Chambers of Commerce Demand a Universal FTTH system In Ireland Costing €2Bn

Sponge Bob

This is a MUCH better idea than IBEC s Tommy McCabe giving us crappy Satellite services and spending €5Bn to deliver that third world network he envisages .

Nor do the Chambers of commerce ponce about with 5Mbits like Tommy does , their 2Bn plan delivers speeds in the 10-100Mbits range.

The suggestion , a great idea in its own right and well worth putting out there for proper discussion (even if €2Bn is optimistic ) , came from their submission to Forfás on what to do to fix Ireland dismal BB availability and is covered on Silicon Republic here

He ( Seán Murphy) also noted that such a significant investment by the Government would not be the first of its kind in the history of the state. “We estimate the cost to the Government of delivering fibre optic cabling to all homes and businesses in Irish towns and cities to be around €2bn. This should be viewed in the context of the multi-billion euro multi-annual investment current ly being undertaken by ESB to upgrade its network.”

Quite correct . Their full Full submission here and fair play to the Chambers of Commerce again

watty

Sponge Bod for Teashuck! Much better idea than Satellite. Keep Satellite for our FTA or FTV RTE TV broadcasts. Like Croatia, Palestine, Slovenia, Morroco, Greece, Tunisa, Wales and all those other rich countries have.

Sponge Bob

Keep satellite for broadcast systems Watty, not for **** latency 2 way cackery packaged up as Broadband. FTTH is the real deal though

There are 1.5M homes in Ireland, who reckons they could all be fibred up for €2Bn (thats about €1330 each ) ???? .

It would be more feasible had the building regulations been modified by Dempsey and Roche in 2002 as suggested because 300,000 homes have been built since and should have ducting installed .....but don't.

It could deffo be done for €5Bn and we could also built a comprehensive footpath network along our regional roads while we were at it. !

watty

They could do it for less if they used the phone pair to pull in the Fibre.

paulm17781

If they could be done for €1330 I would pay that myself!

Sponge Bob

twould be an interesting project on the lines of the information age towns.

set up one of each of

1. a rural village
2. a more compact urbanised village , small rural hinterland
3. a town up to 15k persons , urban only

and then have a national competition. maybe have one of each category in each of the BMW and the SE regions.

The winner in each category gets 100% FTTH done as an experiment and then have a beauty competition among carriers to see who wants all the business on those networks and what they are prepared to offer .

Worth trying to find out what it would cost . In deference to the original idea the towns in category 3 must have a Chamber of Commerce before they bid.

Hows that !

Duiske

Why have the Gov never taken the initiative on this before ? 2bn, which in relation to GDP is not a huge amount, sounds pretty reasonable, and would in one sweep put Ireland up there with the league leaders in DSL availibility.
Going by past record, this report will probably be kept in Minister Dempseys office bathroom, in case of emergencies such as Industrial unrest at Andrex !!!!

cgarvey

Well the €2bn is to cover "Irish towns and cities", so that still leaves the availability issue to be sorted. If we used proper regulation, we can sort out all the line failure issues (over time, granted, but make it an agressive, yet fair, timeline). Which leaves only the 10-15% of people who can't get DSL/cable due to technical limits (not shoddy ericom practices), where wireless can be one of the solutions to bridge. That doesn't require €2bn and does get us fair access to proper BB for all.

Hell, let's just start with dialup. What's unreasonable to suggest that if you can't get DSL on your phone line, you pay the lowest (non-timed) DSL retail price for unlimited dialup calls? Can ComReg make that call, legally? Surely they could as part of a fair USO consultation/compromise? "OK we won't require you to have 80% of lines capable of carrying DSL if you give us flat rate".

.cg

Sponge Bob

Irish towns and cities can be done just as well with HFC (Hybrid Fibre and Co ax for the last few 100m) ....as long as guarantees are given by the cable operators on Universal Service, if the USO were say 10mbits for €19.99 a month by end 2007.

