On Periodic Point-to-Point Bandwidth Scheduling in High-Performance Networks

Siming Wang, Aiqin Hou, Chase Q. Wu, Xinji Qu

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Today's network services and applications are widely distributed across multiple data centers around the world, and require timely transfer of big data between data centers to ensure the quality and reliability of services. Network infrastructure that connects data centers is expensive and such big data transfer would incur high cost, a large portion of which is due to the consumption of bandwidth resources. This paper aims to investigate the problem of minimizing data transfer cost in high-performance networks. We construct rigorous models for link bandwidth and link cost between data centers and formulate a Cost Minimization Bandwidth Scheduling Problem (CMBSP). Considering the differences in time and space of transmission cost, we focus on the problem of bandwidth redundancy con-sumption and cost waste caused by discontinuous time step transmission. We formulate CMBSP as an optimization problem of path selection and link bandwidth allocation, and design a cost timing allocation algorithm (CTAA), which allows users to transfer as much data as possible over inter-data center links and minimize the bandwidth cost. Simulation-based experiments in real-life network topologies such as ESnet and NSFNET show that CTAA outperforms state-of-the-art solutions.

Original languageEnglish (US)
Title of host publicationProceedings - 2023 IEEE International Conference on High Performance Computing and Communications, Data Science and Systems, Smart City and Dependability in Sensor, Cloud and Big Data Systems and Application, HPCC/DSS/SmartCity/DependSys 2023
EditorsJinjun Chen, Laurence T. Yang
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages528-535
Number of pages8
ISBN (Electronic)9798350330014
DOIs
StatePublished - 2023
Externally publishedYes
Event25th IEEE International Conferences on High Performance Computing and Communications, 9th International Conference on Data Science and Systems, 21st IEEE International Conference on Smart City and 9th IEEE International Conference on Dependability in Sensor, Cloud and Big Data Systems and Applications, HPCC/DSS/SmartCity/DependSys 2023 - Melbourne, Australia
Duration: Dec 13 2023Dec 15 2023

Publication series

NameProceedings - 2023 IEEE International Conference on High Performance Computing and Communications, Data Science and Systems, Smart City and Dependability in Sensor, Cloud and Big Data Systems and Application, HPCC/DSS/SmartCity/DependSys 2023

Conference

Conference25th IEEE International Conferences on High Performance Computing and Communications, 9th International Conference on Data Science and Systems, 21st IEEE International Conference on Smart City and 9th IEEE International Conference on Dependability in Sensor, Cloud and Big Data Systems and Applications, HPCC/DSS/SmartCity/DependSys 2023
Country/TerritoryAustralia
CityMelbourne
Period12/13/2312/15/23

All Science Journal Classification (ASJC) codes

  • Artificial Intelligence
  • Computer Networks and Communications
  • Information Systems
  • Information Systems and Management
  • Energy Engineering and Power Technology
  • Safety, Risk, Reliability and Quality
  • Instrumentation
  • Urban Studies

Keywords

  • High-performance network
  • bandwidth scheduling
  • cost minimization
  • periodic scheduling

Fingerprint

Dive into the research topics of 'On Periodic Point-to-Point Bandwidth Scheduling in High-Performance Networks'. Together they form a unique fingerprint.

Cite this