GB938 Video Over IP Application Note, Release 2.0

Two major Video over IP services are discussed in this document: Broadcast Television and Content on Demand; and Video-telephony/conferencing services.

The IP broadcast video market has extremely high expectations regarding video service quality and reliability. Telco video service quality will inevitably be compared to with those offered by satellite and cable providers. If customer expectations are not met, there will be many other providers to choose from, and when customers switch providers, they will most likely switch all their services. Given these considerations, Quality of Experience is perhaps the most important issue to address. By providing best practices for monitoring of quality, this document will enable Service Providers to offer high quality IP video services.

The success of IP Video-telephony and conferencing services will also depend to a great extent on the quality of the video as perceived by the users of the service, as well as the more well-known telephony service quality factors such as speech quality and call connection quality. This document presents best practices for real-time monitoring of calls, drawing on the conclusions of the existing Best Practice for IP Telephony services.

This best practice presents the case for real time monitoring of live IP video sessions, and for the use of well known video quality metrics based on International Standards, rather than the simplistic but potentially misleading packet based measures of jitter and packet loss. The advantages and disadvantages of active monitoring techniques (using test sessions) are also considered.

Key Performance Indicators (KPIs) are defined as measurements of audio & video quality rather than measures of network performance. Examples of Key Quality Indicators (KQIs) based on these KPIs are contained in this version of the document. Methods of measuring and reporting metrics are also discussed. Measurements of system responsiveness (e.g. channel change) are also considered.

The principal recommendations of this document are that:

• Metrics of video quality that are derived from analysis of video signals or estimated from analysis of packet behaviour in real time should be used. KQIs and KPIs should therefore be based on video quality metrics expressed as a MOS or similar metric rather than e.g. simple measurements of jitter and packet loss.
• Measurements of perceived quality should preferably be made in the STB to be close to the user’s real experience. This will require the industry to standardize economic methods for no-reference, measurement of perceptual video quality and support protocols in order to report this information back to the service provider (It is expected that TR-69 [21] / TR-135 [22] and IETF RTCP-XR [32] will eventually evolve to support this requirement).
• The assessment of customer quality of experience should not be limited to video quality but should also include audio/speech quality and system responsiveness (e.g. channel change time in broadcast services, or call connection delays in video-telephony services).