Skip Navigation

BDUK opens consultation on new and improved procurement process
Tuesday 26 January 2016 17:37:30 by Andrew Ferguson

Local authorities and broadband operators keen to get on with more broadband roll-out using public subsidy have a chance to influence the way the procurement will operate as the BDUK is seeking feedback on a new and improved approach to how the tenders will be handled.

The new approach is the result of discussion with the EU to get new State Aid Approval in place (previous approval ran out in June 2015) and for those waiting on the phase 2 roll-out in Devon and Somerset this is seen as increasingly urgent if two county councils go with central approval, or they can go it alone and operate outside the BDUK frameworks. Responses by email or post need to be in by 24th February 2016, and remember this is just about how the framework will operate, not what providers will offer.

A wider set of funding models are under consideration, along with the option to allow non open-access model to win a tender if there are no qualifying open-access bids.

  • Investment Gap Funding: Essentially how the existing contracts with BT for phase 1 and phase 2 projects operate.
  • Public private partnership: Joint Ventures could be used to deliver the service.
  • Concession to Build-Operate-Transfer: A concession is given to operate a network for a period of time, with it returning to public ownership at the end of the contract term.
  • Public sector owned supplier: A company formed by the local authority

A key part of the existing contracts remains to 'maximise coverage for funding available', which means providing point to point FTTH for 3,000 premises in a bid, may lose out to one that is fixed wireless or VDSL2 based that can supply superfast speeds to 8,000 premises with a side effect of may another 1,000 to 1,500 getting improved speeds compared to their current situation.


Posted by chilting 11 months ago
This is excellent news!
Lets hope everyone steps up to the mark and fires us towards 100% superfast coverage.
Posted by ahockings 11 months ago
I've given up. It's all moving too slowly.
My business will be impossible to run very soon on just under 2 meg (Computer engineer).
My EE mast has just gone live and now I get 55Mbps down and 8 up.
EE Christmas SIM helps (200GB over 2 months for 10 quid!)
I'll have to pay 50 odd quid for 50GB after that which is a bit pants but it's better than waiting for BDUK. I wonder if we put them in a wine cellar, they would be able to arrange a drinks party, crucially, before 2030?
Posted by WWWombat 11 months ago
At first glance, this is quite a change.

The new approach acknowledges that LAs might not receive bids for open-access networks, so allows for closed-access bids to still qualify!

That's a hefty shift by the EU.
Posted by gerarda 10 months ago
"The topics covered in this document will be described in BDUK’s State aid notification and is likely to form part of the conditions set out in any corresponding decision from the Commission."

Does this mean that the agreement with the EU which was expected to be in February has been delayed?
Posted by WWWombat 10 months ago
Responses by 24th February, so it looks like it.
Posted by themanstan 10 months ago
Makes perfect sense for sense for there to be closed networks as the diminishing returns of higher investment costs (harder to reach and smaller service base) are offset by more complete ROI. There would simply need to be some form of price regulation.
Posted by TheEulerID 10 months ago
The dilution of revenues by enforced wholesaling has obviously been a major disincentive to investment in marginal areas. Of course, in the fully commercial arena that affects only Openreach & BTW, but the impact on state subsidised projects affects all operators.
There is also the issue of practicality. Integrating hundreds of small networks into a retail ISPs operation is clearly prohibitively expensive.
Posted by mdar5 10 months ago
This was all signaled last year as part of the new EU telecoms commissioner new thoughts
Competition had reduced prices - but also reduced investment.
So the new normal was to be that more mega mergers would be permitted and also that some exclusivity would be allowed to give a company a fighting chance of getting a decent pay back if it made a considerable investment.
Posted by TheEulerID 10 months ago

The unfortunate side-effect of official policy to drive competition as deep into the network as possible has also introduced crippling complexities into systems, interfaces, procedures & technical standards. LLU now makes it very difficult to unravel. The industry is now increasingly depending on VULA products for higher speeds. BT, for obvious reasons, wanted VULA over LLU in the first place.
Posted by TheEulerID 10 months ago
It's difficult to believe it, but at one point Ofcom actually expected the SLU to be a viable competition model for higher speed networks which, a bit of analysis would indicate, was never commercially viable due to the costs.

