MPCC: Online learning multipath transport

Tomer Gilad; Neta Rozen-Schiff; P. Brighten Godfrey; Costin Raiciu; Michael Schapira

doi:10.1145/3386367.3433030

MPCC: Online learning multipath transport

Tomer Gilad
, Neta Rozen-Schiff
, P. Brighten Godfrey
, Costin Raiciu
, Michael Schapira

The Rachel and Selim Benin School of Engineering and Computer Science

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

34 Scopus citations

Abstract

Multipath transport, as embodied in MPTCP, is deployed to improve throughput and reliability in mobile and residential access networks, with additional use-cases including spreading load in data centers and WANs. However, MPTCP is fundamentally tied to TCP Reno's legacy AIMD algorithm, and significantly lags behind the performance of modern single-path designs. Consequently, MPTCP fails to achieve high performance in many real-world environments. We present MPCC, a high-performance multipath congestion control architecture. To achieve our combined goals of fairness and high performance in challenging environments, MPCC employs online convex optimization (a.k.a. online learning). In experiments with a kernel implementation on emulated and live networks, MPCC significantly outperforms MPTCP.

Original language	English
Title of host publication	CoNEXT 2020 - Proceedings of the 16th International Conference on Emerging Networking EXperiments and Technologies
Publisher	Association for Computing Machinery, Inc
Pages	121-135
Number of pages	15
ISBN (Electronic)	9781450379489
DOIs	https://doi.org/10.1145/3386367.3433030
State	Published - 23 Nov 2020
Event	16th ACM Conference on Emerging Networking Experiment and Technologies, CoNEXT 2020 - Barcelona, Spain Duration: 1 Dec 2020 → 4 Dec 2020

Publication series

Name	CoNEXT 2020 - Proceedings of the 16th International Conference on Emerging Networking EXperiments and Technologies

Conference

Conference	16th ACM Conference on Emerging Networking Experiment and Technologies, CoNEXT 2020
Country/Territory	Spain
City	Barcelona
Period	1/12/20 → 4/12/20

Bibliographical note

Publisher Copyright:
© 2020 ACM.

Keywords

congestion control
multipath

Access to Document

10.1145/3386367.3433030

Cite this

Gilad, T., Rozen-Schiff, N., Godfrey, P. B., Raiciu, C., & Schapira, M. (2020). MPCC: Online learning multipath transport. In CoNEXT 2020 - Proceedings of the 16th International Conference on Emerging Networking EXperiments and Technologies (pp. 121-135). (CoNEXT 2020 - Proceedings of the 16th International Conference on Emerging Networking EXperiments and Technologies). Association for Computing Machinery, Inc. https://doi.org/10.1145/3386367.3433030

Gilad, Tomer ; Rozen-Schiff, Neta ; Godfrey, P. Brighten et al. / MPCC : Online learning multipath transport. CoNEXT 2020 - Proceedings of the 16th International Conference on Emerging Networking EXperiments and Technologies. Association for Computing Machinery, Inc, 2020. pp. 121-135 (CoNEXT 2020 - Proceedings of the 16th International Conference on Emerging Networking EXperiments and Technologies).

@inproceedings{94002eb5ca9440d6a2f90d102b343030,

title = "MPCC: Online learning multipath transport",

abstract = "Multipath transport, as embodied in MPTCP, is deployed to improve throughput and reliability in mobile and residential access networks, with additional use-cases including spreading load in data centers and WANs. However, MPTCP is fundamentally tied to TCP Reno's legacy AIMD algorithm, and significantly lags behind the performance of modern single-path designs. Consequently, MPTCP fails to achieve high performance in many real-world environments. We present MPCC, a high-performance multipath congestion control architecture. To achieve our combined goals of fairness and high performance in challenging environments, MPCC employs online convex optimization (a.k.a. online learning). In experiments with a kernel implementation on emulated and live networks, MPCC significantly outperforms MPTCP.",

keywords = "congestion control, multipath",

author = "Tomer Gilad and Neta Rozen-Schiff and Godfrey, \{P. Brighten\} and Costin Raiciu and Michael Schapira",

note = "Publisher Copyright: {\textcopyright} 2020 ACM.; 16th ACM Conference on Emerging Networking Experiment and Technologies, CoNEXT 2020 ; Conference date: 01-12-2020 Through 04-12-2020",

year = "2020",

month = nov,

day = "23",

doi = "10.1145/3386367.3433030",

language = "אנגלית",

series = "CoNEXT 2020 - Proceedings of the 16th International Conference on Emerging Networking EXperiments and Technologies",

publisher = "Association for Computing Machinery, Inc",

pages = "121--135",

booktitle = "CoNEXT 2020 - Proceedings of the 16th International Conference on Emerging Networking EXperiments and Technologies",

