不规则多级胖树网络的优化模型

doi:10.3969/j.issn.1001-2400.2017.06.001

Abstract

Abstract:

To reduce the cost of traditional fat-tree networks, an irregular multi-level fat-tree network (IMLFTN) is proposed, with its network node using a special basic switching module (BSM) in which the number of upward ports is fixed while the number of downward ports remains variable. Analyzing the switching mechanism and traffic distribution of each BSM, the cost function of a BSM considers its numbers of upward and downward ports, and the traffic through it as well. The BSMs used at the first level are chosen according to the total number of users, and the BSMs located at other levels are accordingly determined and connected together to form an IMLFTN. The optimal IMLFTN is selected by the total cost of all IMLFTNs constructed by the combination of BSMs with different numbers of ports. Numerical results show that the proposed IMLFTN has a large advantage in cost when constructing the network with the same number of users.

Key words: irregular multi-level fat-tree networks, basic switching module, switching mechanism, flow distribution, cost function

XU Zhanqi;WANG Na;YANG Fan;ZHANG Yushuai. Optimization model of the irregular multi-level fat-tree network[J].Journal of Xidian University, 2017, 44(6): 1-7.

References

［1］ BOSE A, GHOSAL P, MOHANTY S P. STA: a Highly Scalable Low Latency Butterfly Fat Tree Based 3D NoC Design［C］//Proceedings of the IEEE Computer Society Annual Symposium on Very Large Scale Integration Circuit. Piscataway: IEEE, 2016: 496-501.
［2］ KIM S H, LEE J S, LEE J H, et al. Link-wirelength-aware Topology Generation for High Performance Asynchronous NoC Design［J］. Journal of the Institute of Electronics and Information Engineers, 2016, 53(8): 49-58.
［3］ QUINTIN J N, VIGNERAS P. Transitively Deadlock-free Routing Algorithms［C］//Proceedings of the 2nd IEEE International Workshop on High-performance Interconnection Networks in the Exascale and Big-data Era. Piscataway: IEEE, 2016: 16-24.
［4］ DUAN J, YANG Y. Placement and Performance Analysis of Virtual Multicast Networks in Fat-tree Data Center Networks［J］. IEEE Transactions on Parallel and Distributed Systems, 2016, 27(10): 3013-3028.
［5］ FAN F, HU B, YEUNG K L. Distributed and Dynamic Multicast Scheduling in Fat-tree Data Center Networks［C］//Proceedings of the 2016 IEEE International Conference on Communications. Piscataway: IEEE, 2016: 7511250.
［6］张润生, 刘健, 李艳斌. 利用有限混合模型的高效网络拓扑推断算法［J］. 西安电子科技大学学报, 2015, 42(4): 133-139.
ZHANG Runsheng, LIU Jian, LI Yanbin. Efficient Topology Inference Algorithm Using the Finite Mixture Model［J］. Journal of Xidian University, 2015, 42(4): 133-139.
［7］ MAHAPATRA S, YUAN X, NIENABER W. Limited Multi-path Routing on Extended Generalized Fat-trees［C］//Proceedings of the 2012 IEEE 26th International Parallel and Distributed Processing Symposium Workshops. Piscataway: IEEE, 2012: 938-945.
［8］ ANJUM S, KHAN I A, ANWAR W, et al. A Scalable and Minimized Butterfly Fat Tree(SMBFT) Switching Network for On-chip Communication［J］. Research Journal of Applied Sciences, Engineering and Technology, 2012, 4(13): 1997-2002.
［9］ NAVARIDAS J, MIGUEL-ALONSO J, RIDRUEJO F J, et al. Reducing Complexity in Tree-like Computer Interconnection Networks［J］. Parallel Computing, 2010, 36(2/3): 71-85.
［10］ XU Z Q, WANG C T, ZHOU Z Q, et al. Modeling and Optimization of a Special Multistage Star Switching(SMSSS) System with SEs at Unequal Port Rates［J］. China Communications, 2014, 11(7): 48-63.
［11］ FUENTES P, VALLEJO E, MARTINEZ C, et al. Comparison Study of Scalable and Cost-effective Interconnection Networks for HPC［C］//Proceedings of the International Conference on Parallel Processing Workshops. Piscataway: IEEE, 2012: 594-595.
［12］ RAJAN B, RAJASINGH I, VASANTHI BEULAH P. On the Crossing Number of Generalized Fat Trees［C］//Proceedings of the Communications in Computer and Information Science. Heidelberg: Springer Verlag, 2011: 440-448.
［13］ MINKENBERG C, LUIJTEN R P, RODRIGUEZ G . On the Optimum Switch Radix in Fat Tree Networks［C］//Proceedings of the 2011 IEEE 12th International Conference on High Performance Switching and Routing. Piscataway: IEEE, 2011: 44-51.

[1]	YUAN Wenhao,LOU Yingxi,LIANG Chunyan,WANG Zhiqiang. Speech enhancement method based on the perceptual joint optimization deep neural network [J]. Journal of Xidian University, 2019, 46(2): 89-94.
[2]	LU Jin;SU Hongtao;SHUI Penglang. Target state estimation of the over-the-horizon radar using the cost-reference particle filter [J]. J4, 2013, 40(5): 20-25+59.
[3]	SHEN Yu-long(1;2);PEI Qing-qi(2);MA Jian-feng(1;2). AR routing algorithm for the extensible resource reservation duration [J]. J4, 2006, 33(6): 956-959.

Optimization model of the irregular multi-level fat-tree network

PDF (PC)

Like

Knowledge

Abstract

Cite this article

share this article

References

Related Articles 3

Metrics

Comments

Recommended 10