About | Contact Us | Register | Login
ProceedingsSeriesJournalsSearchEAI
9th EAI International Conference on Performance Evaluation Methodologies and Tools

Research Article

Capacity bounds and robustness in multipath networks

Cite
BibTeX Plain Text
  • @INPROCEEDINGS{10.4108/eai.14-12-2015.2262625,
        author={Andrei Bejan and Richard Gibbens and Robert Hancock and Don Towsley},
        title={Capacity bounds and robustness in multipath networks},
        proceedings={9th EAI International Conference on Performance Evaluation Methodologies and Tools},
        publisher={ACM},
        proceedings_a={VALUETOOLS},
        year={2016},
        month={1},
        keywords={mutlipath capacity bounds performance analysis},
        doi={10.4108/eai.14-12-2015.2262625}
    }
    
  • Andrei Bejan
    Richard Gibbens
    Robert Hancock
    Don Towsley
    Year: 2016
    Capacity bounds and robustness in multipath networks
    VALUETOOLS
    ICST
    DOI: 10.4108/eai.14-12-2015.2262625
Andrei Bejan1, Richard Gibbens2,*, Robert Hancock3, Don Towsley4
  • 1: Smith Institute
  • 2: Computer Laboratory, University of Cambridge
  • 3: Roke Manor Research
  • 4: UMass
*Contact email: richard.gibbens@cl.cam.ac.uk

Abstract

The recent developments of multipath data transport protocols such as Multipath TCP allow end-systems to explore and share available resources within networks. Through dynamic load balancing over subflows these protocols ensure high levels of robustness to network failures and traffic overloads. In this paper we use fluid models to study the benefits that accrue when load is shared across subflows. We combine insights gained from the fluid models with a precise description of the capacity region for the network and show that our models of multipath protocols approach the boundary of the capacity region as the intensity of the offered traffic approaches a critical value. We quantify the extent to which multipath protocols will make a network robust to unforeseen traffic mismatches and link failures and illustrate our results with parameterised models for random fluctuations in the offered traffic.

Keywords
mutlipath capacity bounds performance analysis
Published
2016-01-04
Publisher
ACM
http://dx.doi.org/10.4108/eai.14-12-2015.2262625
Copyright © 2015–2025 ICST
EBSCOProQuestDBLPDOAJPortico
EAI Logo

About EAI

  • Who We Are
  • Leadership
  • Research Areas
  • Partners
  • Media Center

Community

  • Membership
  • Conference
  • Recognition
  • Sponsor Us

Publish with EAI

  • Publishing
  • Journals
  • Proceedings
  • Books
  • EUDL