I think it was Friday night �date-night� about three years ago that I first starting thinking about TV quality of experience (QoE). Trust me, that was not what I wanted to be thinking about. As a road warrior like many of you, �date night� was a chance for me and my wife Kirsten to catch up and relax, with the kids in bed and no distractions. Usually we browsed the movie channels and, if nothing looked good, there we went with DirectTV�s VoD service. Well, as Murphy�s Law would have it, foul weather more often than not seemed to mar the picture quality with the blocking problems that are typical of �rain fade.� Though I am sure it helped my marriage (we had to choose a game to play or gasp! actually talk to each other), I could have written a new coffee table book on ways to curse out the TV.
In frustration, I switched to Comcast (News - Alert)
Cable for a �double play� solution of
broadband Internet and digital TV, but
my TV QoE problems only followed me
there. Channel change speed was horrible
and in the time it took for me to
navigate 100 channels on the electronic
program guide (EPG) I could have
walked to the kitchen, made a ham and
cheese sandwich, hugged the kids, and
returned to the living room without
breaking a sweat. It made me long for
the days when I only had to worry
about TV static �snow� and tuning the
rabbit ears to ju-u-u-u-st the right angle
(at least I felt I had some control).
With IPTV, the stakes for QoE are
raised ten-fold. The level of interactivity,
service blending, and sheer volume of
content options increase the potential to
excite and frustrate. So in these few
brief paragraphs I would like to highlight
the key QoE parameters the IPTV
industry should be thinking about.
First, I would like to expand the
industry definition of IPTV subscriber
�quality of experience.� The term is
built upon a similar networking concept
called Quality of Service (QoS),
which is very much a technical term
referring to things like packet loss, jitter,
delay and other parameters that
many times get written into service
level agreements (SLAs) between
Internet service providers and their
enterprise customers. The TV/video
industry quickly understood that these
parameters alone did not ensure that
what subscribers actually saw on their
TVs measured up to quality standards
of viewing. That�s why they came up
with ways to track the quality of the
video subscribers saw on their screens
(using Mean Opinion Scores or �MOS�
ratings), and then implemented probes
in the delivery chain from the head-end
to the set-top box (STB) to locate any
problems and bottlenecks (such as a
flaky decoder chip or bad Ethernet port
on the STB).
So as far as QoE is concerned, video
quality is indeed a starting point. But I
would also like to add other variables
such as how intuitive the navigation of
the EPG is, EPG time-to-scroll, including
vertically cell-to-cell, page-to-page,
the fast scroll of many channels, and
horizontally (temporal scan of the same
channel), how quickly you can change
channels, how quickly you can identify
what you want to watch and actually
watch it (�time to watch�), content
search speed, and the balance between
providing enough information on the
user guide and too much information
(which becomes even more important
on multiple-line HD EPGs).
When it comes to IPTV, it is important
to remember that most operators
are not implementing IPTV in a vacuum,
but are deploying combinational
triple play (TV plus Internet and VoIP)
and quad play (add mobility) services.
So another key QoE parameter is how
well these combinational services are
blended together. For example, do I
have to go through several clicks and a
separate �Communications� tab on the
user guide or is it easy to send and
receive calls or initiate IM chats?
In this world of device intelligence
and personal style, how smart is the
IPTV service to track my preferences
and what options do I have to customize
the look and feel of my interfaces
(called �skins�) and apply parental
controls to filter what my kids are
watching?
So far, I have seen few IPTV solutions
that combine the intuitive, customizable,
graphically rich user interface
that subscribers have come to
expect with the fast performance they
demand. Those that try usually fall
down on one end of the spectrum or
the other, or they are cost-prohibitive
and don�t scale technically or financially.
For example, IPTV solutions that use a
browser-only approach (appropriate for
some applications such as VoD), may
allow for rich and customizable interfaces,
but their performance and scalability
are lacking and leave subscribers
waiting far too long to refresh the
browser screen with each click of the
remote control. Proprietary IPTV solutions
(you know who I am talking
about) may be a little faster, but they
lack the openness and flexibility operators
are looking for. Worse, they often
take technical shortcuts, such as requiring
three video streams to ensure fast
channel change speed (the up and
down channel is always a click away)
and require expensive, bloated, PC-like
STBs that don�t scale and require a
plethora of processing power and
servers in the back office.
There is a better way. One that combines
open Web development tools for
rich, customizable interfaces with a fast
and cost-effective STB and IPTV component
ecosystem. It is what can loosely
be called a �data-driven IPTV architecture�
and it combines the best of open
development frameworks and the performance
one would expect from
embedded C code. What�s more, this
approach, now seriously being looked at
by some of the leading IPTV vendors
and service providers in the world, does
not break the bank because it can be
run on low- to middle-end STBs and
requires far less network bandwidth for
EPG population than first generation
technologies. And best of all for subscribers,
this �data-driven architecture�
allows service providers the tools to
develop the most user-friendly QoE
available. The end result is a QoE/TCO
(Total Cost of Ownership) ratio that
makes both IPTV subscribers and operators
smile.
In the grand scheme of things, the
IPTV industry is still in its infancy. But
for this market to enjoy wide-scale user
adoption, it will have to provide a superior
QoE and overcome the challenges
of earlier TV services. To grow beyond
lab trials and pilot deployments, IPTV
vendors and service providers alike need
to make the right architectural decisions
and avoid the mistakes of the past. My
date night is counting on it. IT
Brian Mahony is the Vice President of
Marketing for Espial (News - Alert) (www.espial.com), a
leading IPTV middleware and applications
provider. He is a speaker and writer on
next-gen services such as IPTV, VoIP, and
triple/quad play. He can be contacted at
[email protected] and would welcome
feedback and help promoting IPTV architectures with strong QoE/TCO ratios.
If you are interested in purchasing reprints of this article (in either print or PDF format), please visit Reprint Management Services online at www.reprintbuyer.com or contact a representative via e-mail at [email protected] or by phone at 800-290-5460.
|