Content delivery network slicing: QoE and cost awareness

Sara Retal, Miloud Bagaa, Tarik Taleb, Hannu Flinck

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review

45 Citations (Scopus)
377 Downloads (Pure)


Content Delivery Networks (CDNs) emerged to manage the great amount of content, as well as the transmissions over long distances. In recent years, this concept proves to be a promising solution for emergent enterprises. In this paper, we present a Content Delivery Network as a Service (CDNaaS) platform which can create virtual machines (VMs) through a network of data centers and provide a customized slice of CDN to users. CDNaaS manages a great number of videos by means of caches, transcoders, and streamers hosted in different VMs. However, an optimal placement of VMs with adequate flavors for the different images is required to obtain an efficient slice of CDN. In this work, we argue the need to find a convenient slice for the CDN owner while respecting his performance requirements and minimizing as much as possible the incurred cost. We first formulate the VMs placement problem as two Linear Integer problem solutions, aiming at minimizing the cost and maximizing the quality of experience of streaming. Then, extensive simulation results are presented to illustrate the effectiveness of the proposed models.

Original languageEnglish
Title of host publication2017 IEEE International Conference on Communications, ICC 2017
Number of pages6
ISBN (Electronic)9781467389990
Publication statusPublished - 28 Jul 2017
MoE publication typeA4 Conference publication
EventIEEE International Conference on Communications - Paris, France
Duration: 21 May 201725 May 2017

Publication series

NameIEEE International Conference on Communications
ISSN (Print)1550-3607
ISSN (Electronic)1938-1883


ConferenceIEEE International Conference on Communications
Abbreviated titleICC


Dive into the research topics of 'Content delivery network slicing: QoE and cost awareness'. Together they form a unique fingerprint.

Cite this