The LLU operators would, of course, like an FTTH service to replace MPF but want it at comparable cost. That is dreamland stuff.
Posted by gerarda 10 months ago
Have had a close look at the document I cannot see it doing anything to address the problems of the previous framework. Its scoring is hugely biased in favour of existing BDUK suppliers, the number of premises passed is given priority over 100% coverage, and the type of technology to be used has already been limited to fixed wireless, FTTC or FTTP.
Posted by gerarda 10 months ago
and also the maximum amount of public funding required has been specified before it goes out to tender. So a bid for say £2.2m to do 100% of an intervetion would be disqualified in favoour of a bid at the maximum £2m which only covered 80%. The winnig bid not only has a higher cost per premises but does not take into account the future cost of reaching the remaining 20%
Posted by TheEulerID 10 months ago

Do you actually think your example is credible? It's difficult to imagine a scenario where you can go from 80% to 100% by spending just 10% more. Incremental costs for this sort of infrastructure tend to strongly increase, not decrease as you go through the tail. That little cluster of cottages, or the farmhouse costs muc more.
More likely is a progression like £2m for 80%, £4m for 90%, £8m for 95% and so on.
Posted by TheEulerID 10 months ago
Note that is illustrative, and it may not be so severe if compromises (like reduced speed) in the tail is allowed which would apply to wireless and fibre/copper hybrid networks. FTTH doesn't involve speed compromises, but it will be most aggressive increases in marginal costs in the tail end of the distribution.
Posted by gerarda 10 months ago
@EulerId I was making a point, which should have been part of the previous post that if you continue down the same route as before you do end up with the sort of incremental cost increases to cover the tail. The framework should be encouraging, not specifically excluding, innovative solutions
Posted by TheEulerID 10 months ago

Talk of innovative solutions is all well and good, but how about some credible examples? We know that some money has been invested in trials outside the mainline BDUK project, but to date I've not seen much news on progress.
Also, if it is possible to avoid the excessive tail cost issue, then surely it would have been seen in the world somewhere. The Australian NBN project has had to satellite for the (rather large) tail.
Posted by godsell4 10 months ago
The likes of Airbroadband and look interesting in the innovative solution category, I am hoping they fall into the FWA category. Is there a need for the BDUK process to also include more direct discussion between the communities to be served and the local council to find out what their needs and wants maybe too. Is that in scope of suggestion which could be included in this consultation? I expect some councils may do this already better than many.
Posted by TheEulerID 10 months ago
Fixed wireless can work, but it also becomes increasingly expensive as coverage goes up. It needs more masts to cover blackspots due to geographical topology and to counteract the loss of speed with distance. Also, the airwaves can get congested so again you need more, and less powerful masts. I think it could well have a role in the right sort of landscape and the right population density, but I think it would need to be tested on a substantive scale against the sort of requirements in a BDUK contract.
Ultimately fixed wireless does have capacity limits unless you go to very small cells.
Posted by godsell4 10 months ago
Yes agreed, it depends on the geography as to what solution works, if we speculate that many locations in the fine 5% are small groups of properties which are 2km from the FTTC cabinet and each group has one or two properties 1km from it. Use an FWA connection to get to a small DSLAM device near the group of houses, from that point use point-to-point radio to connect the one or two remote properties.
Posted by godsell4 10 months ago
To be clear, make connections to the group of properties.
Posted by chilting 10 months ago
Unfortunately many of our long lines in West Sussex are not in easy to manage groups at all. They tend to be strung out along lanes up to around 5km long in semi-rural locations.
This is basically the urban sprawl of the last 50 odd years.
Posted by gerarda 10 months ago
@fastman I think the onus should be on BDUK to allow the innovative solutions but they are clearly not interested. All their market trials were pretty old hat technology.
Posted by gerarda 10 months ago
@godsell4 and chilting provided you have line of sight because FWA can cover fairly long distances (certainly many times more than FTTC) in a few hops. What is really annoying is that our area has one of the worst FTTC coverages in the UK yet almost all the properties are in sight of the Mendlesham mast
Posted by chilting 10 months ago
If you can find an operator ask the question - Can you provide us with a service using Mendlesham? - if you don't ask you don't get.
Posted by gerarda 10 months ago
I know what the costs for a Mendlesham based service are likely to be and no commercial service could entertain that now BT have cherry picked the larger villages, and the Council is refusing to specify which areas won't be upgraded in phase 2.

