Join 3,574 readers in helping fund MetaFilter (Hide)


European Broadband
August 9, 2012 2:03 PM   Subscribe

Why is Western Europe dragging its heels on implementing Fiber based broadband?

I was perusing this .pdf (slide 5) and was really really surprised to see how slow the incumbent telcos in much of Western Europe are at implementing Fibre. I recognize that the precedence of Satellite over Cable has take away some of the competitive angle we have in the US, but I would think the governments would be "encouraging" their respective telephone monopolies to get their act in gear.

What gives?
posted by JPD to Computers & Internet (18 answers total)
 
I think you might be working off (30 year) old data re: "telephone monopolies". Most of the countries I know of in W. Europe have competitive private companies providing internet/cable services just the same as the USA. My guess is that they're not investing in infrastructure due to the economic problems and are happily taking in the money from their current customers who aren't excited about spending any more money.
posted by merocet at 2:48 PM on August 9, 2012 [1 favorite]


What telephone monopolies? The EU opened up the telecoms market quite some time ago, and the European Commission's DG Competition keeps a beady eye on the legacy operators. Indeed, this is an important factor slowing down the introduction of fibre, as those operators not-so-subtly lobby for a looser competition rules before investing in upgrading the networks.
posted by Skeptic at 2:48 PM on August 9, 2012


This article gives some more colour about the complex interplay between DG Competition, the national regulators, the incumbent operators and the newcomers.
posted by Skeptic at 3:12 PM on August 9, 2012


Look at that presentation. For the most part the incumbents still own the local loop and they provide the bulk of new fibre hookups.

(yes I should have used incumbents, not monopolies)
posted by JPD at 3:17 PM on August 9, 2012


Yes, JPD, but the incumbents complain that the EC doesn't let them charge the new entrants enough for access to the local loop, and have been withholding investment in upgrading it.
posted by Skeptic at 3:28 PM on August 9, 2012


The telcos in the US aren't building out fiber anymore, except where they are required to by agreements they got into when it was Teh Hotness and where people thought the money would be.

If you're a huge telco like Verizon (or any of its Euro equivalents), given the choice between investing in wireline and wireless infrastructure, there's a lot more money to be made in wireless. The profit margin is phenomenally larger.

I think it's just the case of European operators having taken a slight wait-and-see approach, and deciding that they didn't like what they saw. US operators -- really, Verizon -- mostly got into fiber-based services due to the heavy competition in the HSI/phone/TV market by cable companies -- Comcast and Cox, if we're naming names. In Europe, the cable TV companies aren't as dominant and didn't drive the phone companies to make the investment as early (from what I understand), and also there's a lot more DSL options which made fiber less attractive. In the US, Verizon had to deploy fiber in order to compete with coax in many locations; the existing copper infrastructure wouldn't cut it.
posted by Kadin2048 at 4:31 PM on August 9, 2012


Skeptic do you have a cite for the underpricing of local access (cite in the "Hey that sounds interesting and I'd like to learn more" not "I don't believe you and think you are lying" sense)

Yes my hypothesis from afar was that the lack of a cable competitor in most markets is why there has not been a market driven investment. I like to describe VZ's strategy as "investing enough to remain a credible threat to CATV operators". But even with that broadband seems to have penetrated enough in the US for people to being cord cutting and replacing Satellite with something else. I'm wondering if the same thing will happen in Europe wrt to Satellite providers if the governments "encourage" investment in fiber.
posted by JPD at 4:44 PM on August 9, 2012


First of all, the graph shows a count of new deployments by country - not scaled by population. America has about 5 times Britain's population, so the figures for BT should really be scaled up from 3 to 15. Still lagging behind Verizon's 790, obviously.

BT is a big player in the UK internet access market - they were formerly state-owned, and they're subject to a Universal Service Obligation meaning they have to provide service to customers out in the wilderness even when it's uneconomic to do so. Their main technology is ADSL which runs over existing copper telephone lines. Due to local loop unbundling, there are various other companies selling the same service over the same telephone lines. Advertised performance is 16 Mbps downstream, worsening as you get further from the exchange.

The other major player is Virgin Media. They have a network of buried coaxial cable reaching about half of Britain's homes, almost exclusively in urban areas. They have about 25% of the broadband market. No local loop unbundling, so this is only available through Virgin themselves. They market their service as "high speed fiber optic broadband" but really it's fibre-to-the-cabinet which is a lot cheaper to install and doesn't count in FTTH lists. Each street cabinet serves a few hundred homes. Virgin advertise speeds up to 100 Mbps.

None of the services offer the advertised speeds on average, but Virgin reportedly get closer than anyone who offers service over ADSL. On the other hand, Virgin have a reputation for selling customer data to advertisers and blocking sites all willy-nilly; some ADSL suppliers have a better reputation in this regard.

According to this short but informative report the main reasons not to move to next generation broadband (perhaps in the form of FTTH) include:
* There's limited evidence of demand, as few customers opt for the highest speed service packages with current broadband technology.
* If we wait, better technology may become available.
* It would cost £10 billion to get FTTH to 90% of homes, and looking at what customers are willing to pay for better performance, it would take a super long time to break even.

Also, the report doesn't say it, but BT is heavily regulated, e.g. forced into local loop unbundling and regulated in terms of what they can charge. So they go into debt building a bunch of infrastructure and they might be forced to let their debt-free competitors use it too. They probably figure by dragging their feet they can get a better deal from the regulators.

