Broadband: Some rural areas will lose out

crawler

http://www.rte.ie/business/2009/0311/eircom.html

An Eircom executive has said that people in rural Ireland who live more than 5km from a telephone exchange will never get broadband.

Eircom spokesman Paul Bradley told the Oireachtas Communications Committee that even when the local exchange is upgraded to handle broadband, after 5km the signal becomes so weak that a modem will not connect. He said it was a limitation of the technology.

Fine Gael's Noel Coonan accused Eircom of abandoning rural Ireland. But Mr Bradley said Eircom has now set up broadband in exchanges where there are between 300 and 400 customers. He described these areas as 'very rural'.
Advertisement

Another Eircom executive, Pat Galvin, said that even when these exchanges are connected, some communities will have to rely on slower broadband because of costs.

Mr Galvin said it would not be economical to extend fixed line broadband nationwide and that some parts of rural Ireland would have to reply on wireless broadband, which is slower.

Earlier, Mr Galvin told the committee that even though the company has a large amount of debt, Eircom has not been hindered in investing in better broadband. He said the company's debt was manageable.

He was responding to a question from Fine Gale's Simon Coveney who said there was a perception that Eircom's parent company was broke and that Eircom would not be able to pay for next generation broadband.

watty

We have always known that more than 5km that Fixed wireless is best option.

In reality it's possible to get faster lower latency Fixed Wireless at 20km than DSL at 4km, it there are not too many subscribers on it. The Issue of speed with Fixed Wireless is only Spectrum and Cost to lower contention. The issue with DSL is it needs repeaters.

Why can't they use a repeater at 4km and that extends the 4km speed to 8km?
-- Cost.

fergalfrog

This is currently being debated on Radio 1...

(Noel Coonan and Paul Bradley)

bealtine

crawler wrote: »

Mr Galvin said it would not be economical to extend fixed line broadband nationwide and that some parts of rural Ireland would have to reply on wireless broadband, which is slower.

I'd have to disagree with this comment, some wireless providers offer excellent service and speeds, equal if not better than most of eircom offerings.

But then he could have been referring to the NBS...

Sponge Bob

This is eircom throwing a hissy fit in public at not getting the NBS contract

Rural people piss off , take what we give you vibe .

Fixing the most expensive lines in the World to an adequate standard is the last thing on their minds, what exactly do we get for all that line rental anyway ?

lucernarian

Eircom have also said that they will try to give DSL to anyone within 8km of the phone exchange as the wire goes. In my experience (personal and what I've read here) the 8km limit is quite realistic.

The amber line program has given broadband to an awful lot of people.

Also the bit about wireless broadband being simply slower than DSL is just nonsense. That looked like a swipe at the NBS.

yellowcurl

I live 5 mins outside a town in county kildare and we have been trying to get broadband for the last good few years and each time Eircom came up with one excuse after another until they finally admitted that there was actually a fault with our line and that it "wasn't profitable" to fix it. Hmm.... Fixed wireless isn't half as reliable as broadband. I can definitely vouch for that one.

PogMoThoin

yellowcurl wrote: »

Hmm.... Fixed wireless isn't half as reliable as broadband. I can definitely vouch for that one.

Very general there, Its the provider not the technology.

broadbend

Did you know that the Australian company Telstra, installed a brand new 3G Wireless network, now running rates up to 21Mbits/sec.

They claim coverage of 98.8% of the population.

Now some of the Australian population is VERY REMOTE, maybe even several thousand miles from a telephone exchange!

So how did the Aussies solve the Broadband to the Bush problem.

Yes, Satellite was one option, but they were smarter.

A company called CALYPTECH (www.calyptech.com) have a VERY neat solution, that Telstra deployed in the bush, called an INTEGRATED WIRELESS TERMINAL.

Essentially, this terminal, connects to the 3G network, and can drive a copper telephone pair, to the customers premises. This unit is put in a cabinet, with a high antenna to pick up good signal from a 3G tower up to 40km away, and drives ADSL2+ down the copper!

All powered remotely with Solar/Battery.

Surely Ireland can use this technology??? What do you think?

-Broadie

bealtine

broadbend wrote: »

Did you know that the Australian company Telstra, installed a brand new 3G Wireless network, now running rates up to 21Mbits/sec.

3G will never deliver anything approaching broadband, never in their wildest dreams...
So any "imaginative" solutions are a waste of money basically.

crawler

4G might...or something that looks like BB, Sometimes.

watty

I did the sums on 100Mbps LTE and 21MBps LTE.

About 3% of cell area (rural non-MIMO) gets the top speed. About 50% area gets the 1Mbps (21Mbps HSPA or LTE 5MHz FDD) or 4Mbps on 100MBps LTE (20MHz FDD spectrum).

That's one user.

With 20 to 40 users evenly distributed simultaneous download:
half the users get
(4MBps / number of users) for LTE@ 20MHz -- 200kbps down to 100kbps
(1MBps / number of users) for LTE@ 5MHz -- 50kbps down to 25kbps
(0.5MBps / number of users) for HSPA -- 25kbps down to 12kbps

Loaded heavily the LTE is twice as good as HSDPA/HSPA/HSUPA as it OFDM rather than CDMA.

For smaller Urban cell with 4 x4 MIMO you can have 4 times the LTE capacity closer to mast. If you have N=9 spectrum (unlikely unless there is one national operator), then Urban cells can overlap to make loaded speed closer to 1Mbps per person than 2Mbps per person.

