Wednesday, February 16, 2011

Conclusions from BDUK’s theoretical exercise

Broadband Delivery UK’s (BDUK’s) theoretical exercise report, drawing conclusions from the exercise conducted last summer following the industry day in July, provides an interesting overview of the current array of technologies, challenges and opportunities in the provision of broadband in rural areas. It was published late last year alongside Britain’s Superfast Broadband Future but has slipped under my radar until now. Coffee warning: this is a long post, so you might want to go and get one before reading any further.

The purpose of BDUK’s exercise was “to understand the choice of cost effective solutions available to provide a minimum level of coverage to all premises without access to at least 2Mbps broadband”, based on three locations (one in south Wales, another in the north west of England and one in the north of Scotland). Suppliers were invited “to explain what solutions they would deploy in those areas to provide coverage, how much it would cost and what revenues they would forecast over time, and a calculation of the subsidy needed to make the project viable.”

A key finding was that the USC commitment should not be separated from the drive to deliver NGA, as also acknowledged in Britain's Superfast Broadband Future:
“BDUK’s Universal Service Commitment (USC) objective should not be seen as separate from the superfast broadband objective, but rather an integral part of pushing next generation networks deep into rural Britain.”
The importance of backhaul was also recognised, which in my view further reinforces the case for investigating the potential to consolidate and re-use existing education and other public sector networks that often have significant reach into rural areas already:
“The provision of affordable backhaul in these rural areas as part of the government’s delivery approach would positively impact the business case for private sector investment in fibre, fixed wireless and mobile solutions in those areas, reducing the subsidy needed to provide universal coverage.”
More on re-use later in this post. In total, BDUK received 26 responses, but:
 “BDUK was disappointed that despite some original expressions of interest no mobile operators decided to participate in the process, meaning BDUK saw no practical illustrations from the mobile operators of how their current plans for mobile broadband can contribute to the coverage challenge.”
There were some question marks over how representative the areas BDUK chose for the exercise actually were:
“Several suppliers shared their perception that the areas BDUK chose for the exercise were atypical and particularly hard areas to solve – in the 3% of least populated areas in the UK, as calculated by one supplier – and may not be typical of many of the not-spots and slow-spot locations. These suppliers warned BDUK against directly extrapolating costs from just these example areas for the whole of the UK’s not-spots and slow-spots. BDUK recognises that these areas may not be representative of the nature and cost of the most typical areas to be addressed, but maintains that there are a significant number of similarly difficult and costly areas in the UK, which in aggregate may require a significant proportion of an overall subsidy, if not of overall physical area covered. Nevertheless, the point about extrapolation of costs is noted.”
Many responses took a similar approach to provisioning the exercise locations:
“The majority of submissions received by BDUK proposed fixed wireless technology to solve not-spots and slow-spots in the exercise areas. These suppliers saw fixed wireless having a role in all three exercise areas, although identified limitations of point-to-multipoint solutions below a minimum scale in a particular area. For example in the North of Scotland scenario, several postcodes that weren’t clustered close together couldn’t be served cost-effectively by wireless solutions in most respondents’ solutions since the number of customers in a sector would generate too little revenue to make the service sustainable.”
Responses also highlighted the key gotcha between using licensed and unlicensed spectrum:
“Suppliers presented solutions using either licensed or unlicensed spectrum, but only those using licensed spectrum could assure service availability.” 
And an interesting finding emerged in relation to the potential to re-use of electricity pylons to deliver services:
“Even where high-voltage (HV) electricity pylons are available as options for site locations, the cost of developing new sites is often prohibitively expensive – ironically due to the cost of getting LV power to them, so construction of new masts may be necessary.”
Alanis Morissette should take note here perhaps? To quote Ed Byrne: "The only ironic thing about that song is it's called 'Ironic' and it's written by a woman who doesn't know what irony is. That's quite ironic." On fixed access, some suppliers proposed a solution new to Openreach:
 “At least one supplier proposed cabinet ‘splitting’, i.e. placing an additional cabinet closer to customers, as a way of increasing the reach of VDSL. Suppliers were told by BT Openreach that there is currently no product available from BT Openreach to enable them to ‘cut into’ the network to create / co-locate cabinets further out into the network, so at the moment this is an option that requires BT Openreach’s cooperation to implement. However the existing regulatory environment does allow Communications Providers (CPs) to request the implementation of a new equivalent product from BT Openreach.”
So would this effectively be sub-sub-loop unbundling, or SSLU? Powerline could be a possible future option for fixed access, but this is still some way from becoming an effective solution:
“BDUK received responses proposing the use of the electricity distribution network infrastructure to propagate broadband signals into the premise. Respondents were able to demonstrate the deployment of high-speed access over the power infrastructure into premises in other countries, and had the cooperation of the local electricity distribution network operator (DNO) in developing the solution. However, this solution has been piloted before, but not yet been progressed in the UK, so respondents were unable to demonstrate the commercial sustainability or practical deployment of this technology in these rural areas. The robustness of power infrastructure would suggest that this is worthy of investigation again. The powerline access technology must still connect to an appropriate backhaul route. This is not necessarily straightforward if the power infrastructure does not intercept the telecoms infrastructure locally, unless backhaul is provided via the DNO’s passive infrastructure back to an aggregation point – but these products are not yet available from the DNOs.”
Satellite options were also considered, but latency remains an issue, particularly in relation to certain applications:
“With a ~70,000km round-trip communication path, the inevitable latency of satellite means that it is more suitable for broadcasting and multicasting and less suitable for real-time services (e.g. voice, virtual private networks, two-way video) needed for home-working, relative to ground-based technologies, despite optimising technologies that improve the browsing and downloading experience.”
The importance of affordable backhaul was stressed in relation to all proposed solutions:
“A common supplier conclusion was that the provision of affordable backhaul in these areas would positively impact the economics of supplying wireline, fixed wireless and mobile technologies and so increase the business case for private sector investment.”
Some interesting insights into backhaul too, in terms of how it’s provisioned:
“All networks are shared resources, engineered to accommodate busy-period loads. Operators size networks on the basis of a backhaul allocation per customer, which enables them to offer service levels for anticipated customer experience during the day, e.g. “2Mbps available 90% of the time during peak three hours of the day”. A typical backhaul allowance from suppliers when designing solutions was 30-60Kbps per customer, which is consistent with mass-market products available today.”
The proposed solutions varied quite considerably in this regard:
“However, as the choice of backhaul infrastructure in these solutions was influenced by environmental, customer density and economic constraints, BDUK saw a wide spread in allocations: e.g. from <20Kbps for a wireline connection where backhaul would be very expensive to provision, to >200Kbps for wireless backhaul where the low density of customers on a mast meant that a significant per-customer backhaul allowance was available. These examples would result in a significantly different customer experience in peak hours of the day, but this experience is also dependent on the allowance for data transport between the operator’s point of handover and the internet as well although this is generally a commercial decision for the CP.”
Which is perhaps an important additional consideration in relation to having the “best superfast broadband in Europe” within the lifetime of this parliament, and beyond last mile bandwidth and a competitive marketplace? Responses also provided further commentary on physical infrastructure access (PIA) concerns, as I described in this previous post:
“Several suppliers had indicated that they hoped that Ofcom’s wholesale local access (WLA) consultation might provide new opportunities for backhaul. However, the recent Ofcom decision following the WLA market review accepts that access to BT Openreach’s passive infrastructure (their PIA product) should currently only apply to the access network up to the NGA access node (typically the nearest major town). Connections beyond this are provided via regulated products that sit within the business connectivity market, which are subject to cost orientation and other regulation. Suppliers also noted that while the industry is able to develop fibre access products under Ofcom’s general access obligation on BT Openreach, it is not obliged to provide dark fibre backhaul products in the current regulatory environment. This is in contrast to some other European countries, e.g. in Germany, Deutsche Telekom AG is required to offer dark fibre in certain circumstances. The use of BT passive infrastructure for leased lines and backhaul will be re-examined by Ofcom in the first half of 2011 as part of the business connectivity market assessment.”
There were both positive and negative comments in relation to the potential for re-using existing public sector networks:
"Responses that included Welsh schools or Pathfinder public sector locations highlighted the importance of previous investments in networking public sector locations; however it also became clear that there are still significant issues involved to make them accessible to public use. Specifically, it can be a challenge to transform networks created for a single customer or purpose in order to set up services that can be offered and billed to individual members of the public. This requires a different operating and security model and also upgrades to the active components of the network, which represents significant additional cost. In some instances, existing contractual arrangements may also prohibit the network operator from serving additional customers in their current form, particularly for commercial gain. Alternatively, extension of these contracts may be incompatible with public procurement regulations and also state aid guidelines for broadband.”
It’s good that such existing networks were at least acknowledged, but there is clearly more to do to explore how these challenges (most of which are contractual/commercial, rather than technical) might be addressed, as I've mentioned in this previous post (and also here). Similarly, the challenges of securing providers to offer services over new networks were also recognised  – “developing multiple interfaces into the range of alternative access networks would run into the millions of pounds” –  this is where the JON Exchange (or its successor?) has an important role, primarily in reducing the duplication of effort needed to connect multiple networks, operators and service providers, but the changing nature of the Internet and the consequential changing role played by ISPs may also help here too, as discussed in this previous post.

