Broadband to CLAR

thegills

Yo,

There was a Call for Broadband Proposals in last Tuesdays Irish Times.

"A pilot trial of broadband technology and services in selected industrial estates, business, educational and training centres in the CLÁR areas has been identified as a key factor in promoting the economic and business development of these areas"

15 locations have been earmarked for e500k - Cootehill, Tulla, Ballingeary, Rockchapel, Kilcar Industrial Estate, Clifden, Carna Industrial Estate, Milltown Industrial Estate – Dingle, Mohill, Carlingford, Corraun, Ballycastle, Boyle, Tubbercurry, Lismore.

Don't ask me how they select the locations. Projects to be implemented by end 2003

Any thoughts on the matter?

spongebob

Yes. They all need proper wireless backhaul too or else an Eircom 1-2Mbit HDSL delivered on schedule (6 weeks) of course :D .
Otherwise they will end up with the VSAT/ Wi-Fi solution that the SWRA is adopting in similar locations in Cork and Kerry.

It strikes me that a VSAT/Wi-Fi solution in an area with no satellite and wireless expertise whatsoever is akin to making the locals pass the Driving Test in a Ferrari ! Inappropriate and overcomplex technology is a no-no in my opinion.

Then again the Clár areas are desperate for anything that could work seeing as Eircom have never done anything for them so its a dificult one to call IMO.

M

antoinolachtnai

do you know the people involved muck?

sounds like some interesting technical fun to come up with a simple solution.

spongebob

I do know the people involved , I have to give them a better solution before I tell that that their plan is inadequate.

So I am working on it, believe me.

M

antoinolachtnai

groovy enough. let me know if i can be of any help.

spongebob

an article somewher on the maths of VSAT and W-Fi schemes, costing about 12k in equipment terms , being subsidised at 30k each by the government , would be nice if you get around to it.

The vsat links are worse than 64k isdn once shared.

it is NOT Broadband, it is a scabby form of always on.

M

Urban Weigl

Originally posted by Muck
It is NOT Broadband, it is a scabby form of always on

What VSAT provider/technology were they planning on using; Hughes, Satlynx (formerly Gilat) or Aramiska? I have heard that Aramiska works very well in a shared environment, though obviously it would not be suitable for gaming due to the latency (400ms quoted by Aramiska).

spongebob

Aramiska is supposed to be the best but it is ferociously expensive , see Here

antoinolachtnai

well, it's not too expensive if you share it. Service is still crap because of latency though.

Here's my view: you have to find some way of having a terrestrial link to the head-end. Ireland is neither that big nor that mountainous, so this shouldn't be difficult.

You just have to find a point within wireless range (maybe 25 miles although more if you're prepared to do multiple hops) that has access to a choice of national networks. Looking at the list, this doesn't look like it would be too difficult to set up.

You then pick your equipment to match your budget and capacity requirement. El Cheapo wi-fi with directional antenna if you're on a shoestring, 5.8 if you have a few bob, and something like 10 GHz if you've loadsamoney.

To be honest, linking into the network is going to be expensive for such small amounts of bandwidth (2Mbps chunks) but it would prove the principle and provide a framework for expansion later.

I will write an article about it if you want!

a.

spongebob

Thanks Antoin.

Generally , when a mobile network gets a FULL NATIONAL licence, it is obliged to provide 80% population coverage by date xxxx . No mobile network has ever been obliged by licence condition to go above 80% that I know of.

This condition applied to Meteor and will also apply to Hutchison '3' when they launch (I suspect their target is end 2007 for the 80% figure)

The Clár areas are the chunks of the country with the highest cumulative population loss since the 1920's. The 'celtic tiger' era did nothing to stop that as it happens. Typically the Clár areas will be in the OTHER 20% of the country when it comes to an 80% population coverage target.

Now look at the Meteor coverage map (with its 80% coverage requirement completed and rollout now stopped). It is a bit optimistic looking because their coverage is not as complete as is indicated but nevertheless instructive. The Hutchison 3 coverage matrix will be pretty similar ....by 2007 that is.

http://www.meteor.ie/network/index.html

Now have a look at this (PDF map) of the Clár areas.

http://www.pobail.ie/ie/RAPIDagusCLR/CLR/TuilleadhfaoiCLR/CLR2002/d6420.ie.v1.0.t4.PDF

You will probably notice the near perfect NON overlap between them. They would also tend to be hillier than most of the country .

My basic thesis is this.

1. Clár areas are the most likely to be deficient in Mast infrastructure owing to the manner in which licence compliance is measured (POP % ) and population density .

2. Consequently they are denied adequate Mobile AND Fixed Wireless communications services by the market and by the regulator.

3. Therefore the Department ITSELF should be provisioning the mast infrastructure in those areas, that is the key. Market forces and coverage requirements will not achieve this without direct intervention by the department.

Until the Mast infrastructure is provisioned and opened to all Wireless operators ....be it Fixed Nomadic or Mobile.... then the infrastructure deficit will never be addressed in a systemic way.

The VSAT/Wi Fi patches being applied to this problem through the call for proposals are analagous to a botched water supply effort.

