Information Technology Reference
In-Depth Information
The interfaces between the EIMS managers have been defined in a generic way
and implemented as traditional Web Services. However, these management ser-
vices are just the foundation enabling end-to-end QoS/QoE and the management
thereof. The next section will describe means how to effectively measure QoE and
its mapping to standardized description formats enabling interoperability among
the involved entities.
4 Measuring Quality of Experience for MPEG-21-Based
Cross-Layer Multimedia Content Adaptation
4.1 Introduction
The requirement to access multimedia content such as video and audio streams
during everyday's life is omnipresent. Research and standardization efforts around
to what is commonly known as Universal Multimedia Access (UMA) has gained
momentum and offer a rich set of tools enabling such an access from a technologi-
cal point of view. However, most of these techniques exclude the human end user
who is actually the source of the above mentioned requirements and ultimately
wants to consume multimedia content independent of his/her context. The issues
resulting from a more user-centric perspective are collectively referred to as Uni-
versal Multimedia Experience (UME) [23] where the user takes a center stage.
An important aspect with regard to UME is to measure the quality experienced
by the user in an objective way and to signal the required quality metrics by stan-
dardized, i.e., interoperable, description formats. As the objective measures may
require quality metrics coming from various layers (i.e., service, application, and
network) we propose to adopt cross-layer interactions, especially when transmit-
ting multimedia content over wireless channels [24].
Therefore, we propose a QoS/QoE model enabling to measure the perceptual
quality of video transmissions taking into account quality metrics coming from
different layers and following the philosophy of the ITU-T's E-model for audio. In
order to enable interoperability among the involved parties - mainly the service
provider and the content consumer - we propose to adopt description formats (i.e.,
tools) from the MPEG-21 DIA standard (introduce earlier already). In particular,
we demonstrate how our QoS/QoE model can be instantiated using MPEG-21
DIA tools enabling a generic metadata-driven decision-taking component to de-
termine which parameters of the content needs to be adjusted and how in order to
provide a satisfactory quality experienced by the end user.
4.2 Probing Quality of Service
For audio streams, in 1996 the European Telecommunications Standards Institute
(ETSI) and the ITU-T published a model that estimates the perceived quality