NTL have done more or less most of Dublin to something capable close to 5Mbits for about €50m tops in Capex over the past 3 years . More advanced cable systems like Dungarvan and Longford do 9Mbits already

This costs the project at around €2200 per household in towns and cities containing 60% of the population and tells the other 40% who do not live in towns and cities to **** off. Hmmm, not good enough that . We need Universal FTTJH or a mix of Universal urban standards to leverage short distances and Universal rural standards....probably wireless.

We need sats for Telly, thats all.

Still, its ambitious , a quality sadly lacking over in IBEC .

bk

Sponge Bob wrote:

NTL have done more or less most of Dublin to something capable close to 5Mbits for about €50m tops in Capex over the past 3 years . More advanced cable systems like Dungarvan and Longford do 9Mbits already

The modems of the oldest NTL BB customers, in the oldest BB enabled areas (Tallaght) use DOCSIS 1 and are only capable of 5m. The network itself is capable of >10m, all customers for atleast the last two years have been getting DOCSIS 2 modems which are capable of >10m. NTL is currently replacing all the older DOCSIS 1 modems with the new DOCSIS 2 modems. That should say a lot about what they are planning to do

You are right, I recently read a very interesting article that explained that a HFC network using wideband DOCSIS technology can deliver speeds equal to or greater then what is realistically possible with FTTH, for a much cheaper price.

Interestingly the article also explains how shared local contention can actually be an advantage!!!!

watty

Sponge Bob wrote:

This costs the project at around €2200 per household in towns and cities containing 60% of the population and tells the other 40% who do not live in towns and cities to **** off. Hmmm, not good enough that . We need Universal FTTJH or a mix of Universal urban standards to leverage short distances and Universal rural standards....probably wireless.

We need sats for Telly, thats all.

Agreed, Cable for dense areas, adsl for anything not to far from exchange (all except rural), Wireless (Metro, Wimax, not mickymouse wifi style stuff) with POTS ports in rural and some urban.

And communal multisat dish / distribution on any apartment block with more than 16 units. Definately Sat Birds for Telly! Tenants should have choice of Sat AND/OR cable in apartments with no more than 50 Euro a year for right to connect your own sat receiver.

Blaster99

I suppose FTTC/vDSL is a good start on the way to proper FTTH, but it seems to me that most forward-thinking countries are going for FTTH directly.

bk

Blaster99 wrote:

I suppose FTTC/vDSL is a good start on the way to proper FTTH, but it seems to me that most forward-thinking countries are going for FTTH directly.

Very few are going for true FTTH. Verizon is about the only company who is. In asia, they are doing Fibre to The Apartment building and then delivering to each apartment via Ethernet or Coax.

A good quality HFC network will give you much higher speeds then FTTC/vDSL, speeds equivalent or above those delivered by FTTH, for the same price or less then FTTC/vDSL.

Blaster99

I presume you're talking theory, which is fine. It's just that a lot of houses has copper, not coax. And if you're going to dig to deliver coax, you may as well do fibre.

I suppose there's another factor, which is that in reality HFC networks don't deliver the speeds of vDSL or FTTH, at least as deployed today. The latter are usually 50Mbps and 100Mbps respectively, the former 10Mbps. I guess FTTC + coax could give FTTH a run for its money, but I'm not sure I've ever seen that deployed anywhere. I suppose FTTB + coax is a variant, but that's essentially a fibre network with a bit of coax at the end of it. Hardly what one would call a cable network...

lucernarian

I think bk is referring to the current capabilities of cable vs a single fibre pair.

A fibre pair does not really operate above 2.5 Gb as can be seen on the ESB network and on the MANs aswell I think.

If cable dedicated all the channels available to it to broadband using channel bonding or whatever it would exceed that figure.

SeaSide

To_be_confirmed wrote:

A fibre pair does not really operate above 2.5 Gb as can be seen on the ESB network and on the MANs aswell I think.