HSPA can do 1 to 2Mbps loaded cell throughput. At best rural LTE in 5MHz does 2Mbps to 4MBps loaded cell throughput. N=9 Urban 4x4 MIMO (very expensive in gear and licence) can manage maybe 20Mbps loaded sector throughput, i.e. 1Mbps on average for 20 simultaneous users.

Any flaws in my analysis Crawler? (all round approximate figures)

Because fixed wireless has the LTE Mbps/MHz centre of cell speed to edge of range, its sector throughput is close to its peak speed, about 25Mbps with no MIMO, normalised to 5MHz channel.

So Metro was outperforming hypothetical future LTE by more than x10 , for a whole sector on average, four years go.
http://irishwattystuff.com/comparewireless/CompareHSPAandFixed-v4.html

Also mobile omnidirectional aerials very constrained on uplink compared to power of TX and gain of fixed outdoor directional aerial of Fixed Wireless. Unless every 3rd street lamp is a Wireless base, mobile can never deliver BB. It's mathematics and physics, not technology issue.

crawler

Watty - looks about right based on the assumptions you have made. Naturally if you move from a 5Mhz channel to a 20Mhz one then you need to look at the higher interference issue - you will not get a perfect X4 uplift.....to be honest the problem for LTE will probably not be the technology, it will be the available spectrum in the right band and (unless the can get a nuclear power plant into a mobile) the Tx power and antenna size and type in the UD.

I like MIMO and I like xOFDM (which is actually an OLD technology) - actually if you look at the next advances in DSL (Ericsson got 500Mbps on 6 pairs over 1km on vectorized VDSL and in theory could get more on fewer pairs) they use MIMO along with cross talk elimination techniques. N reuse will also be impacted by the frequency used and the antenna beam width and also possibilities for frequency stacking - there is also the possibility of more tones being used so it could be squeezed.

Like all things though, the laws of physics remain a constant and even improvements introduce new problems, in telecoms things ALWAYS get worse

In all LTE is a move in the right direction but it too has suffered from the standards bun fight between it and UMB (only a year dead and almost forgotten already - poor Qualcomm!) - good parts were left out of the standard due to heavy lobby pressure....

watty

83 MBps per pair @ 1km. We had that, what 7 years ago? Simply the DSL pair bonding equivalent of DOCSIS 3.0 channel bonding. No actual extra speed per resource. (DOCSIS 3.0 creates no additional capacity, just lets one user have all of 2 to 5 channel)

But at Irish line rental prices who would want two lines never mind six ?

bealtine

crawler wrote: »

Naturally if you move from a 5Mhz channel to a 20Mhz one then you need to look at the higher interference issue -

Where is the bandwidth supposed to come from? (magic?)
All the relevant bands are fairly crowded already.

In all LTE is a move in the right direction but it too has suffered from the standards bun fight between it and UMB (only a year dead and almost forgotten already - poor Qualcomm!) - good parts were left out of the standard due to heavy lobby pressure....

That's the main reason we have a CDMA based system rather than say a TDMA based system, stupid lobbyists

crawler

Re bandwidth- EXACTLY the point - where will all the spectrum come from?

There is a fair chunk with Digital Dividend and IOFFL need to ensure it remains technology neutral and not sent off to "standards/vested interest bun fight jail" forever - Digital Dividend consultation V important

crawler

watty wrote: »

83 MBps per pair @ 1km. We had that, what 7 years ago? Simply the DSL pair bonding equivalent of DOCSIS 3.0 channel bonding. No actual extra speed per resource. (DOCSIS 3.0 creates no additional capacity, just lets one user have all of 2 to 5 channel)

But at Irish line rental prices who would want two lines never mind six ?

Which is my point on physics

On the six lines Watty, I would pay for it at the distribution network layer if it was a choice between that or a leased line At the access layer, it is too rich for ressie alright but not for Business. DSL migrates to DSM eventually anyway (as you said, this type of stuff is around a long time, just not commercial in any great way)

watty

Leased lines here are funny money. Indeed 10 DSL lines could be cheaper

trekkypj

Hmmmm I'm just wondering how much would it cost to replace a rural network with underground fibre?

For example:

Let's suppose the government got funds from the EU to create jobs and stabilise the economy by investing in infrastructure. (pie in the sky i know but just suppose)

Couldn't you do a joined-up thinking approach where the next time the roads were re-done, the council could install ducting for use by the ESB and telephone utilities? This would hugely cut the cost of laying fibre, I think.

You could also make it a requirement for one off houses to install standardised ducting from the house to the road to join up with the network.

There's so many things that could be done.... if only there was the will. The rural networks (copper on poles) are rotting and needs to be replaced - why not just put the ducting in now and save the pain later?

Tom Young

Circa 300 Million. I think someone has done a study on it - May have been Magnet Networks.....not sure.

trekkypj

Tom Young wrote: »

Circa 300 Million. I think someone has done a study on it - May have been Magnet Networks.....not sure.

Hmmm well it might not be feasible everywhere but if you used common utility ducts for the ESB and the fibre and water and all wouldn't that cut costs right down?

Kilkenny Co. Co. are redoing our local village, spending tons on a new car park, playground and roadworks in the village, and the exchange is slap bang in the middle - you could incorporate that sort of work into the scheme and remove the poles.

But are they doing it? :pac:

PogMoThoin

You could wrap fibre optic round existing ESB cable, the state still own this network and they don't leave out houses that are hard to reach

crawler

Loads of ways to do this - armoured fibre or siamese cables being one way - exisitng trunk lines, exisitng power lines, narrow bore technology (Witch ditch), smelly fibre - ya da ya da ya da - problem is lack of vision and joined up thinking.

Better still would be muni type schemes and local ownersip with grants etc etc...

So many ways....