overhead caused by virtualization. Our evaluation shows that Solo not only guarantees the VM performance isolation, but also improves VM performance to the level of traditional OS, and thus meets the requirements of the high performance applications without special hardware support.","PeriodicalId":346815,"journal":{"name":"2009 IEEE International Symposium on Parallel and Distributed Processing with Applications","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-08-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134455616","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Dayong Ye, Q. Bai, Minjie Zhang, K. Win, Zhiqi Shen
Recently, task allocation in multi-agent systems has been investigated by many researchers. Some researchers suggested to have a central controller which has a global view about the environment to allocate tasks. Although centralized control brings convenience during task allocation processes, it also has some obvious weaknesses. Firstly, a central controller plays an important role in a multi-agent system, but task allocation procedures will break down if the central controller of a system cannot work properly. Secondly, centralized multi-agent architecture is not suitable for distributed working environments. In order to overcome some limitations caused by centralized control, some researchers proposed distributed task allocation protocols. They supposed that each agent has a limited local view about its direct linked neighbors, and can allocate tasks to its neighbors. However, only involving direct linked neighbors could limit resource origins, so that the task allocation efficiency will be greatly reduced. In this paper, we propose an efficient task allocation protocol for P2P multi-agent systems. This protocol allows not only neighboring agents but also indirect linked agents in the system to help with a task if needed. Through this way, agents can achieve more efficient and robust task allocations in loosely coupled distributed environments (e.g. P2P multi-agent systems). A set of experiments are presented in this paper to evaluate the efficiency and adaptability of the protocol. The experiment result shows that the protocol can work efficiently in different situations.
{"title":"An Efficient Task Allocation Protocol for P2P Multi-agent Systems","authors":"Dayong Ye, Q. Bai, Minjie Zhang, K. Win, Zhiqi Shen","doi":"10.1109/ISPA.2009.14","DOIUrl":"https://doi.org/10.1109/ISPA.2009.14","url":null,"abstract":"Recently, task allocation in multi-agent systems has been investigated by many researchers. Some researchers suggested to have a central controller which has a global view about the environment to allocate tasks. Although centralized control brings convenience during task allocation processes, it also has some obvious weaknesses. Firstly, a central controller plays an important role in a multi-agent system, but task allocation procedures will break down if the central controller of a system cannot work properly. Secondly, centralized multi-agent architecture is not suitable for distributed working environments. In order to overcome some limitations caused by centralized control, some researchers proposed distributed task allocation protocols. They supposed that each agent has a limited local view about its direct linked neighbors, and can allocate tasks to its neighbors. However, only involving direct linked neighbors could limit resource origins, so that the task allocation efficiency will be greatly reduced. In this paper, we propose an efficient task allocation protocol for P2P multi-agent systems. This protocol allows not only neighboring agents but also indirect linked agents in the system to help with a task if needed. Through this way, agents can achieve more efficient and robust task allocations in loosely coupled distributed environments (e.g. P2P multi-agent systems). A set of experiments are presented in this paper to evaluate the efficiency and adaptability of the protocol. The experiment result shows that the protocol can work efficiently in different situations.","PeriodicalId":346815,"journal":{"name":"2009 IEEE International Symposium on Parallel and Distributed Processing with Applications","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-08-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130110141","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Fault-tolerance is an important issue for the design of interconnection networks. In this paper, a new fault-tolerant routing algorithm is presented and is applied in mesh networks employing wormhole switching. Due to its low routing restrictions, the presented routing algorithm is so highly adaptive that it is connected and deadlock-free in spite of the various fault regions in mesh networks. Due to the minimal virtual channels it uses, the presented routing algorithm only employs as few buffers as possible and is suitable for fault-tolerant interconnection networks with low cost. Since it chooses the path around fault regions according to the local fault information, the presented routing algorithm takes routing decisions quickly and is applicable in interconnection networks. Moreover, a simulation is conducted for the proposed routing algorithm and the results show that the algorithm exhibits a graceful degradation in performance.
{"title":"Fault-Tolerant Routing Schemes for Wormhole Mesh","authors":"Xinming Duan, Dakun Zhang, Xuemei Sun","doi":"10.1109/ISPA.2009.62","DOIUrl":"https://doi.org/10.1109/ISPA.2009.62","url":null,"abstract":"Fault-tolerance is an important issue for the design of interconnection networks. In this paper, a new fault-tolerant routing algorithm is presented and is applied in mesh networks employing wormhole switching. Due to its low routing restrictions, the presented routing algorithm is so highly adaptive that it is connected and deadlock-free in spite of the various fault regions in mesh networks. Due to the minimal virtual channels it uses, the presented routing algorithm only employs as few buffers as possible and is suitable for fault-tolerant interconnection networks with low cost. Since it chooses the path around fault regions according to the local fault information, the presented routing algorithm takes routing decisions quickly and is applicable in interconnection networks. Moreover, a simulation is conducted for the proposed routing algorithm and the results show that the algorithm exhibits a graceful degradation in performance.","PeriodicalId":346815,"journal":{"name":"2009 IEEE International Symposium on Parallel and Distributed Processing with Applications","volume":"51 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-08-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114853373","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In this paper, a Model Driven Architecture (MDA) approach is applied to Semi-automatically translate sequential programs into corresponding distributed code. The novelty of our work is the use of MDA in the process of translating serial into distributed code. The transformation comprises automatic generation of platform independent and then platform specific models from the sequential code. In order to generate the PIM, a meta-model defining the overall architecture of the resultant distributed code is developed. The meta-model is used as a basis for the development of platform independent models (PIM) for the resultant distributed code. A set of transformation rules are defined to transform the resulted PIM into a corresponding platform-specific model. These transformation rules can be modified by the user, depending on the details of the underlying middle-ware applied for the distribution. The platform independent model provides a better understanding of the distributed code and helps the programmer to modify the code more easily.
{"title":"Semi-automatic Transformation of Sequential Code to Distributed Code Using Model Driven Architecture Approach","authors":"S. Karimi, Saeed Parsa","doi":"10.1109/ISPA.2009.71","DOIUrl":"https://doi.org/10.1109/ISPA.2009.71","url":null,"abstract":"In this paper, a Model Driven Architecture (MDA) approach is applied to Semi-automatically translate sequential programs into corresponding distributed code. The novelty of our work is the use of MDA in the process of translating serial into distributed code. The transformation comprises automatic generation of platform independent and then platform specific models from the sequential code. In order to generate the PIM, a meta-model defining the overall architecture of the resultant distributed code is developed. The meta-model is used as a basis for the development of platform independent models (PIM) for the resultant distributed code. A set of transformation rules are defined to transform the resulted PIM into a corresponding platform-specific model. These transformation rules can be modified by the user, depending on the details of the underlying middle-ware applied for the distribution. The platform independent model provides a better understanding of the distributed code and helps the programmer to modify the code more easily.","PeriodicalId":346815,"journal":{"name":"2009 IEEE International Symposium on Parallel and Distributed Processing with Applications","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-08-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124324187","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In this paper, we analyze the basic characteristics of the propagation network. Using the modeling framework Algebra of System (AoS), we convert the complexity of the network structure into the quantity of data. By using OPN(Object-Process Network) which is an executable meta-language based on the mathematical specification of AoS, we build a framework to describe and evaluate the performance of internet networking during Beijing Olympic Games, which is collected by GPERF, a globally distributed performance measurement system of China Education and Research Network (CERNET). In this framework, we deal with the information of QoS for simulation, visualization, and analysis of user’s needs in a uniform algebraic way.
{"title":"Analysing Internet Information Propagation Using Quality of Service Methodology Based on Algebra of System","authors":"T. Wang, Q. Qu, Benjamin H. Y. Koo","doi":"10.1109/ISPA.2009.87","DOIUrl":"https://doi.org/10.1109/ISPA.2009.87","url":null,"abstract":"In this paper, we analyze the basic characteristics of the propagation network. Using the modeling framework Algebra of System (AoS), we convert the complexity of the network structure into the quantity of data. By using OPN(Object-Process Network) which is an executable meta-language based on the mathematical specification of AoS, we build a framework to describe and evaluate the performance of internet networking during Beijing Olympic Games, which is collected by GPERF, a globally distributed performance measurement system of China Education and Research Network (CERNET). In this framework, we deal with the information of QoS for simulation, visualization, and analysis of user’s needs in a uniform algebraic way.","PeriodicalId":346815,"journal":{"name":"2009 IEEE International Symposium on Parallel and Distributed Processing with Applications","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-08-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121634208","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
M. Hedayat, Wentong Cai, S. Turner, Shayan Shahand
Grid computing is a fundamental technology for large scale distributed resource sharing. Workflow management is becoming one of the most important Grid services. A lot of research work has been done on different issues involved in workflow management systems. The focus of this paper is on three areas: workflow partitioning, enactment and data movement. A new workflow management system called parallel and distributed workflow management system (PDWMS) is proposed. In this system the execution of workflow is done by a network of collaborative engines. To achieve this target, the original abstract workflow (input of the system) is partitioned into parallel parts, using a new proposed partitioning algorithm. PDWMS’s data movement, which is categorized into local and global models, uses a peer-to-peer approach.
{"title":"Distributed Execution of Workflow Using Parallel Partitioning","authors":"M. Hedayat, Wentong Cai, S. Turner, Shayan Shahand","doi":"10.1109/ISPA.2009.96","DOIUrl":"https://doi.org/10.1109/ISPA.2009.96","url":null,"abstract":"Grid computing is a fundamental technology for large scale distributed resource sharing. Workflow management is becoming one of the most important Grid services. A lot of research work has been done on different issues involved in workflow management systems. The focus of this paper is on three areas: workflow partitioning, enactment and data movement. A new workflow management system called parallel and distributed workflow management system (PDWMS) is proposed. In this system the execution of workflow is done by a network of collaborative engines. To achieve this target, the original abstract workflow (input of the system) is partitioned into parallel parts, using a new proposed partitioning algorithm. PDWMS’s data movement, which is categorized into local and global models, uses a peer-to-peer approach.","PeriodicalId":346815,"journal":{"name":"2009 IEEE International Symposium on Parallel and Distributed Processing with Applications","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-08-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124471305","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","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：