2.5Gigs would be one colour on a fibre and you can get several hundred colours onto the one fibre

Sponge Bob

Blaster99 wrote:

I presume you're talking theory, which is fine. It's just that a lot of houses has copper, not coax. And if you're going to dig to deliver coax, you may as well do fibre.

The cost of repairing / joining pure fibre is a lot higher than for co ax .

Co Ax need not be dug, it can be hung from Irish semi to Irish semi. Fibre is difficult to hang as it is not as resilient as co-ax and faults more easily if blown around and stretched .

Co ax can carry voltage which may be needed for powering the CPE in certain cases. With fibre you must also present the CPE where power is available.

Docsis 3 cable will do up to 100Mbits which is what the Universal target should be in urban areas with a population of 1500 or more in Ireland by 2010. Forget about Mary O Rourke and her 5Mbits bull in 2002 . Its not worth investing in for urban areas now because technology has moved on .

Once you set the target you then invite proposals and costings to achieve the target on the basis of Universal Provision . It would cost the government feck all to invite proposals on a non binding basis and to find out if such a project is worth tendering out.

All it takes is ambition ....like the ambition that was evident in the late 1970s that originallygot us a telephone network by 20 years ago . Regrettably the ambition to privatise this asset led to the onset of deterioration and under investment starting in the early 1990s.

This is what haunts this country now and stunts is development as surely as the crapheap that was the Post and Telegraph network in around 1979.........and which took c. 8 years and about €2bn to sort out back then .

watty

You only need fibre to the end of each street.

Blaster99

I'd say it doesn't really matter whether you do FTTC + copper or FTTC + coax in the scheme of things. The end result is vastly better than the crap we have now.

Anyways, I think SB's idea of doing a pilot in three towns with different characteristics is an outstanding idea. Surely the price of a couple of Comreg surveys could pay for it or use some of the unused GBS money to pay for it.

Sponge Bob

If nothing else it will start a feeding frenzy nationwide for ultra high capacity BB as towns and williges prepare their cases and pitches and thoroughly diss their neighbours in order to get their grubby hands on the money

That was kinda how the Information age eircom competition thing ran in the mid 1990s (won by Ennis) although

a) it was only Universal PCs (maybe ISDN)
b) it was turned off years ago and
c) was confined to towns over 10k population

bk

Sorry, I didn't mean to start an argument. Realisitcally no one is going to listen to us. The decisions are gonig to be made by companies based on cost and the technology they have already invested in and have expertise in.

So in the long term the telcos will probably try FTTC + Unshielded Twisted Pair and the cablecos will go with Shielded coax.

Blaster99 wrote:

I'd say it doesn't really matter whether you do FTTC + copper or FTTC + coax in the scheme of things. The end result is vastly better than the crap we have now.

However having said the above, from a purely geeky point of view it is interesting to talk about the merits of the various options, even if it has no impact on what is rolled out.

VDSL2+ can deliver speeds of about 40Mbps using FTTC. Coax can deliver up to 5Gbps. This is down to the pure physical characteristics of the cable used.

Unshielded Twisted Pair (telephone cable - you can't really call it Copper, as Coax is also Copper) can deliver a little over 1Mhz of bandwidth. Coax carries about 860MHz of bandwidth. Currently cable uses one single 8MHz channel (out of the 860Mhz) to deliver 50Mbps. All the excitment in the cable world is about DOCSIS 3, this new standard will allow for channel bonding, so you can use multiple 8MHz channels to deliver much higher speeds. Cable companies in the US are seriously talking about delivering 100+MBit speeds using this tech and much higher in the future.

These speeds will never be possible with unshielded twisted pair, purely because of the physical attributes, even with VDSL2+ and FTTC, you won't be able to go higher then 40 - 50Mbits, it is simply the ultimate limitation of the physics of the cable medium. Some companies in the US are talking about running two or even three UTP cables to each persons home and to bond them to offer higher speeds, but that is going to be very expensive.

