Search This Blog

Monday 31 May 2010

What is NGA?

Dispelling a few myths, this is an excerpt from an email thread we have permission to publish on this blog. This blog post can be read at http://5tth.blogspot.com

Barry Forde: Firstly what is NGA - "Well it has to be high speed symmetrical low latency capacity able to handle next generation applications such as HD, 3DHD and interactive video for gaming, tele-medicine and social interaction. Also probably more than one stream as most homes have more than one person in them and each is likely to want to do its own thing. Put that together and you need full duplex/symmetrical >50Mbs low contention capacity. I cannot see how FTTC, DOCIS, terrestrial wireless, satellite or data over power can deliver that, the only technology that has the range and bandwidth is fibre all the way. If we accept that fibre is the way then due to the way ethernet technology goes up in 10s you have either 10Mbs, 100Mbs or 1Gbs and of those only the higher two meet the NGA bandwidth requirements. So I'd say for NGA we need FTTP running either 100Mbs or 1Gbs back to the serving PoP.
What keeps happening is that we merge together the issue of NGA and the 2Mbs USC and get ourselves in a twist about which we are discussing. If we want to limit the public sector investment to only deliver 2Mbs then that's fine and dandy but lets not refer to it as NGA, it isn't. I think there is a genuine desire to make sure the 2Mbs USC is delivered in a timely manner and to everyone. However what the NWDA, BDUK and everyone else also realises is that any investment in USC that cannot then be reused for NGA is wasted money. We get one bite at this cherry and we had better make it work or the window of opportunity will shut for quite a while. Government will believe it has solved the problem for the rural areas and will not want to listen to people telling it that hundreds of millions have been spent but by the way we need to do it all again as its no longer fit for purpose.
So coming back to your delivery via FTTC, Wireless, Satellite etc these may have a part to play in the USC delivery. However I believe that public funds should only be used for the parts of project that do put real fibre in the ground and can therefore be reused when moving to NGA. For instance getting fibre to PoPs in the rural areas but letting the local community fund the final delivery via wireless is fine. Putting public money into the wireless side is not. Similarly getting fibre to the cabinet is fine, funding VDSL2 for final/first mile is not etc. etc.
How do we deliver USC 2Mbs under this sort of regime? Well I think if you can get high capacity pipes into local communities which can be used to provide low cost middle mile connectivity/backhaul then lots of options open up for the first/last mile connections. If communities wish to bid for funding for using fibre in an FTTP style roll out then that's eligible for support but because the total cost is high they might well have to contribute something for the local bit.( Note their contribution might be money but equally it can be things such as donating wayleaves and free/cheap labour doing duct digs or fibre installation). If on the other hand they want to take an interim position and roll out USC only then a low cost wireless/WiMax or possibly Data over power rollout is a runner. Its down to them to make the decision.
If the local community doesn't want to get involved at all then what? Well if we have funded fibre to the community then Openreach or anyone else can make use of it to do an FTTC via sub-loop unbundling or anything else they want to do. I think there are likely to be small start up ISPs who could make a living filling that space.
So to summarise I think public sector funding should only be made available for the areas that are unlikely to get NGA in a reasonable time frame otherwise (2015?). It should only be used to support fibre laying projects and finally it should mandate a delivery speed of no less than 100Mbs symmetrical. Where public sector money is targeted at USC delivery only then it should again mandate that the funding can only be used for the portions of the build that result in pure fibre installation."

reply by Miles Mandelson:
"That’s the clearest exposition of NGA and the role of public funding I’ve seen yet and, setting aside matters of self-interest, is nigh-on impossible to take issue with. However in small communities such as my own, the cost differential between FTTP and high-performance wireless (for example) is so great that, while the latter is affordable and therefore very tempting, the former feels almost completely out of reach. Accordingly, in trying to make progress locally, I find myself making compromises as regards performance in order to bring advancement into the realms of affordability and, in doing so, bending the definition of NGA to make a case for funding.

So to adhere to the principles you have set out, we need to find more ingenious (and affordable) ways of delivering FTTP in sparsely populated rural communities over long distances and across challenging terrain".

Barry's reply:
"I dont think anything I said would prove problematical to somewhere like Gxx. My thoughts were that when looking at any area with a view to rolling out FTTP there would be an analysis of the geography, distribution of premises and number of premises. From that would come an indication of what the cost would be for full FTTP rollout via a standard commercial contract. From that would flow the idea of the degree of subsidy needed for that particular community. Somewhere very difficult/expensive to connect would be eligible for higher subsidy than one that's easier/cheaper to do. I envisage the subsidy reflecting the difficulty of rollout. So the contribution from people within the community would remain broadly constant across community types. I would expect any attempt to roll out the USC 2Mbs to be subject to the same analysis and the equations to line up with each other.
In the case of Gxxx if we could deliver a centrally funded fibre into the village with a 1/10GbE backhaul, charged at a nominal fee, linking back to an Internet exchange where you could get metro Internet Transit pricing you would have quite a bit of your isolation removed. From the PoP in the Village you might well go for a quick and cheap WiMax solution and that's fine. As users wanted more then the move to putting in a fibre at a time to their premises or getting them to club together to service clusters is quite on the cards. So long as there was no subsidy for the wimax bit it would be perfectly OK for them to request subsidy for that last mile bit as and when they were ready to do it."

(this was an excerpt from an email thread we have permission to publish on this blog.
The discussion can continue in the comments if anyone has anything they wish to add ...)

2 comments:

Unknown said...

This makes a lot of sense to me. Anything other than fibre all the way is an expensive stopgap.

It makes me think back around 10 years to when I was doing some specification work for a semiconductor device that was going into a home gateway for the far eastern market.

At that time - 10 years ago remember - 60Mbps was considered to be the target requirement for a dwelling, taking into consideration all the expected multiple video streaming and gaming needs of the future. With 60Mbps as a baseline we assumed that 100Mbps would therefore be the standard.

Now technology has moved on and video encoding has got better so good quality video doesn't need the bandwidth it might have 10 years ago. However our usage of all things internet has probably exceeded what was forecast at that time and for anyone to think that less than 100Mbps is "Next Generation" has their head buried in concrete.

Oliver said...

I guess your definition of NGA depends where you are now. If you're in a rural environment with a dial up modem, then 2mbps broadband is NGA.

It's a good post though - we need to do what is best for everyone in the long term, and if that means some people in the sticks are left behind in the short term, then so be it.