Have a look at put IP14 5ND in the location and zoom out four steps to see what I mean.
Posted by JNeuhoff 10 months ago
@gerarda: "Council is refusing to specify which areas won't be upgraded in phase 2."

Usually the councils hide behind dubious confidentiality clauses. We know, because we tried to get detailed coverage (or non-coverage) information via various FoI requests. This is one of the reasons why the BDUK is a farce - at the expense of taxpayers.

Posted by JNeuhoff 10 months ago
@TheEulerID (Steve Jones): "There is also the issue of practicality. Integrating hundreds of small networks into a retail ISPs operation is clearly prohibitively expensive. "

We repeatedly hear statements from you along the lines of why certain things can't be done. For a change, would do you propose should be done?

Personally, I think option 2 (Public private partnership: Joint Ventures could be used to deliver the service.) makes a lot more sense than the BDUK gap-funding model.
Posted by chilting 10 months ago
We had to wait a fair time but West Sussex now have a very detailed map of the areas to benefit from BDUK2.
Posted by chilting 10 months ago
I wouldn't give up on a Fixed Wireless option. There must be more than enough potential customers in a 25km radius of the Mendlesham mast. When we canvassed for support we found that residents who got 2Mbps or less were very enthusiastic to spread the world around and sign up.
Posted by WWWombat 10 months ago
A 300m mast used for FWA? It'll give you range, that is for sure.

But capacity? Just how many properties would you expect to cover with it? And how many sectors and antenna would you give it?
Posted by godsell4 10 months ago
@chitling in that situation, to serve a line of houses over a range of 5km, if there is no easy access to a fibre connection at either end of that line of houses. Use FWA to get a node in the middle of the properties and then or fibre along the poles if you have them. the option might require some wire rearranging at such a cost which would make you go straight to fibre and leave the copper where it is. It does however depend on the geography a great deal. At least you have been told what is happening in Phase2 in your location.
Posted by chilting 10 months ago
Some of these locations may benefit from FTTP - that is probably the only workable fixed wired option. The copper network cannot be upgraded!
Looking at the BDUK2 map makes you realise just how much it would cost to get to 100%.
The Fixed Wireless option is the only viable alternative for the final 5%.
Posted by gerarda 10 months ago
@wwwombat Sizing the capacity is beyond my technical ability especially as the technology has moved on since I was involved but it would probably need to cover 5,000 premises to fill in the gaps.
Posted by WWWombat 10 months ago
Ubiquiti forums have lots of people trying to run WISPs with equipment in the 5.8GHz band.

In there, you see mention of
- Access Points with specs of 100 clients, but wisps aiming at 30-50 clients
- Towers with 8-14 antenna/sectors, sectors of 120 degrees, each with an AP
- Range of 7km for OK performance (probably not superfast)

In the UK, you might use 4 sectors on a mast, each using 1 5GHz "Band C" frequency. That's 120-200 clients in range of the one tower.

Or 6 sectors on a mast, using "Band B" frequencies; lower power and range. 180-300 clients.
Posted by WWWombat 10 months ago
If you want to target thousands, you start needing multiple masts, aiming to keep a restricted number of clients per access point.

