Cellular architectures not great for TV: study A study by the KTH Royal Institute of Technology in Sweden has found that cell-based distribution of broadcast TV content can be a glutton for spectrum. Around the world, mobile network owners have their eyes on spectrum being vacated by the move to digital TV, and the received wisdom is that video is going to keep driving …
The whole point of broadcast is the huge number who are happy to watch an event simultaneously, either because it is convenient or it is a special "live" event.
But don't knock the study too much, I bet they knew the result all along but needed to have the technical evidence so those who make the license decisions could see the downside to handing over all of the spectrum to broadband companies when then promis to deliver TV-like service as part of it.
Yes, it needs to be said. It may be stating the obvious but Governments and Regulators are ignoring the Obvious.
Mobile, Satellite, Fibre, IP etc is all wonderful but complementary to Radio & TV Broadcast.
It was greedy short-sighted stupidity to sell off part of the TV spectrum to Mobile. That part of spectrum is only good for larger cells and isn't enough channels big enough for decent 4G any better than 3G. A total waste.
Terrestrial TV needs the spectrum to compete with cable & satellite TV, which never can be portable or instantly available at a new location. Satellite is very vulnerable and cable only viable for Urban areas, both are tethered so though "Broadcast" are not replacements for Terrestrial TV, despite Ofcom's and Comreg's agenda to abolish Terrestrial TV entirely.
As we read this, Ofcom and Comreg are planning to reduce DTT further to sell off a 2nd block (even poorer mobile capacity and larger cells).
Mobile to give decent 4G coverage and speed needs MORE cells than 3G. This isn't possible on UHF.
Beamforming works best on higher frequencies and smaller cells. Planning permissions and cost mitigate against any useful "Beamforming" on UHF. You need huge aerial arrays and much more electronics.
MIMO and Beamforming is way hyped and available on WiFi, 2G and 3G also but rarely deployed on Mobile.
When you have small cells (= higher capacity per area) then 2.1 to 2.6GHz bands are fine. There should be nearly 200MHz at 2.6Hz. Since the aerial size is less than 1/3rd size of 800MHz, MIMO and Beamforming is more practical. Also BECAUSE the 2.6 GHZ signal can't travel as far as 800MHz, frequency reuse to have higher capacity is 10 to 20 times more capacity.
But operators are more interested in maximizing revenue than performance and the Licence conditions let them. The most efficient use of any band is a SINGLE wholesale RAN, not Ofcom's and Comreg's pack in as many operators on separate channels (more licence revenue). On average a single RAN gives about x4 capacity or speed compared to splitting up the band the way Ofcom and Comreg do.
This has been wishful thinking since day one. While the bandwidth to broadcast dtt or satellite remains constant irrespective of the number of viewers, if each and every one of them wants their own personalised TV then you either need a hell of a lot of distribution, or you, er, broadcast...
Reading the paper, it looks like they did assume broadcast on LTE for broadcast-type networks, thus using some of the LTE bandwidth stolen from traditional TV for that. Even so, the numbers don't look good. A DTT network is now extremely well optimised for its purpose, so it's not too surprising that a network optimised for more general-purpose communication, including mostly single station to single station communication, would not compare favourably on spectrum utilisation.
This is a study that might as well have said that fire is hot and ice is cold. A single stream broadcast at the multi-KW level is going to cover a much larger area with a single transmission. This iisn't rocket science. It is how the networks are built with different goals in mind.
It's proving a negative and (very occasionally) it produces counter intuitive results.
But not in this case. Which is good.
Now will Ofcom take a blind bit of notice?
Thumbs up for proving the idea is b***ocks, not just thinking it is.
A network designed for broadcast is better suited for broadcast than a network designed for unicast. Wow.
Also the much bigger problem is that cellular networks tend to include brain dead DRM which tends to be incompatible with the way people use TV since the advent of widespread digitalisation. (ca 2000)
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
