Data-driven approaches to edge caching

Guangyu Li, Qiang Shen, Yong Liu, Houwei Cao, Zifa Han, Feng Li, Jin Li

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

Abstract

Content caching at network edge is a promising solution for serving emerging high-throughput low-delay applications, such as virtual reality, augmented reality and Internet-of-Things. The traditional caching algorithms need to adapt to the edge networking environment since old traffic assumptions may no longer hold. Meanwhile, user/group content interest as a new important element should be considered to improve the caching performance. In this work, we propose two novel caching strategies that mine user/group interests to improve caching performance at network edge. The static user-group interest patterns are handled by the Matrix Factorization method and the temporal content request patterns are handled by the Least-Recently-Used or Nearest-Neighbor algorithms. Through empirical experiments with a large-scale real IPTV user traces, we demonstrate that the proposed caching algorithms outperform the existing caching algorithms and approach the caching performance upper bound in the large cache size regime. Leveraging on offline computation, we can limit the online computation cost and achieve good caching performance in realtime.

Original languageEnglish (US)
Title of host publicationNEAT 2018 - Proceedings of the 2018 Workshop on Networking for Emerging Applications and Technologies, Part of SIGCOMM 2018
PublisherAssociation for Computing Machinery, Inc
Pages8-14
Number of pages7
ISBN (Electronic)9781450359078
DOIs
StatePublished - Aug 7 2018
EventACM SIGCOMM 2018 Workshop on Networking for Emerging Applications and Technologies, NEAT 2018 - Budapest, Hungary
Duration: Aug 20 2018 → …

Publication series

NameNEAT 2018 - Proceedings of the 2018 Workshop on Networking for Emerging Applications and Technologies, Part of SIGCOMM 2018

Other

OtherACM SIGCOMM 2018 Workshop on Networking for Emerging Applications and Technologies, NEAT 2018
CountryHungary
CityBudapest
Period8/20/18 → …

Fingerprint

Caching
Data-driven
IPTV
Augmented reality
Factorization
Virtual reality
Throughput
Internet of Things
Factorization Method
Matrix Factorization
Augmented Reality
Matrix Method
Virtual Reality
Costs
Networking
Cache
High Throughput
Experiments
Nearest Neighbor
Trace

Keywords

  • Data-driven
  • Edge caching

ASJC Scopus subject areas

  • Software
  • Theoretical Computer Science
  • Computer Science Applications
  • Hardware and Architecture

Cite this

Li, G., Shen, Q., Liu, Y., Cao, H., Han, Z., Li, F., & Li, J. (2018). Data-driven approaches to edge caching. In NEAT 2018 - Proceedings of the 2018 Workshop on Networking for Emerging Applications and Technologies, Part of SIGCOMM 2018 (pp. 8-14). (NEAT 2018 - Proceedings of the 2018 Workshop on Networking for Emerging Applications and Technologies, Part of SIGCOMM 2018). Association for Computing Machinery, Inc. https://doi.org/10.1145/3229574.3229582

Data-driven approaches to edge caching. / Li, Guangyu; Shen, Qiang; Liu, Yong; Cao, Houwei; Han, Zifa; Li, Feng; Li, Jin.

NEAT 2018 - Proceedings of the 2018 Workshop on Networking for Emerging Applications and Technologies, Part of SIGCOMM 2018. Association for Computing Machinery, Inc, 2018. p. 8-14 (NEAT 2018 - Proceedings of the 2018 Workshop on Networking for Emerging Applications and Technologies, Part of SIGCOMM 2018).