So in summary, why is the UK lagging behind in that graph? Because the graph doesn't include fiber to the cabinet or Virgin Media, and because the two main companies don't think spending a bunch of money on a FTTH network would be a good investment.
posted by Mike1024 at 4:49 PM on August 9, 2012


Could you do TV/ VOD over VDSL?
posted by JPD at 4:49 PM on August 9, 2012


According to theOECD fibre + cable (which is mostly fiber to the block/cabinet no?)on a per 1000 person rate
US 17.4
UK 7.3
France 2.5
Germany 4.5
So its not a question of scaling. (also Verizon's service area's population is more like 100 mil, so much less than the cumulative populations of those other ILECs on that page.)
posted by JPD at 5:05 PM on August 9, 2012


> could you do TV/ VOD over VDSL?

Yes. In the USA, AT&T's "UVerse"-branded voice, TV and Internet services use VDSL2 for communication between customers' gateways and the local node.
posted by BrandonW at 5:06 PM on August 9, 2012


BT have conducted a significant roll-out of FTTC across the UK in the last year. It seems adequate for current needs. I about 30Mb/s down and about 6Mb/s up, (advertised at 40/10). I'm not sure where you'd currently find significant customer demand for speeds much above this at a higher cost.
posted by howfar at 7:06 PM on August 9, 2012


Skeptic do you have a cite for the underpricing of local access (cite in the "Hey that sounds interesting and I'd like to learn more" not "I don't believe you and think you are lying" sense)

This Reuters article is quite good.
posted by Skeptic at 10:44 PM on August 9, 2012


Could you do TV/ VOD over VDSL?

I get triple play (Internet, phone and TV/VOD) over ADSL. There are some bandwidth problems, because I'm relatively far from a node, but it works. The local cable operator offers a fibre link, and I'm trying to change to that, but I'm getting some problems (the cable shafts in my building are apparently cluttered).
posted by Skeptic at 1:49 AM on August 10, 2012


The issue has been ramping up in importance here in the UK, at least. By some measures the UK is well served relative to its present demand, but is apparently very poorly served in terms of potential, which, it is suggested, will stifle creativity and economic activity.

There are definitely issues around universal access. There are also big costs involved in getting fibre down the last mile.

An editorial in this Sunday's Observer talks about some of the political and commercial issues.

FWIW, I live in a rural location and our DSL broadband just clocked in at a disappointing 1.8Mbps. We were 'upgraded' within the past six months. Prior to that, we got about 1.4Mbps, so clearly not a transformative upgrade.
posted by sagwalla at 2:41 AM on August 10, 2012


Because VDSL and DOCSIS are good enough? FTTH is still a rare thing in the world. In the US only Verizon is doing it and only in its Verizon territories and it still has not reached 100% of those territories. I'm in AT&T country, so I will never get Verizon service here, but I can get Uverse which is a VDSL based product. Google has gotten into the game, but one have one trial city.

VDSL isn't bad. Its essentially a Fiber to the Curb scheme. AT&T runs fiber to every neighborhood and then the VDSL box plugs into there and the customer connects to it via short run of twisted pair running VDSL. I think AT&T is able to get 30-50mbps, but some of that is partially allocated for TV service. I can buy as high as 24mbps of internet from AT&T. The VDSL2+ spec claims a maximum of 250mbps at short distances and 100mbps at 1 kilometer.

I think VDSL2+ technology is going to be very attractive for telcos with copper everywhere. Running new fiber is just expensive. They probably want to milk another 10+ years out of their cable infrastructure. Who knows VDSL3 or whatever the future hold may allow higher speeds and some countries could be on copper for a very long time.
posted by damn dirty ape at 7:42 AM on August 10, 2012


Is there some theoretical maximum bandwidth as a function of distance that can be put through copper - like violates core principles of physics sort of thing? Or could they just keep improving the equipment at the two ends of the copper and keep increasing capacity? Not that they know how to do this yet, but there is no hard barrier?

In other words if you could magical eliminate the loss from distance that comes from copper could it carry as much data as fibre?
posted by JPD at 9:40 AM on August 10, 2012


Probably not. There will always be noise on copper lines. Fiber is largely noiseless because its immune from EM interference.

I'm sure there's some Shannon's law calculation about the maximum theoretical limit, but the practical limits seem to be stuck on our ability to do digital signal processing (both from a hardware and algorithm perspective), usable frequencies, and lots of little things like legacy concerns, price, regulation, etc.

The neat part is that vanilla DSL in 1999 was around 1.3mbps and now VDSL2 floats around 200mbps. If you're a kilometer from the CO in 1999 you probably could get a solid 500k connection. Now you can probably get a solid 50-100mbps connection. That's a 100x improvement.

I also suspect there's a pretty huge law of diminishing returns here. If today you can get 30mbps to your home for $30 a month, but in a couple of years can get 100mbps for $100, would you even bother upgrading? There might be a natural consumer speed where if movies and software delivery are percieved to be fast enough then they won't really need to upgrade too often. Kind of how like 9600 baud was the defacto modem speed for many years. Something tells me if your customer can get 50mbps from copper than why bother with 100 or 500mbs fiber? Grandma isn't downloading many 5gb linux ISOs.
posted by damn dirty ape at 9:58 AM on August 10, 2012


« Older Can I play Bioshock on my Macb...   |  I'm getting married soon, and ... Newer »
This thread is closed to new comments.