Here is the thing, laying coax is only very slightly more expensive then UTP (about 70% of the cost for both is labor and civil ground work), so if you are starting with a complete blank slate then coax is a much better option then UTP or even do with NTL:UK did and lay a UTP cable alongside the Coax cable.

As for FTTH, while Fibre has potentially far greater speeds if you were to use all the light frequency (1000Gbps), the reality is such equipment is incredibly expensive and can only be feasibly used for international backhaul. For residential use you can only use a very small part of the spectrum for a realistic cost, which limts the bandwidth to speeds equivalent to HFC. The problem for Fibre is that it is signifcantly more expensive to terminate and lay then either Coax or UTP, while only delivering speeds equal or less to coax using an all IP network.

For those of you who are interested from a purely academic point of view, here is a very interesting article which compares HFC to Fibre:
http://www.iec.org/newsletter/feb06_1/analyst_1.html

Although fiber offers more bandwidth theoretically, in practice HFC can be made to offer similar or even higher useful bandwidth than fiber at a cost that customers and service providers are more willing to bear.

Of course while all of tihs is great fun to talk about, it does nothing to help the poor souls who are still stuck on 56k dial-up and really that is what we need to focus on for now.

Sponge Bob

bk wrote:

However having said the above, from a purely geeky point of view it is interesting to talk about the merits of the various options, even if it has no impact on what is rolled out.

Unless somebody on high shows some ambition you mean .

I would look at it differently. As technology moves on it is possible on a national basis to:

a) set down a service marker . State that the plan is 10mbits in 2007 and 100Mbits in 2010 and that nothing short of that will be entertained.
b) State that this service must be Universal in towns above 10,000 persons or the CSO definition of urban in Ireland which is 1,500 but pick a number .
c) Invite proposals from various industry players to find out how this can be achieved and for how much. If the figures are loopy then forget it.
d) If you want the traditional copper pair (AWG23 or AWG24) you have clearly pointed out to us BK that no matter how close you bring the fibre to the houses the copper itself will only carry 40-50Mbits. It is theoretically feasible that Bonded Pair VDSL3++ will carry 100Mbits at some future stage so do allow for that too
d) Then invite costings from all interested parties on the basis that the project will be for a Universal service by date nn and for cost xx and find out if its worth going to tender based on those proposals.

These speeds will never be possible with unshielded twisted pair, purely because of the physical attributes, even with VDSL2+ and FTTC, you won't be able to go higher then 40 - 50Mbits, it is simply the ultimate limitation of the physics of the cable medium. Some companies in the US are talking about running two or even three UTP cables to each persons home and to bond them to offer higher speeds, but that is going to be very expensive.

Correct, particularly in the absence of a standard at this time, but in 2007 ya never knows.

Here is the thing, laying coax is only very slightly more expensive then UTP (about 70% of the cost for both is labor and civil ground work), so if you are starting with a complete blank slate then coax is a much better option then UTP or even do with NTL:UK did and lay a UTP cable alongside the Coax cable.

Or another option which is to lay one cable to a hose in which is Co Ax , UTP (phone wires shall we say, not as thick as co-ax ) and fibre, all in one encapsulated armoured cable.

Of course while all of tihs is great fun to talk about, it does nothing to help the poor souls who are still stuck on 56k dial-up and really that is what we need to focus on for now.

It is fair to say that there are issues in rural areas but that is NOT a good reason why the 50% of the irish population who could get 100Mbits should not be sorted out in one package.

A Universal Rural scheme would not come in at 100Mbits by 2010 but then again if the rural dwellers all got a co ax out as far as the road from their own houses and if all you had to do 'nationally' or 'regionally' was to follow the road and join (no digging or driling off road, the householder must do that themselves) then the economies would change utterly .

Failing that the basic technology in rural areas would be wireless and we could easily replicate the urban plan albeit at lower speeds.