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

Li, G, Shen, Q, Liu, Y, Cao, H, Han, Z, Li, F & Li, J 2018, Data-driven approaches to edge caching. in NEAT 2018 - Proceedings of the 2018 Workshop on Networking for Emerging Applications and Technologies, Part of SIGCOMM 2018. NEAT 2018 - Proceedings of the 2018 Workshop on Networking for Emerging Applications and Technologies, Part of SIGCOMM 2018, Association for Computing Machinery, Inc, pp. 8-14, ACM SIGCOMM 2018 Workshop on Networking for Emerging Applications and Technologies, NEAT 2018, Budapest, Hungary, 8/20/18. https://doi.org/10.1145/3229574.3229582
Li G, Shen Q, Liu Y, Cao H, Han Z, Li F et al. Data-driven approaches to edge caching. In NEAT 2018 - Proceedings of the 2018 Workshop on Networking for Emerging Applications and Technologies, Part of SIGCOMM 2018. Association for Computing Machinery, Inc. 2018. p. 8-14. (NEAT 2018 - Proceedings of the 2018 Workshop on Networking for Emerging Applications and Technologies, Part of SIGCOMM 2018). https://doi.org/10.1145/3229574.3229582
Li, Guangyu ; Shen, Qiang ; Liu, Yong ; Cao, Houwei ; Han, Zifa ; Li, Feng ; Li, Jin. / Data-driven approaches to edge caching. NEAT 2018 - Proceedings of the 2018 Workshop on Networking for Emerging Applications and Technologies, Part of SIGCOMM 2018. Association for Computing Machinery, Inc, 2018. pp. 8-14 (NEAT 2018 - Proceedings of the 2018 Workshop on Networking for Emerging Applications and Technologies, Part of SIGCOMM 2018).
@inproceedings{552685e1c7a34e188d0bdfbfb370c310,
title = "Data-driven approaches to edge caching",
abstract = "Content caching at network edge is a promising solution for serving emerging high-throughput low-delay applications, such as virtual reality, augmented reality and Internet-of-Things. The traditional caching algorithms need to adapt to the edge networking environment since old traffic assumptions may no longer hold. Meanwhile, user/group content interest as a new important element should be considered to improve the caching performance. In this work, we propose two novel caching strategies that mine user/group interests to improve caching performance at network edge. The static user-group interest patterns are handled by the Matrix Factorization method and the temporal content request patterns are handled by the Least-Recently-Used or Nearest-Neighbor algorithms. Through empirical experiments with a large-scale real IPTV user traces, we demonstrate that the proposed caching algorithms outperform the existing caching algorithms and approach the caching performance upper bound in the large cache size regime. Leveraging on offline computation, we can limit the online computation cost and achieve good caching performance in realtime.",
keywords = "Data-driven, Edge caching",
author = "Guangyu Li and Qiang Shen and Yong Liu and Houwei Cao and Zifa Han and Feng Li and Jin Li",
year = "2018",
month = "8",
day = "7",
doi = "10.1145/3229574.3229582",
language = "English (US)",
series = "NEAT 2018 - Proceedings of the 2018 Workshop on Networking for Emerging Applications and Technologies, Part of SIGCOMM 2018",
publisher = "Association for Computing Machinery, Inc",
pages = "8--14",
booktitle = "NEAT 2018 - Proceedings of the 2018 Workshop on Networking for Emerging Applications and Technologies, Part of SIGCOMM 2018",

}

TY - GEN

T1 - Data-driven approaches to edge caching

AU - Li, Guangyu

AU - Shen, Qiang

AU - Liu, Yong

AU - Cao, Houwei

AU - Han, Zifa

AU - Li, Feng

AU - Li, Jin

PY - 2018/8/7

Y1 - 2018/8/7

N2 - Content caching at network edge is a promising solution for serving emerging high-throughput low-delay applications, such as virtual reality, augmented reality and Internet-of-Things. The traditional caching algorithms need to adapt to the edge networking environment since old traffic assumptions may no longer hold. Meanwhile, user/group content interest as a new important element should be considered to improve the caching performance. In this work, we propose two novel caching strategies that mine user/group interests to improve caching performance at network edge. The static user-group interest patterns are handled by the Matrix Factorization method and the temporal content request patterns are handled by the Least-Recently-Used or Nearest-Neighbor algorithms. Through empirical experiments with a large-scale real IPTV user traces, we demonstrate that the proposed caching algorithms outperform the existing caching algorithms and approach the caching performance upper bound in the large cache size regime. Leveraging on offline computation, we can limit the online computation cost and achieve good caching performance in realtime.

AB - Content caching at network edge is a promising solution for serving emerging high-throughput low-delay applications, such as virtual reality, augmented reality and Internet-of-Things. The traditional caching algorithms need to adapt to the edge networking environment since old traffic assumptions may no longer hold. Meanwhile, user/group content interest as a new important element should be considered to improve the caching performance. In this work, we propose two novel caching strategies that mine user/group interests to improve caching performance at network edge. The static user-group interest patterns are handled by the Matrix Factorization method and the temporal content request patterns are handled by the Least-Recently-Used or Nearest-Neighbor algorithms. Through empirical experiments with a large-scale real IPTV user traces, we demonstrate that the proposed caching algorithms outperform the existing caching algorithms and approach the caching performance upper bound in the large cache size regime. Leveraging on offline computation, we can limit the online computation cost and achieve good caching performance in realtime.

KW - Data-driven

KW - Edge caching

UR - http://www.scopus.com/inward/record.url?scp=85056385826&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85056385826&partnerID=8YFLogxK

U2 - 10.1145/3229574.3229582

DO - 10.1145/3229574.3229582

M3 - Conference contribution

T3 - NEAT 2018 - Proceedings of the 2018 Workshop on Networking for Emerging Applications and Technologies, Part of SIGCOMM 2018

SP - 8

EP - 14

BT - NEAT 2018 - Proceedings of the 2018 Workshop on Networking for Emerging Applications and Technologies, Part of SIGCOMM 2018

PB - Association for Computing Machinery, Inc

ER -