Predicting throughput of cloud network infrastructure using neural networks

Derek Phanekham; Suku Nair; Nageswara Rao; Mike Truty

doi:10.1109/INFOCOMWKSHPS51825.2021.9484520

Predicting throughput of cloud network infrastructure using neural networks

Derek Phanekham
, Suku Nair
, Nageswara Rao
, Mike Truty

Autonomous and Complex Systems

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

8 Scopus citations

Abstract

Throughput prediction of network infrastructures is an important aspect of capacity planning, scheduling, resource management, route selection and other network functions. In this paper, we describe throughput measurements collected over a network infrastructure that supports cloud computing spanning the globe. We train deep learning models to predict TCP throughput using these measurements, which show performance improvements with buffer tuning and parallel streams. We also compare the accuracy of machine learning and conventional methods in predicting both single thread and mutli-stream throughput in a public cloud environment.

Original language	English
Title of host publication	IEEE INFOCOM 2021 - IEEE Conference on Computer Communications Workshops, INFOCOM WKSHPS 2021
Publisher	Institute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)	9781665404433
DOIs	https://doi.org/10.1109/INFOCOMWKSHPS51825.2021.9484520
State	Published - May 10 2021
Event	2021 IEEE Conference on Computer Communications Workshops, INFOCOM WKSHPS 2021 - Virtual, Online Duration: May 9 2021 → May 12 2021

Publication series

Name	IEEE INFOCOM 2021 - IEEE Conference on Computer Communications Workshops, INFOCOM WKSHPS 2021

Conference

Conference	2021 IEEE Conference on Computer Communications Workshops, INFOCOM WKSHPS 2021
City	Virtual, Online
Period	05/9/21 → 05/12/21

Keywords

Cloud network infrastructure
Deep learning
Machine learning
Neural networks
TCP throughput prediction

Access to Document

10.1109/INFOCOMWKSHPS51825.2021.9484520

Cite this

Phanekham, D., Nair, S., Rao, N., & Truty, M. (2021). Predicting throughput of cloud network infrastructure using neural networks. In IEEE INFOCOM 2021 - IEEE Conference on Computer Communications Workshops, INFOCOM WKSHPS 2021 Article 9484520 (IEEE INFOCOM 2021 - IEEE Conference on Computer Communications Workshops, INFOCOM WKSHPS 2021). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/INFOCOMWKSHPS51825.2021.9484520

Phanekham, Derek ; Nair, Suku ; Rao, Nageswara et al. / Predicting throughput of cloud network infrastructure using neural networks. IEEE INFOCOM 2021 - IEEE Conference on Computer Communications Workshops, INFOCOM WKSHPS 2021. Institute of Electrical and Electronics Engineers Inc., 2021. (IEEE INFOCOM 2021 - IEEE Conference on Computer Communications Workshops, INFOCOM WKSHPS 2021).

@inproceedings{49833787292549c1bd0dbcce4f54f3cd,

title = "Predicting throughput of cloud network infrastructure using neural networks",

abstract = "Throughput prediction of network infrastructures is an important aspect of capacity planning, scheduling, resource management, route selection and other network functions. In this paper, we describe throughput measurements collected over a network infrastructure that supports cloud computing spanning the globe. We train deep learning models to predict TCP throughput using these measurements, which show performance improvements with buffer tuning and parallel streams. We also compare the accuracy of machine learning and conventional methods in predicting both single thread and mutli-stream throughput in a public cloud environment.",

keywords = "Cloud network infrastructure, Deep learning, Machine learning, Neural networks, TCP throughput prediction",

author = "Derek Phanekham and Suku Nair and Nageswara Rao and Mike Truty",

note = "Publisher Copyright: {\textcopyright} 2021 IEEE.; 2021 IEEE Conference on Computer Communications Workshops, INFOCOM WKSHPS 2021 ; Conference date: 09-05-2021 Through 12-05-2021",

year = "2021",

month = may,

day = "10",

doi = "10.1109/INFOCOMWKSHPS51825.2021.9484520",

language = "English",

series = "IEEE INFOCOM 2021 - IEEE Conference on Computer Communications Workshops, INFOCOM WKSHPS 2021",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

booktitle = "IEEE INFOCOM 2021 - IEEE Conference on Computer Communications Workshops, INFOCOM WKSHPS 2021",

