|
Work item:
|
P.DiAQoSE
|
|
Subject/title:
|
Diagnostic assessment of QoS and QoE for adaptive video streaming sessions
|
|
Status:
|
[Carried to next study period]
|
|
Approval process:
|
AAP
|
|
Type of work item:
|
Recommendation
|
|
Version:
|
New
|
|
Equivalent number:
|
-
|
|
Timing:
|
-
|
|
Liaison:
|
-
|
|
Supporting members:
|
Ericsson Research, L.M. Ericsson
Nippon Telegraph and Telephone Corporation (NTT)
Ilmenau University of Technology
AT&T
|
|
Summary:
|
There are a variety of measurement approaches and corresponding Key Performance Indicators (KPIs) for characterizing HTTP-based Adaptive Video streaming sessions and services. Typical Quality of Service (QoS) type indicators are, for example, bandwidth, latency, loss, and delay variation. With P.1203 and P.1204, also standardized indicators for Quality of Experience (QoE) are available, sometimes referred to as KQIs (Key Quality Indicators). In a practical network monitoring and measurement context, a certain set of KPIs and KQIs is typically collected. Currently, the challenge of a coherent presentation of these different measures and of their integration in terms of conclusions on network quality is unsolved. The new Work Item P.DiAQoSE targets to inform the market about how the respective measurement information should best be presented. Also, approaches for a better linking between QoE and QoS will be considered. Here, the work will e.g. address aspects such as the specification of events or measurement results that should be signalled as an “alarm” with regard to network or service performance. The models described in P.1203 and P.1204 deliver QoE estimations based on different types of input parameters, which include aspects such as the audio and video codec used, the bitrate, framerate, resolution, as well as stalling and initial loading delay event characteristics. These parameters can directly be used as diagnostic information about service performance. Besides this, there may be underlying KPIs such as latency or loss that lead to certain adaptive streaming configurations regarding bitrate, resolution and buffering (and consequently initial loading delay or stalling). The Work Item addresses both the link between underlying KPIs and P.1203 and P.1204 input parameters, as well as a better understanding of the relation between P.1203 and P.1204 input parameters and the resulting QoE estimation in terms of Mean Opinion Score (MOS). For the latter, another aspect is of high interest: Can individual parameters be identified that may be the most relevant since deteriorating characteristic of the service at a given moment in time, and to be improved so that the resulting quality in terms of MOS may be optimized?
The Work Item will have three threads: (1) P.1203/4 measurement result presentation and diagnostic QoS-related analysis with regard to P.1203/4 input parameters. The result will be an “application guide” for users of P.1203/4. (2) As an Annex or Appendix for the P.DiAQoSE Recommendation, an analysis and optimization scheme for adaptive video streaming sessions will be developed, targeting the most relevant deteriorating factors that may affect quality in terms of MOS. (3) In a further set of Annexes or Appendices to P.DiAQoSE, guidelines regarding the relation between QoS-type KPIs and the P.1203/4 input parameters will be provided.
|
|
Comment:
|
-
|
|
Reference(s):
|
|
|
Historic references:
|
|
Contact(s):
|
|
| ITU-T A.5 justification(s): |
|
|
|
|
First registration in the WP:
2020-04-28 16:35:42
|
|
Last update:
2021-11-03 13:03:11
|
|
