On ml methods for network powered by computing infrastructure

封面

如何引用文章

全文:

开放存取 开放存取
受限制的访问 ##reader.subscriptionAccessGranted##
受限制的访问 订阅存取

详细

The paper considers the application of machine learning methods for optimal resource management for Network Powered by Computing (NPC) – a new generation computing infrastructure. The relation between the proposed computing infrastructure and the GRID concept is considered. It is shown how machine learning methods applied to computing infrastructure management make it possible to solve the problems of computing infrastructure management that did not allow the GRID concept to be fully implemented. As an example, the application of multi-agent optimization methods with reinforcement learning for network resources management is considered. It is shown that the application of multi-agent machine learning methods makes it possible to increase the speed of distribution of transport flows and ensure optimal NPC network channel load according to the criterion of uniform load distribution, and that such management of network resources is more effective than a centralized approach.

作者简介

R. Smeliansky

Lomonosov Moscow State University

编辑信件的主要联系方式.
Email: smel@cs.msu.su

Corresponding Member, Faculty of computational mathematics and cybernetics, Department of computing systems and automation

俄罗斯联邦, Moscow

E. Stepanov

Lomonosov Moscow State University

Email: estepanov@lvk.cs.msu.ru

Faculty of computational mathematics and cybernetics, Department of computing systems and automation

俄罗斯联邦, Moscow

参考

  1. Smeliansky R. Hierarchical edge computing // Int. Conf. Modern Network Tech., MoNeTec-2018. Moscow, 2018. P. 97–105.
  2. Smeliansky R. et al. On hpc & cloud environments integration. chapter 1 // Performance evaluation models for distributed service networks. Springer: Springer Nature Switzerland AG Gewerbestrasse 11, 6330 Cham, Switzerland, 2020.
  3. Smeliansky R. Network Powered by Computing: Next Generation of Computational Infrastructure // Edge Computing Technology, Management and Integration. IntechOpen, 2023. ISBN 978-1-83768-862-3. P. 47–70.
  4. Topology and Orchestration Specification for Cloud Applications.http://docs.oasis-open.org/tosca/TOSCA/v1.0/os/TOSCA-v1.0-os.html [Accessed: 2024-19-03]
  5. https://www.itprotoday.com/serverless-computing/what-serverless-computing (accessed: March 19, 2024).
  6. Foster I., Kesselman C. The Grid 2: Blueprint for a new computing infrastructure. Elsevier, 2003.
  7. Foster I., Kesselman. C. The history of the grid //arXiv:2204.04312, 2022.
  8. https://www.cpubenchmark.net/year-on-year.html (accessed: March 19, 2024).
  9. https://www.visualcapitalist.com/cp/charted-history-exponential-growth-in-ai-computation/ (accessed: March 19, 2024).
  10. https://habr.com/ru/companies/yota/articles/283220/ (accessed: March 19, 2024).
  11. Моисеев Н.Н., Иванилов Ю.П., Столярова Е.М. Методы оптимизации. М.: Наука, 1978. 352 с.
  12. Karimireddy S.P. et al. Scaffold: Stochastic controlled averaging for federated learning. International conference on machine learning. PMLR, 2020.
  13. Vogels T., Karimireddy SP., Jaggi M. PowerSGD: Practical low-rank gradient compression for distributed optimization // Advances in Neural Information Processing Systems. 2019.
  14. Oseledets I., Tyrtyshnikov E. TT-cross approximation for multidimensional arrays // Linear Algebra and its Applications. 2010. Т. 432. № 1. P. 70–88.
  15. Gusak J. et al. Automated multi-stage compression of neural networks // Proceedings of the IEEE/CVF Int. Conf. on Computer Vision Workshops, 2019.
  16. Novikov A. et al. Tensorizing neural networks // Advances in neural information processing systems. 2015. N 28.
  17. Gong Y. et al. ETTE: Efficient tensor-train-based computing engine for deep neural networks // Proceedings of the 50th Ann. Int. Symp. on Computer Architecture. 2023. P. 1–13.
  18. Смелянский Р.Л., Антоненко В.А. Концепции программного управления и виртуализации сетевых сервисов в современных сетях передачи данных. М.: Курс, 2019. 160 с.
  19. Bernardez G. et al. Is machine learning ready for traffic engineering optimization? // 2021 IEEE29th International Conference on Network Protocols (ICNP). IEEE, 2021.
  20. You Xinyu et al. Toward packet routing with fully distributed multiagent deep reinforcement learning. // IEEE Transactions on Systems, Man, and Cybernetics: Systems 52.2 (2020): 855–868.
  21. Mai Xuan, Quanzhi Fu and Yi Chen. Packet routing with graph attention multi-agent reinforcement learning // 2021 IEEE Global Communications Conference (GLOBECOM). IEEE, 2021.
  22. Stepanov E. et al. On Fair Traffic allocation and Efficient Utilization of Network Resources based on MARL // Preliminary on ResearchGate. Available from: https://www.researchgate.net/publication/371166584_On_Fair_Traffic_allocation_and_Efficient_Utilization_of_Network_Resources_based_on_MARL (accessed: November 14, 2023).
  23. ECMP Load Balancing. Available from: https://www.cisco.com/c/en/us/td/docs/ios-xml/ios/mp_l3_vpns/configuration/xe-3s/asr903/mp-l3-vpns-xe-3s-asr903-book/mp-l3-vpns-xe-3s-asr903-book_chapter_0100.pdf (accessed: November 14, 2023).
  24. UCMP Load Balancing. Available from: https://www.cisco.com/c/en/us/td/docs/ios-xml/ios/mp_l3_vpns/configuration/xe-3s/asr903/17-1-1/b-mpls-l3-vpns-xe-17-1-asr900/m-ucmp.pdf (accessed: November 14, 2023).

补充文件

附件文件
动作
1. JATS XML

版权所有 © Russian Academy of Sciences, 2024