}

Phanekham, D, Nair, S, Rao, N & Truty, M 2021, Predicting throughput of cloud network infrastructure using neural networks. in IEEE INFOCOM 2021 - IEEE Conference on Computer Communications Workshops, INFOCOM WKSHPS 2021., 9484520, IEEE INFOCOM 2021 - IEEE Conference on Computer Communications Workshops, INFOCOM WKSHPS 2021, Institute of Electrical and Electronics Engineers Inc., 2021 IEEE Conference on Computer Communications Workshops, INFOCOM WKSHPS 2021, Virtual, Online, 05/9/21. https://doi.org/10.1109/INFOCOMWKSHPS51825.2021.9484520

Predicting throughput of cloud network infrastructure using neural networks. / Phanekham, Derek; Nair, Suku; Rao, Nageswara et al.
IEEE INFOCOM 2021 - IEEE Conference on Computer Communications Workshops, INFOCOM WKSHPS 2021. Institute of Electrical and Electronics Engineers Inc., 2021. 9484520 (IEEE INFOCOM 2021 - IEEE Conference on Computer Communications Workshops, INFOCOM WKSHPS 2021).

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

TY - GEN

T1 - Predicting throughput of cloud network infrastructure using neural networks

AU - Phanekham, Derek

AU - Nair, Suku

AU - Rao, Nageswara

AU - Truty, Mike

PY - 2021/5/10

Y1 - 2021/5/10

N2 - Throughput prediction of network infrastructures is an important aspect of capacity planning, scheduling, resource management, route selection and other network functions. In this paper, we describe throughput measurements collected over a network infrastructure that supports cloud computing spanning the globe. We train deep learning models to predict TCP throughput using these measurements, which show performance improvements with buffer tuning and parallel streams. We also compare the accuracy of machine learning and conventional methods in predicting both single thread and mutli-stream throughput in a public cloud environment.

AB - Throughput prediction of network infrastructures is an important aspect of capacity planning, scheduling, resource management, route selection and other network functions. In this paper, we describe throughput measurements collected over a network infrastructure that supports cloud computing spanning the globe. We train deep learning models to predict TCP throughput using these measurements, which show performance improvements with buffer tuning and parallel streams. We also compare the accuracy of machine learning and conventional methods in predicting both single thread and mutli-stream throughput in a public cloud environment.

KW - Cloud network infrastructure

KW - Deep learning

KW - Machine learning

KW - Neural networks

KW - TCP throughput prediction

UR - https://www.scopus.com/pages/publications/85113334028

U2 - 10.1109/INFOCOMWKSHPS51825.2021.9484520

DO - 10.1109/INFOCOMWKSHPS51825.2021.9484520

M3 - Conference contribution

AN - SCOPUS:85113334028

T3 - IEEE INFOCOM 2021 - IEEE Conference on Computer Communications Workshops, INFOCOM WKSHPS 2021

BT - IEEE INFOCOM 2021 - IEEE Conference on Computer Communications Workshops, INFOCOM WKSHPS 2021

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2021 IEEE Conference on Computer Communications Workshops, INFOCOM WKSHPS 2021

Y2 - 9 May 2021 through 12 May 2021

ER -

Phanekham D, Nair S, Rao N, Truty M. Predicting throughput of cloud network infrastructure using neural networks. In IEEE INFOCOM 2021 - IEEE Conference on Computer Communications Workshops, INFOCOM WKSHPS 2021. Institute of Electrical and Electronics Engineers Inc. 2021. 9484520. (IEEE INFOCOM 2021 - IEEE Conference on Computer Communications Workshops, INFOCOM WKSHPS 2021). doi: 10.1109/INFOCOMWKSHPS51825.2021.9484520

Predicting throughput of cloud network infrastructure using neural networks

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this