Telco Video: The Quality Scale

Telcos are making some big and well-publicized strides in going after the video services business that has been dominated by cable MSOs for more than a generation. But to turn this quick start into a long-term victory, telcos have to make sure that they can deliver high-quality video services reliably to a mass market. A look under the hood suggests that there's still plenty of ground to cover in terms of the quality of service (QOS) and quality of experience (QOE) of IPTV delivery.

Let's first take QOS. You'd think QOS for video would not be much of an issue, since the IP infrastructure has successfully evolved to be carrier-class, capable of handling the most demanding business applications. IPTV, however, is different, not just because of the bandwidth demands it will place on the network, but also because the IP network will be operating in a highly dynamic environment, as a result of subscribers frequently wanting to leave one multicast group and join a different one – in other words, to channel surf. IP networks to date have not had to deal with what happens when millions of end users start changing channels simultaneously during commercial breaks.

My current research into IPTV quality issues leaves me reasonably confident that the QOS challenge will be met with some effort, but I'm a lot less confident when it comes to the QOE issue. The notion of IPTV QOE is really about subscriber perception of service quality – which, of course, is highly subjective. To ensure that telcos have a successful IPTV rollout, test and measurement vendors are attempting to provide the tools for objectively measuring the quality of the video as displayed on TV screens – in other words, to use objective measurements to validate subjective attitudes. It's not exactly logical, but it's the best that can be done.

The best tools that telcos have can, in a sense, only guess at the picture quality the subscriber is seeing. They can check to see whether the video quality has been degraded by the encoders, or they can look at metrics such as IP jitter and packet loss and arithmetically compute the impact on the video stream. But neither approach can be exact, unless you know what the original video signal looked like.

What this means for telcos is that they are going to have to test for QOE in the lab and in the operational network, and then judiciously uses the QOE tools available for live testing in the network and/or at the subscriber location. After that, it's mostly crossing the fingers and hoping that they are getting an accurate QOE reading that will keep subscribers from bolting to the competition.

— Sam Masud, Senior Analyst, Heavy Reading

Be the first to post a comment regarding this story.
Sign In