Node selection in simulation grid
-
摘要: 仿真网格是一个将网格技术和建模与仿真技术相结合的网络化仿真平台,节点选择是提高仿真网格下仿真执行性能的重要手段.根据仿真网格下资源异构性的特点,提出一套形式化的节点性能评价标准,以此标准为依据建立节点选择的数学模型,并给出选择最优节点集合的节点选择算法.算法实验表明,该算法以节点的计算能力为优先考虑,同时顾及节点之间的通信性能,并能够回避通信性能参差不齐时好的通信性能掩盖住差的通信性能的情况,能够起到提高仿真执行效率目的.
-
关键词:
- 仿真网格 /
- 高层体系结构(HLA) /
- 节点选择
Abstract: Simulation grid is a networked simulation platform by combining modeling and simulation (M&S) technology and grid technology. Node selection is an important method to increase the performance of simulation execution in simulation grid. Based on the feature of resource heterogeneousness in simulation grid, a formula appraising criterion for node selection was proposed, a mathematic model for node selection was built on the appraising criterion, and finally a node selection algorithm was designed. Experiment result shows that the algorithm puts emphasis on selecting nodes with good computation capacity, communication performance among nodes is considered at the same time, and the situation where good communication performance overlaps poor communication performance is avoided in this algorithm. So the node selection method is able to achieve the purpose of raising simulation execution efficiency.-
Key words:
- simulation grid /
- high level architecture (HLA) /
- node selection
-
[1] Li Bohu, Chai Xudong, Zhu Wenhai, et al. Research on service oriented simulation grid //Qian Qingquan. The 7th International Symposium on Autonomous Decentralized Systems (ISADS 2005). USA: IEEE Computer Society Press, 2005:7-14[2] Foster I. The grid: a new infrastructure for 21st century science[J]. Physics Today, 2002, 32(1):42-47[3] Kilby M. Overview of the DOD high level architecture . US: Department of Defense, 2000 . [4] Weismann J B. Metascheduling: A scheduling model for metacomputing systems //Salim Hariri. The Seventh IEEE International Symposium on High Performance Distributed Computing. USA: IEEE Conputer Society Press, 1998: 348-349[5] Reiher P L, Jefferson D. Virtual time based dynamic load management in the time warp operation system //Proceedings of the 1990 SCS Multiconference on Distributed Simulation. US: Trans of the Soc for Computer Simulation, 1990:103-111[6] Wolski R, Spring N, Hayes J. The network weather service: a distributed resource performance forecasting service for metacomputing[J]. Future Generation Computer Systems, 1999, 15(5-6):757-768 -
点击查看大图
计量
- 文章访问数: 2766
- HTML全文浏览量: 64
- PDF下载量: 1021
- 被引次数: 0
下载:
