Network operators worldwide are looking closely at the opportunities associated with content delivery, and especially television.
Unfortunately, both business/opportunity planning and network planning processes may be complicated by inconsistencies in defining just what IPTV is.
According to some forecasts, it might appear that everyone on the planet is either already using IPTV or will be using it within a couple of years. At the same time, publications are reporting slow roll-outs of IPTV in some areas, and no roll-outs at all in others.
Defining the terms
While bias and errors can account for some of the discrepancies, the biggest cause of confusion in the market space we call "IPTV" is the lack of a good definition. To get one, we have to decompose the term to "television over IP." As it happens, there's plenty of room for confusion in both the pieces of this phrase.
To start off, what exactly is "television"? In the last decade, that question would have been easily answered, as in, television is a form of video entertainment that involves the broadcast of multiple network channels to a user receiver that can tune to the channel of choice to permit viewing the material. ABC, NBC, Fox, Warner, Discovery, Disney, HBO and others are networks that provide "television" shows. This is the most constraining view of television, so it's a handy place to start.
What then does "over IP" mean? The obvious answer "Over the Internet" is clearly incorrect if the current IPTV services are examined, because they don't send TV over the Internet, but over broadband connections in parallel with Internet access. Even the definition "Over the Internet protocol" may not be fully correct because it is very feasible to transmit "IPTV" over Ethernet or ATM, and not even use the IP protocol at all. The most straightforward interpretation of "over IP," however, is "using the IP protocol for delivery."
The combination of these starting points would mean that at a minimum, "IPTV" is the delivery of a broadcast television experience over the IP protocol. This definition would include AT&T's U-Verse, other Alcatel-Microsoft-based installations, and similar competing offerings. What it would not include may be even more instructive.
What IPTV is not
The definition of IPTV would not include Verizon's FiOS service because FiOS broadcast channels are sent using linear video directly over fiber, not using the IP protocol at all. It would not include any downloaded form of video, or streaming video from YouTube or a network's video portal. In fact, if we assume that the foundation requirement for IPTV is "standard or HD broadcast television over IP," it is quite likely that many geographies in the world will never see any IPTV at all.
A completely permissive definition is equally unrealistic. Internet video delivery, even of TV episodes, can't qualify as IPTV because it can't replicate a TV viewing experience reliably. Best-effort video is likely to generate customer complaints due to delay jitter. In addition, compressed video is tremendously vulnerable to packet loss and pixelization problems. Some degree of QoS management is essential for true IPTV.
What IPTV really is
The best definition of IPTV is "the QoS-assured delivery of standard- or high-definition network or studio programming in either broadcast or on-demand form using the IP protocol." This does not demand that both broadcast channels and streaming VoD be supported, but that at least one of them needs to be provided. Thus, it would embrace both the AT&T U-verse and Verizon FiOS models, since both models have at least one form of TV delivered over IP.
The broadcast dimension represents the most complex issue under this definition. If broadcast material is to be delivered over IP, current technology and practices suggest that the only option is a multicast enrollment strategy that allows consumers to subscribe to some limited number of concurrent streams from a larger selection broadcast channels (best known as the Microsoft TV model). The maximum number of concurrent subscriptions would then be based on the access bandwidth and the type of video (SD or HD).
This model works best where copper loop length can be controlled to permit DSL speeds in excess of 30 Mbps. Where DSL delivery rates are likely to be less than 20 Mbps, it may not be practical to support multiple HD streams, which may make a telco video model non-competitive. Where this form of "slot enrollment" is used, VoD should also be considered a separate channel slot to prevent VoD consumption from overloading the loop.
If broadcast channels are not to be offered as IPTV (they could be provided through a partnership with a satellite provider over FTTH in linear form via a separate wavelength in the FiOS model, etc.), it will still be necessary to manage concurrent VoD feeds over the same loop to eliminate the risk that congestion will impact video performance and cause customer complaints.
A download or store-for-play model of video content delivery could in theory create an "IPTV" experience if a set-top box assembled the material for viewing on a standard TV. But this would still require that the access network have the bandwidth needed to pull in the channels for viewing in near real time, and there are no models of this type of service currently in use. It seems likely that store-for-play video will be ultimately competitive with DVD delivery rather than with traditional television viewing.
It's tempting to think of all of this as an exercise in semantic hair-splitting, but it is critical for planners to understand the exact mission of "IPTV", which is impossible without understanding exactly what form it might take. It is equally critical to be sure that the media and business positioning of an "IPTV service" be precise in setting the outlines of the service, since a misunderstanding could result in hostile press reaction or investor confusion. Finally, the definition of IPTV may change over time in response to market trends, and planners should always be aware of how these changes will impact planning and positioning over time.
About the Author: Tom Nolle is president of CIMI Corporation, a strategic consulting firm specializing in telecommunications and data communications since 1982. He is a member of the IEEE, ACM, Telemanagement Forum, and the IPsphere Forum, and the publisher of Netwatcher, a journal in advanced telecommunications strategy issues. Tom is actively involved in LAN, MAN and WAN issues for both enterprises and service providers and also provides technical consultation to equipment vendors on standards, markets and emerging technologies. Check out his SearchTelecom networking blog Uncommon Wisdom.