I got an interesting suggestion from a colleague – would it make sense to provide streaming TV over an LTE system using fixed antennas operating in the UHF bands – i.e. replacing current DVB-T systems and freeing “prime” mobile service spectrum for more flexible use than just traditional TV ?
Well, I think my colleague is right – we need more flexible systems that can adapt to the needs of the users – the age of “one-trick ponys” has certainly passed. 10 years ago most people in most northern European countries had “Linear” Broadcast TV as their main source of information and entertainment which has made TV Broadcasting a politically prioritized technology. In 10 years very few will watch TV the way one did 10 years ago and then the motivation for using “prime” mobile spectrum for serving a few percent of the users will be very week. So if we can solve these few users problem with a general, IP based solution, thats fine. On the other hand, building a system designed for traditional (linear streaming) TV based on new technology I think is meaningless – the users will not be interested by the time we are done. DVB-H is a very good example of solving yesterdays problem with tomorrows technology.
On the political scene consumer broadband internet access is now becoming the topic of the day and my guess is that this will take priority over “Broadcast TV access” in the next few years – eventually leading to a plan for dismantling the DVB-T networks, freeing spectrum for mobile access. In the densely populated areas where most households already are on cable this is a “no-brainer” already today. When the technology shift will happens (not if!) depends on political inertia and some problems mainly related to rural access – where DVB-T is still quite efficient. LTE systems do not get that much help from more spectrum in the coverage department. Admittedly, lower frequencies provide somewhat more favorable coverage properties, but its not dramatically different to 8-900 MHz). It will still be very expensive to provide TV-grade 10-20 Mbit/s data rates in rural areas – regardless of technology. In Sweden/Finland maybe we are talking about 10- 20% of the households that (for economical reasons) will never get fiber access, nor 10 Mb+ mobile broadband. Even though you can watch standard TV at lower data rates, I doubt this would be politically accepted in Sweden (e.g. having lower quality if you live in the “outback”). Maybe it will be cheaper to provide these users with satellite dishes! Compared with spectrum fees being paid, I think this would be a good deal for the operators. ;-)
This is an interesting idea, and there seems to be activity on several fronts pushing in a similar direction. See, for example, the powerpoint attached to a document which was submitted to the FCC recently (the title is “Universal Broadband Broadcasting”): http://fjallfoss.fcc.gov/ecfs/…021027035. And almost exactly a year ago, Ulrich Reimers from the Technical University in Braunschweig gave a talk on “The Prospects of LTE (Long Term Evolution) for the Delivery of Broadcast Services” at DVB World in Lisbon. David Wood, who attended that talk, posted these notes in his blog: “Ulrich Reimers dared to ask the question… Can the next generation mobile phone system, LTE, be used for broadcast services? Ulrich explained that a lot of the LTE elements come from DVB systems… Actually, Ulrich calculated that DVB-T2 is about 50% more bit rate efficient than LTE… LTE networks need very dense networks, dramatically smaller than DVB-T Ulrich concluded that although LTE is more efficient than UMTS, as a means of providing a ‘broadcast’ it is much less efficient than DVB systems, and cannot be used for high quality video delivery. Are we saved?” (from http://dvbworld.wordpress.com/…redundant/)
I think your right. The key issue is probably not if DVB-T or LTE are more efficient on the PHY-layer, its rather the broadcast/unicast issue. If your assumption is that everyone in some area is going to watch the same set of channels (“Linear TV”), then (broadcast) DVB-T no doubt is the most efficient way to provide the service. However, if your task is to cater for thousands of users in “video-on-demand” mode with a variety of tastes, you need an unicast capable solution (if you can afford it). Further, if you deploy an LTE system, you are likely to do this as an two-way IP access system which considerably requires more BS:s -in particular if there is an uplink to handle. Just copying the broadcast mode of DVB-T is meaningless – it does not add anything new.