Most of the country gets mains water from lakes. Rural areas are (by analogy) being plumbed back to a well. Wells are simply not good enough to provide a reliable steady supply and will not be able to compete with those areas which do have same beacuse the supply is INADEQUATE

The same Clár areas have another disadvantage . The copper deficit. Before anybody suggests that this should be remedied rather than have the situation remedied through Wireless I would like to add one salient point.

http://www.metoffice.gov.uk/climate/uk/averages/atddata_djf.html

Clár areas are precisely those with the highest statistical risk of serious degradation of the already crummy copper plant.

Over to you!

M

antoinolachtnai

Well, an alternative way of doing it would be for the community to provide the masts and towers.

A mast is fundamentally not an expensive structure. You can erect a very basic mast for as little as EUR 500. You can erect quite an elaborate mast for EUR 40k. There is maintenance, but it wouldn't be much on a small installation.

The main costs and problems with towers in Ireland (as I see it, and it's not my area of special expertise) arise from difficulties in acquiring property and getting planning permission. The local community should be able to deal with these issues much better than a central agency could.

I don't see the relevance of the lightning stats. Lightning is a problem for both wireless and wired systems. It is a surmountable problem.

I don't see that there would be a big issue doing the long-haul wireless links, even if it is a bit hilly. A bit of a hill can even be an advantage in certain circumstances. You'd just have to draw some cross-sections (using OS maps) to figure that out.

It isn't quite true to say that these areas are unserviced by mobile phone. These areas are generally serviced by at least one mobile phone company (in my experience from visiting them and talking to people who have lived there). Naturally enough, there will be plenty of 'dark' patches.

We have to accept that there is no intention to cover more than a fraction of the land mass with 3G service. 3G service is only economic in areas of relatively high population density. Even if masts were available in these areas for free, it is extremely unlikely that 3G service would be deployed there.

spongebob

80% of the population MUST have 3G by end 2007 . That is a licence condition, this will more than likely mean a map like the Meteor one I provided.

The structure by which masts are provisioned (government or community or community in conjunction) is not the issue.

The issue is that there is no SYSTEMIC approach to this deficit. A couple of mickey mouse VSAT systems is not the answer to a SYSTEMIC problem. That is my point. The SYSTEMIC solution along with the the deficit it addresses is neatly discussed in this article here.

http://www.telecommagazine.com/default.asp?journalid=2&func=articles&page=0306i09&year=2003&month=6

"However, BFWA is more cost effective than DSL and cable for density areas of less than 100 lines per square mile…in low density areas of less than 5 lines per square mile, BFWA is the only viable choice to broadband services as costs of DSL and cable networks are extremely high due to the high costs of local loop plant."

The costs are depicted in the following link!



Wireless is Easier to isolate from lightning than fixed. When have you ever seen a surge protector on an Eircom line (I have seen one provided by Eircom after a number of hits).

The planning system will fast-track masts under the essential infrastructure category.

thegills

80% of the population MUST have 3G by end 2007 . That is a licence condition, this will more than likely mean a map like the Meteor one I provided.

Does this not just apply to 3, and can this be achieved via roaming or as an MVNO on either O2 or Vodafone?

You can erect a very basic mast for as little as EUR 500. You can erect quite an elaborate mast for EUR 40k.

For decent coverage you would need at least a 30m mast which is about 35k for mast and install and 35k for concrete base and fencing, not to mention annual land rental.

Well, an alternative way of doing it would be for the community to provide the masts and towers.

Or the Local Authority which has premises through out their Counties; their buildings can be used for antennae or small mast structures and the planning process can be fast-tracked (they would alos be glad of the rental revenue!)

antoinolachtnai

you are right, there is a systemic problem. it needs a broad policy to resolve it, not just a few installs here and there.

As I understand it (from an eircom linesman) eircom don't use surge protectors because they are uneconomic. I understand that the cost of the call-out to fix the surge protectors is generally greater than the cost of the damage caused to equipment.

Surely the best reason for not bothering with this VSAT stuff is that an equivalent or better service is already available to these areas, via ISDN?

Antoin.

spongebob

He He

They do use them but only at the exchange end.

I do believe that an always on system should be protected from electrical storms such as the one in March 2003 which frazzled half the rural ISDN Hi-Speed boxes in County Galway.....not for the first time either.

No surge protectors were installed then either so they'll all go again within a year or so. Muppets.

Would you recommend an always on solution plugged into a copper line in those circumstances? I wouldn't. It is a strong argument in favour of Wireless being the only feasible solution for always on on the western seaboard.

M

antoinolachtnai

Maybe they should install user-serviceable surge protectors then?

I don't think it is fair to tar the people who run eircom as 'muppets'. They seem to run the company on a sound commercial basis. If there is a business case for surge protectors, I'm sure they will be fitted.

I think the matter of which technology is used requires a little more analysis than a blanket rejection of copper technologies.

Wireless technologies are vulnerable to the weather too. The mast may be struck by lightning, causing damage. The antenna may be blown out of position. The mast may fall, potentially causing major damage and injury. It might well be that the risk of this is lower than the risk of damage to telephone lines (though I doubt it.)