Instead of 10Mbits in 2007 and 100Mbits in 2010 we would look at Universal 1Mbits by end 2007, 5 Mbits at end 2009 and 20Mbits by end 2011 in rural areas . Thats because of frequency planning and judicious reuse and because you are providing a baseband business for mast owners. A tax scheme would also be needed so that farmers could put up masts on a large scale between now and end 2008 ....with tax relief . It is as important to the country dweller as the tax backed car park is in central Dublin .

The magic word in every case is Universal , current schemes are all piecemeal and fragmented and we would have to leverage economies of scale to do this on a National basis .

watty

Universal rural wireless with 5M is possible.
Up to 200M bit wireless is relatively cheap but is much shorter range (4km compared with 15km for 5Mps 10GHz and 30km for 2Mbps 3.5GHz).

Rural wireless should include a POTS port that supports at least 14.4k modem (alarms, Fax, legacy direct dialup etc) and ideally 33k or 49k. (Chorus Wireless phone supported error free 49k)

Sponge Bob

My main argument is that you really have to have 2 Universal standards, the urban one and the rural one and you invite 2 lots of proposals and costings .

If you set the cutoff for 'urban' at 5000 people in a town ..or upwards of that...you will find that the 'Urban' scheme will benefit 50% of the population and most of the large employers and yer National Spacer Strategy Gateway towns and yer Decentralisation locations to a considerable degree and all in one fell swoop more or less .

While the Mans are going/have gone to a lot of these towns the problem with the MAN is that it is not a final mile solution but a localised backbone network .

watty

Universal (including rural) 3M by Q2 2007 is very feasible. With satellites limited to TV /Radio.


The MANs are pointless, a solution looking for a problem. No small user connectivity and no backhaul in the plan. They would only be useful for high speed B2B in a SINGLE town. Most Internet /data connectivity is in/out of a town not around it.

If each MAN had high speed backbone connection (none do) and a cabinet connected at each street it passed to interconnect with existing phone pairs and cable TV coax, then it could bypass the exchange and cable head ends. But it wasn't designed to do that. The MANs where the silliest comms idea for a country with non-existent user connectivity. It was adressing the wrong problem.

SeaSide

Most of the MANs do have backhaul which was payed for by .gov/EU

CIE fibre along the railway, Western Digital Corridor, ESB.

Only ESB is available showing the negotiating skills of the parties involved

watty

It wasn't in the original design and AFAIK some are still arguing with Eircom over price.

If it wasn't for ESB fibre a lot of the ISPs, especially wireless ones would be in serious trouble due to the anti-competitve stance of Eircom. I think maybe Eircom should be split in three:
* Calls / Retail / retail line rental / retail BB etc
* wholesale LL and Exchange
* wholesale Network services
They have too much of a strangle hold and quoted insane prices to connect up the MANs to Internet.

Sponge Bob

watty wrote:

The MANs are pointless, a solution looking for a problem. No small user connectivity and no backhaul in the plan

The national backhaul relies entirely on the ESB fibre.

Where you have both ESB fibre and a MAN , in a town, the proposition works . The proposition is in itself limited so we now need to look at a Universal system not at a system that is available in Industrial estates and Business parks but not in residential areas .

The ESB/MAN combination explains why you have Smart and Magnet (amonst others) live in Galway city today .

watty

It's mad that a semistate electricity company is doing more for Broadband than the ex-semistate telecom company. I hope it isn't sold to some carpet bagging asset stripper who reduces the electricity price and doubles the "line rental".

Blaster99

SeaSide wrote:

Most of the MANs do have backhaul which was payed for by .gov/EU

CIE fibre along the railway, Western Digital Corridor, ESB.

Only ESB is available showing the negotiating skills of the parties involved

Isn't that CIE fibre actually owned by BT?

paulm17781

watty wrote:

I hope it isn't sold to some carpet bagging asset stripper who reduces the electricity price and doubles the "line rental".

As opposed to now when price inreases are frequent anyway?