}

Gilad, T, Rozen-Schiff, N, Godfrey, PB, Raiciu, C & Schapira, M 2020, MPCC: Online learning multipath transport. in CoNEXT 2020 - Proceedings of the 16th International Conference on Emerging Networking EXperiments and Technologies. CoNEXT 2020 - Proceedings of the 16th International Conference on Emerging Networking EXperiments and Technologies, Association for Computing Machinery, Inc, pp. 121-135, 16th ACM Conference on Emerging Networking Experiment and Technologies, CoNEXT 2020, Barcelona, Spain, 1/12/20. https://doi.org/10.1145/3386367.3433030

MPCC: Online learning multipath transport. / Gilad, Tomer; Rozen-Schiff, Neta; Godfrey, P. Brighten et al.
CoNEXT 2020 - Proceedings of the 16th International Conference on Emerging Networking EXperiments and Technologies. Association for Computing Machinery, Inc, 2020. p. 121-135 (CoNEXT 2020 - Proceedings of the 16th International Conference on Emerging Networking EXperiments and Technologies).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - MPCC

T2 - 16th ACM Conference on Emerging Networking Experiment and Technologies, CoNEXT 2020

AU - Gilad, Tomer

AU - Rozen-Schiff, Neta

AU - Godfrey, P. Brighten

AU - Raiciu, Costin

AU - Schapira, Michael

PY - 2020/11/23

Y1 - 2020/11/23

N2 - Multipath transport, as embodied in MPTCP, is deployed to improve throughput and reliability in mobile and residential access networks, with additional use-cases including spreading load in data centers and WANs. However, MPTCP is fundamentally tied to TCP Reno's legacy AIMD algorithm, and significantly lags behind the performance of modern single-path designs. Consequently, MPTCP fails to achieve high performance in many real-world environments. We present MPCC, a high-performance multipath congestion control architecture. To achieve our combined goals of fairness and high performance in challenging environments, MPCC employs online convex optimization (a.k.a. online learning). In experiments with a kernel implementation on emulated and live networks, MPCC significantly outperforms MPTCP.

AB - Multipath transport, as embodied in MPTCP, is deployed to improve throughput and reliability in mobile and residential access networks, with additional use-cases including spreading load in data centers and WANs. However, MPTCP is fundamentally tied to TCP Reno's legacy AIMD algorithm, and significantly lags behind the performance of modern single-path designs. Consequently, MPTCP fails to achieve high performance in many real-world environments. We present MPCC, a high-performance multipath congestion control architecture. To achieve our combined goals of fairness and high performance in challenging environments, MPCC employs online convex optimization (a.k.a. online learning). In experiments with a kernel implementation on emulated and live networks, MPCC significantly outperforms MPTCP.

KW - congestion control

KW - multipath

UR - https://www.scopus.com/pages/publications/85097616396

U2 - 10.1145/3386367.3433030

DO - 10.1145/3386367.3433030

M3 - ???researchoutput.researchoutputtypes.contributiontobookanthology.conference???

AN - SCOPUS:85097616396

T3 - CoNEXT 2020 - Proceedings of the 16th International Conference on Emerging Networking EXperiments and Technologies

SP - 121

EP - 135

BT - CoNEXT 2020 - Proceedings of the 16th International Conference on Emerging Networking EXperiments and Technologies

PB - Association for Computing Machinery, Inc

Y2 - 1 December 2020 through 4 December 2020

ER -

Gilad T, Rozen-Schiff N, Godfrey PB, Raiciu C, Schapira M. MPCC: Online learning multipath transport. In CoNEXT 2020 - Proceedings of the 16th International Conference on Emerging Networking EXperiments and Technologies. Association for Computing Machinery, Inc. 2020. p. 121-135. (CoNEXT 2020 - Proceedings of the 16th International Conference on Emerging Networking EXperiments and Technologies). doi: 10.1145/3386367.3433030

MPCC: Online learning multipath transport

Abstract

Publication series

Conference

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this