Damage to a mast or property might well be less likely than a lightning strike to telephone cables, but it is also likely to cost more and take longer to repair than damage to electronic equipment.

In addition, wireless transmissions in the ISM and other 'open' bands are vulnerable to interference that you may not be able to do anything about. In some bands, there is also the issue of interference due to weather conditions to be considered.

I'm not saying that wired is better than wireless in all or any cases. I'm just saying that you have to look at the statistics and measure up the pros and cons of each for a particular situation.

In practice, business issues rather than technical or meteorological issues will generally drive the choice of wired or wireless for broadband schemes in rural areas, so the discussion is moot. (i.e., eircom probably won't be interested in cooperating, so the scheme will have to use wireless.)

Antoin.

spongebob

Originally posted by thegills
Does this not just apply to 3, and can this be achieved via roaming or as an MVNO on either O2 or Vodafone?
For decent coverage you would need at least a 30m mast which is about 35k for mast and install and 35k for concrete base and fencing, not to mention annual land rental.Or the Local Authority which has premises through out their Counties; their buildings can be used for antennae or small mast structures and the planning process can be fast-tracked (they would alos be glad of the rental revenue!)

ONLY 3 are obliged to allow others to run virtual bnetworks on their system (an MVNO)

ONLY 3 are oliged to provide 80% population covrage, for what that is worth given that Meteor have left out huge chunks of the country while doin just that,

The other 3G operators have a c.55% population coverage requirement by end 2007, Oh Pooh and Voodoofone that is.

One of the keys is indeed the Local Authorities and in particular their reservoirs which tend to be on high ground. While a quality mast does cost 70k-100k to get up and running it will then attract a revenue stream over a long period of ime, most modern masts will last 30-50 years. By contrast, a quality road rebuild for a regional road comprising a regrade and resurface with a 20 year pavement life would cost 100k per kilometre at least.

It is instructive that many counties (the flatter ones) can be fully covered by 4 or 5 masts within one year. That is the cost of 4 or 5 kilometres of quality road work.........nothing on the big scale of things. Unlie roads they will be a revenue source over 30 years and will cut down a lot of unneccesary road journeys over that time.


M

spongebob

Originally posted by antoinolachtnai
Maybe they should install user-serviceable surge protectors then?

Liek a fuse in a fusebox, outside, absolutely.

I think the matter of which technology is used requires a little more analysis than a blanket rejection of copper technologies.

Wireless technologies are vulnerable to the weather too. The mast may be struck by lightning, causing damage. The antenna may be blown out of position. The mast may fall, potentially causing major damage and injury. It might well be that the risk of this is lower than the risk of damage to telephone lines (though I doubt it.)

Yes, all of this can happen but rarely does. The damage is localised to the mast (the equipment on the mast , some 50K worth) . One small house fire would cost 50k to repair too.

Masts are properly earther unlike telephone poles (of which there are many ...each a potential target)

Serious wind damage is rare. Misalignment is a matter of a ladder and spanner job unless there is windblown debris damage to a sectoral antennae.

Damage to a mast or property might well be less likely than a lightning strike to telephone cables, but it is also likely to cost more and take longer to repair than damage to electronic equipment.

I disagree, the damage is localised.

In addition, wireless transmissions in the ISM and other 'open' bands are vulnerable to interference that you may not be able to do anything about. In some bands, there is also the issue of interference due to weather conditions to be considered.

I am nota fan of ISM for carrier grade solutions, any more than VSAT. There is adequate licenced NLOS spectrum for all clar areas. Weather (fog in particular) is a bigger problem in the LOS bands it seems.

I'm not saying that wired is better than wireless in all or any cases. I'm just saying that you have to look at the statistics and measure up the pros and cons of each for a particular situation.

Take out the Clar map again. Tell me how much of the terrain is Granite and how much is Limestone . Then lets discuss burying fibre instead . I love fibre .

(i.e., eircom probably won't be interested in cooperating, so the scheme will have to use wireless.)

They won't provide wireless themselves either, they have 2.3Ghz 2.4Ghz and 3.5Ghz spectrum that they will not use properly.

M

bminish

Damage to a mast or property might well be less likely than a lightning strike to telephone cables, but it is also likely to cost more and take longer to repair than damage to electronic equipment.

Lighting damage to properly installed towers (or equipment installed on site) is extremely rare IF the tower is correctly installed and grounded.
A correctly installed tower can take direct lighting hits and many even take regular direct hits with no structural or electronic damage to equipment.

PolyPhaser are one company that specialize in lightning protection for towers and telecommunications equipment installed at mast sites.
http://www.polyphaser.com

They have quite a lot of useful information on lightning protection for budding tower installers in their engineering notes page, here
http://www.polyphaser.com/ppc_pen_home.asp

Everything installed on site has to be correctly done or there is the danger of compromising the protection systems leading to expensive equipment damage. This is especially important on shared sites!


By the way, tower height is not an automatic requirement, in many FWA type applications if you are already on a good site extra height (once you get above local obstacles) may make very little difference whilst adding considerably to cost and complexity of the installation.

Brendan