Parallel Design and Load Balance of OpenMP Based Warfighting Simulation

doi:10.16182/j.issn1004731x.joss.201806026

Abstract

Abstract: Parallel simulation is a main trend of large scale and high precision war fighting simulation (WS). However, currently there is no unified framework to design a parallel WS and no effective method to load balance of it. Based on the parallel agent theory, an OpenMP based parallel WS design framework (PWSF) is proposed. To solve the problem of the complex simulation load distribution resulted from the diversity of entities' models and the position independence of the interactions, a load balance method based on METIS multilevel graph partition tool is suggested. Based on the improved PHOLD test model, experiments on different types and amounts of simulation loads and different numbers of CPU cores are conducted. The results show that METIS based PWSF is with good universality, timeliness and expandability.

Key words: OpenMP, parallel war fighting simulation (PWS), METIS, load balance

CLC Number:

TP391.9

Wang Xiao, Liu Yaqi, Tao Yuben. Parallel Design and Load Balance of OpenMP Based Warfighting Simulation[J]. Journal of System Simulation, 2018, 30(6): 2206-2215.

References

[1] 徐晓东, 赵建亭, 许春雷. 分导式多弹头弹道导弹并行计算技术研究[J]. 弹箭与制导学报, 2013, 33(4): 146-148.
Xu Xiaodong, Zhao Jianting, Xu Chunlei.Parallel Technology of Multiple Independently Targeted Reentry Vehicle's Ballistic Trajectory[J]. Journal of Projectiles,Rockets,Missiles and Guidance, 2013, 33(4):146-148.
[2] 陈昊. 基于Open MP的并行蚁群算法求解协同空战火力分配[J]. 传感器与微系统, 2013, 32(1): 20-24.
Chen Hao.Parallel ant colony algorithm for solving weapon-target assignment based on OpenMP in cooperative air combat[J]. Transducer and Microsystem Technologies, 2013, 32(1): 20-24.
[3] Macal C M, North M J.Agent-based Modeling and Simulation[C]// Proc. of the 2009 Winter Simulation Conference[S.l.]: IEEE Press, 2009: 86-98.
[4] Paul K Davis, Robert H Anderson.Improving the Compos ability of DoD Models and Simulations[J]. The Journal of Defense Modeling and Simulation: Applications, Methodology, Technology (S1548-5129), 2004, 1(1): 5-17.
[5] 杜国红, 韦伟, 李路遥. 作战仿真实体组件化建模研究[J]. 系统仿真学报, 2015, 27(2): 234-239.
Du Guohong, Wei Wei, Li Luyao.Research on Component-based Modeling of Combat Simulation Entity[J]. Journal of System Simulation, 2015, 27(2): 234-239.
[6] J Todd McDonald, Michael L Talbert. Agent-based Architecture for Modeling and Simulation Integration[C]// Proc. of the National Aerospace & Electronics Conference. [S.l.]: IEEE Press, 2000: 375-382.
[7] Lees M, Logan B, Theodoropoulos G.Using Access Patterns to Analyze the Performance of Optimistic Synchronic Algorithms in Simulations of MAS[J]. Simulation (S2310-4791), 2008, 84(10/11): 481-492.
[8] 余文广, 王维平, 侯洪涛, 等. 基于多核CPU-GPU异构平台的并行Agent仿真[J]. 系统工程与电子技术, 2012, 34(8): 1716-1722.
Yu Wen-guang, Wang Wei-ping, Hou Hong-tao, et al.Parallel Agent-based Simulation on Multi-core CPU and GPU Heterogeneous Platforms[J]. Sysems Engineering and Electronics, 2012, 34(8): 1716-1722.
[9] 余文广, 王维平, 李群. 并行Agent仿真研究综述[J].系统仿真学报, 2012, 24(2): 245-251.
YU Wen-guang, WANG Wei-ping, LI Qun.Review of Parallel Agent-based Simulation[J]. Journal of System Simulation,2012, 24(2): 245-251.
[10] 姚益平, 张颖星. 集群计算环境下基于复杂网络的社会学仿真负载划分优化算法[J]. 计算机研究与发展, 2011, 48(9): 1759-1767.
Yao Yiping, Zhang Yingxing.An Optimized Partitioning Algorithm for Complex Network Based on Social Simulations on Cluster Computing Platform[J]. Journal of Computer Research and Development, 2011, 48(9): 1759-1767.
[11] 王维平, 余文广, 侯洪涛, 等. 多核CPU-GPU异构平台下并行Agent仿真负载均衡方法[J]. 系统工程与电子技术, 2012, 34(11): 2366-2373.
Wang Weiping, Yu Wenguang, Hou Hongtao, et al.Load Balancing Mechanism for Parallel Agent-based Simulation on Multi-core CPU and GPU Heterogenous Platform[J]. Sysems Engineering and Electronics, 2012, 34(11): 2366-2373.
[12] Biagio Cosenza, Gennaro Cordasco, Rosario De Chiara, et al.Distributed Load Balancing for Parallel Agent-based Simulation[C]// 2011 19th International Euromicro Conference on Parallel, Distributed and Network-Based Processing.[S.l.]: IEEE Press, 2011: 62-69.
[13] Chapman B, Jost G, Van Der Pas R. Using OpenMP: Portable Shared Memory Parallel Programming[M]. London: MIT Press, 2008: 35-49.
[14] 周伟明. OpenMP编程指南[EB/OL]. http://blog.csdn. net/drzhouweiming/article/details/4093624.
Zhou Wei-ming. A guide to OpenMP Programming [EB/OL]. http://blog.csdn.net/drzhouweiming/article/ details/4093624.
[15] 余文广, 王维平, 侯洪涛, 等. 面向并行Agent仿真的合成基准测试模型[J]. 系统工程与电子技术, 2012, 34(4): 813-819.
Yu Wen-guang, Wang Wei-ping, Hou Hong-tao, et al.Synthetic Benchmark Model for Parallel Agent-based Simulation[J]. Sysems Engineering and Electronics, 2012, 34(4): 813-819.
[16] AJ Alt, Philip A Wilsey.Profile Driven Partitioning of Parallel Simulation Models[C]// Proceedings of the 2014 Winter Simulation Conference. Florida, USA: IEEE Press, 2014: 2750-2761.
[17] George Karypis, Vipin Kumar.A Fast and High Quality Multilevel Scheme for Partitioning Irregular Graphs[J]. SIAM Journal on Scientific Computing (S1064-8275), 1998, 20(1): 359-392.
[18] George Karypis.Metis: A Software Package for Partitioning Unstructured Graphs, Partitioning Meshes, and Computing Fill-Reducing Orderings of Sparse Matrices, Version 5.1.0[R]. Minneapolis, US: University of Minnesota, 2013: 1-32.