一种高性能嵌入式云计算平台架构被引量：2

A High-Performance Embedded Cloud Computing Platform Architecture

下载PDF

导出

摘要在商用云计算技术发展背景下,基于计算资源虚拟化、虚拟网络技术、虚拟资源集群管理技术以及MapReduce、Spark等大数据处理框架的理论基础,提出了一种嵌入式云计算平台体系架构。该架构基于轻量级容器化的方法解决了嵌入式资源的虚拟化问题。通过内存映射核间通信中间件构建了嵌入式虚拟资源网络,并实现了一种可支持回溯的有向无环图计算模型,应用内各个算法单元之间的通信时延可低至10μs,可支持高速实时信号处理、高速实时数据分析等分布式应用的运行。 Under the background of commercial cloud computing technology development,based on the theories of computing resource virtualization,virtual network technology,virtual resource cluster management technology and MapReduce,Spark and other big data processing frameworks,this paper proposes an embedded cloud computing platform architecture.The architecture is based on a lightweight containerization method,which solves the virtualization problem of embedded resources.An embedded virtual resource network is constructed through the memory-mapped inter-core communication middleware,and a directed acyclic graph computing model that can support backtracking is realized.In the application,the communication delay between each algorithm unit can be as low as ten microseconds,which can support the operation of distributed applications such as high-speed real-time signal processing and high-speed real-time data analysis.

作者潘灵 PAN Ling(Research Institute of Electronic Technology Southwest China,Chengdu Sichuan 610036,China)

机构地区中国西南电子技术研究所

出处《通信技术》 2021年第5期1184-1188,共5页 Communications Technology

关键词资源虚拟化虚拟网络有向无环图实时信号处理 resource virtualization virtual network direct acyclic graph real-time signal processing

分类号 TN919 [电子电信—通信与信息系统]

引文网络
相关文献

参考文献2

1张京妹,李英俊,苏广文,欧阳盼.嵌入式云计算系统的资源实时共享方法研究[J].西北工业大学学报,2016,34(3):393-398. 被引量：8
2卢罡,徐勤良,许南山,郭俊霞.复杂网络松耦合分布式计算框架的设计与实现[J].计算机工程,2015,41(11):73-76. 被引量：3

二级参考文献20

1Bader D A,Madduri K. Designing Multithreaded Algo- rithms for Breadth-first Search and st-connectivity on the Cray MTA-2 [ C ]//Proceedings of ICPP' 06. Washington D. C. , USA : IEEE Press ,2006:523-530.
2Bader D A,Madduri K. Parallel Algorithms for Evaluat- ing Centrality Indices in Real-world Networks[C]// Proceedings of ICPP' 06. Washington D. C. , USA: IEEE Press ,2006:539-550.
3Zhao Xiaohan, Sala A, Zheng Haitao, et al. Efficient Shortest Paths on Massive Social Graphs[Cl// Proceedings of the 7th International Conference on Collaborative Computing: Networking, Applications and Worksharing. Orlando, USA: [ s. n. ] ,2011:77-86.
4Hardin D S, Hardin S S. ACL2 Meets the GPU: Formalizing a CUDA-based Parallelizable All-pairs Shortest Path Algorithm in ACL2 [ C ]//Proceedings of ACL2' 13. [ S. 1. ] :EPTCS,2013:127-142.
5Pandey M, Pandey H, Sharma S. In-place Recursive Approach for All-pairs Shortest-path Problem Using OPENCL I C 1//Proceedings of ICIT' 13. Washington D. C, , USA :IEEE Press ,2013.
6Ortega-Arranz H,Torres Y, Llanos D R, et al. The All- pair Shortest-path Problem in Shared-memory Hetero- geneous Systems E EB/OL]. (2013-01-03). http://www. infor, uva. es/- yuri. torres/docs/hector_Complex_2013, pdf.
7Sariyiice A E. Betweenness Centrality on GPUs and Heterogeneous Architectures [ C ]//Proceedings of the 6th Workshop on General Purpose Processor Using Graphics Processing Units. New York, USA: ACM Press ,2013.
8Aapo K, Blelloch G E, Guestrin C. GraphChi: Large- scale Graph Computation on Just a PC [ C ]// Proceedings of OSDI' 12. Washington D. C., USA: IEEE Press ,2012:31-46.
9Ediger D, Jiang K, Riedy J, et al. GraphCT: Multi- threaded Algorithms for Massive Graph Analysis[J]. IEEE Transactions on Parallel and Distributed Systems, 2013,24( 11 ) :2220-2229.
10Kang U, Tsourakakis C E. PEGASUS: A Peta-scale Graph Mining System Implementation and Observa- tions[ C]//Proceedings of ICDM' 09. Washington D. C. , USA: IEEE Press,2009 : 229-238.