For comparison:
The trial in West Witton (1 village!) was planned with with 4 masts and 10 antenna (which'll be 10 access points) to cover 180 properties: 20 properties per AP.

The Airwave Feasibility report reckoned on each 100Mbps link (which an AP will provide on a 20MHz channel) supporting 20 clients with a "superfast" service.
Posted by WWWombat 10 months ago
I guess that means a hamlet of 50 properties, and an intention to serve them all, needs one mast with 3 120 degree sector antenna, and 3 access points. Plus one pt-pt backhaul
Posted by gerarda 10 months ago
where Suffolk were suggesting FWA to between 30-50,000 premises. If Airwaves trial is correct it appears Superfast brings a multiplicity of masts and APs which doesnt appear to be reflected in the pricing or coverages of those WISPs offering superfast. for example. Maybe the topology of the trial caused this discrepancy?
Posted by WWWombat 10 months ago sounds like it uses equipment like the Ubiquiti Nanostation M5 at client, and Rocket M access point, with air interface of 150Mbps (100Mbps"real" throughput) on 20MHz channels. That's standard 2x2 MIMO performance, hard to get much better.

To give a quality superfast experience shared over this air interface, how many clients can you have in parallel? Airwave reckon 20.

The limit is the capacity on limited spectrum, not topology.

Same question, but for range? You might want to have a maximum range of 1-2km for superfast, but 4km will work OK.
Posted by WWWombat 10 months ago
If you want countywide coverage of 40,000 and get 30,000 clients, you need 1,500 access points. Maybe from 250-500 sites.

That's a lot of investment, and needs time to recover. Suffolk's "vision" suffers a flaw: it reckoned it was "interim" - which immediately kills the business case for 250-500 masts that might not be needed for long.

The Airwave feasibility document reckons on 8 years payback. That's not interim to me.
Posted by gerarda 10 months ago
@wwwombat Certainly it looks like spectrum availability does add to the amount of kit needed for wireless compared to lower speeds, however the kit can be put on almost any structure with line of sight and the model we used of free service to host an AP reduces the price to a fraction of a mobile mast or FTTC cabinet.
Posted by andrew (Favicon staff member) 10 months ago
Small WISP and community efforts often fair a lot better at being able to use almost any high object, but as soon as property/tower/land owners find out their is Government/Council/Big Commerce money involved the cost of wayleaves often jumps to the leave where its more than the kit costs.
Posted by WWWombat 10 months ago
A community mesh design can certainly reduce costs, but likely at the expense of a less efficient network design. The transmitters are likely to be placed in sites of opportunity, you won't get the best coverage. I don't imagine that many households would want a sectorised site with multiple antenna either, so it won't give you best capacity either. Community buildings are a better choice there, but occur less frequently.

Pluses and minuses.
Posted by tallmattuk 10 months ago

Have a look at the wiSpire concept in Norfolk, using churches as the PoP in the community and then backhauling, via other churches to the Hub.

as for how a Wimax network might operate, one idea would be to use the Band C frequencies (which have a limited licence cost) for the backhaul and Mesh links, and the Band B frequencies for connections out to the individual houses and businesses. If you get Acccess Points that are both 2.4 and 5 Ghz enabled, the user density can be increased substantially.
Posted by WWWombat 10 months ago
Nice plan!

Hasn't Wimax died a death?

I thought more WISPs have been oriented at 802.11n-based hardware for the last 5-ish years - perhaps influenced by licencing costs more than spectral efficiency. If you started now, you'd probably be looking at 802.11ac hardware, even if we cannot make use of all the possible MIMO modes.

If you want to go the licenced route, some of the LTE-advanced options with multiple MIMO can work better.

I'm not convinced you'll see much from adding 2.4GHz; the noise floor from existing routers' wifi might make it unusable.
You must be logged in to post comments. Click here to login.