Advanced

Analysis and characterization of a video-on-demand service workload

Ali-Eldin, Ahmed; Kihl, Maria LU ; Tordsson, Johan and Elmroth, Erik (2015) 6th ACM Multimedia Systems Conference In [Host publication title missing] p.189-200
Abstract
Video-on-Demand (VoD) and video sharing services account for a large percentage of the total downstream Internet traffic. In order to provide a better understanding of the load on these services, we analyze and model a workload trace from a VoD service provided by a major Swedish TV broadcaster. The trace contains over half a million requests generated by more than 20000 unique users. Among other things, we study the request arrival rate, the inter-arrival time, the spikes in the workload and their cause, the video popularity distribution, the streaming bit-rate distribution and the video duration distribution. Contrary to some previously analyzed workloads in the literature, our results show that the user and the session arrival rates for... (More)
Video-on-Demand (VoD) and video sharing services account for a large percentage of the total downstream Internet traffic. In order to provide a better understanding of the load on these services, we analyze and model a workload trace from a VoD service provided by a major Swedish TV broadcaster. The trace contains over half a million requests generated by more than 20000 unique users. Among other things, we study the request arrival rate, the inter-arrival time, the spikes in the workload and their cause, the video popularity distribution, the streaming bit-rate distribution and the video duration distribution. Contrary to some previously analyzed workloads in the literature, our results show that the user and the session arrival rates for the TV4 workload does not follow a Poisson process. The arrival rate distribution is modeled using a lognormal distribution while the interarrival time distribution is modeled using a stretched exponential distribution. We observe the “impatient user” behavior where users abandon streaming sessions after minutes or

even seconds of starting them. Both very popular videos and non-popular videos are specially affected by impatient users. We also show that this behavior is an invariant in VoD workloads and is neither affected by the average bit-rate nor by the number of videos a user watch. (Less)
Please use this url to cite or link to this publication:
author
organization
publishing date
type
Chapter in Book/Report/Conference proceeding
publication status
published
subject
in
[Host publication title missing]
pages
189 - 200
publisher
ACM
conference name
6th ACM Multimedia Systems Conference
external identifiers
  • scopus:84942543893
ISBN
978-1-4503-3351-1
DOI
10.1145/2713168.2713183
project
LCCC
EIT_VR CLOUD Cloud Control
language
English
LU publication?
yes
id
490fc8a7-214b-44a6-9d10-534f09819a2a (old id 8835022)
date added to LUP
2016-03-10 15:43:20
date last changed
2017-10-01 05:08:16
@inproceedings{490fc8a7-214b-44a6-9d10-534f09819a2a,
  abstract     = {Video-on-Demand (VoD) and video sharing services account for a large percentage of the total downstream Internet traffic. In order to provide a better understanding of the load on these services, we analyze and model a workload trace from a VoD service provided by a major Swedish TV broadcaster. The trace contains over half a million requests generated by more than 20000 unique users. Among other things, we study the request arrival rate, the inter-arrival time, the spikes in the workload and their cause, the video popularity distribution, the streaming bit-rate distribution and the video duration distribution. Contrary to some previously analyzed workloads in the literature, our results show that the user and the session arrival rates for the TV4 workload does not follow a Poisson process. The arrival rate distribution is modeled using a lognormal distribution while the interarrival time distribution is modeled using a stretched exponential distribution. We observe the “impatient user” behavior where users abandon streaming sessions after minutes or<br/><br>
even seconds of starting them. Both very popular videos and non-popular videos are specially affected by impatient users. We also show that this behavior is an invariant in VoD workloads and is neither affected by the average bit-rate nor by the number of videos a user watch.},
  author       = {Ali-Eldin, Ahmed and Kihl, Maria and Tordsson, Johan and Elmroth, Erik},
  booktitle    = {[Host publication title missing]},
  isbn         = {978-1-4503-3351-1},
  language     = {eng},
  pages        = {189--200},
  publisher    = {ACM},
  title        = {Analysis and characterization of a video-on-demand service workload},
  url          = {http://dx.doi.org/10.1145/2713168.2713183},
  year         = {2015},
}