The section on revenue and project returns provides further insight into the nature of responses received:
“The majority of responses were received by network operators rather than the ISPs packaging their services to provide to the end-customer. Therefore most respondents had modelled wholesale prices / revenues, and only indicated the likely prices / products that would be paid / received by the end-customer.”
Which rather suggests that ISPs regard rural broadband as a problem for network operators to solve? This also has implications on the level of competition likely in rural areas, in terms of service availability and diversity:
“…standardisation has enabled the mainstream retail ISP market to operate on thin margins. If a wholesale operator does not charge the same £7 per month or through a similar business-to-business platform, mainstream ISPs will not be able to make money on these connections unless they segment their product in order to charge more to the user in these areas. The lack of availability of recognised, mainstream ISP brands risks lowering take-up among consumers and therefore an increase in demand risk for an investor. The network operator may instead be able to attract smaller, niche ISPs with a business model and products designed to target these rural coverage areas.  This outcome could be expected to result in less competition and choice for consumers – although perhaps more services tailored closer to their needs – which is less likely to provide the same benefits of innovation and price competition that has been typical in the mainstream ISP market. BDUK needs to ensure that the availability of ISP services that are valued by customers is incorporated into any procurement process or investment in the network.”
This is borne out by community implementations to date, where only a single operator is available, but I’m sure few beneficiaries would disagree that one service provider is a whole lot better than no service providers, especially if that service provider is community-owned and operated. The section on revenues also identified an additional advantage of re-using public sector networks, in relation to the public sector's provision of anchor tenants (schools, GP surgeries, libraries - if there are any left that is) to guarantee a certain level of return:
“The prospect of public sector revenues may be more limited and substantially dependent on the timing and structure of existing contractual arrangements. However, where the public sector can be relied on as an ‘anchor’ customer for a project for a significant duration – e.g. potentially through public sector network procurements – the certainty of these additional revenues makes a substantial improvement to the economics of an area.”
So the question is whether this anchor tenant advantage and the reach of existing public sector networks present enough of an opportunity to offset the commercial/contractual (and to a lesser extent technical) hurdles identified previously. To my mind, the potential benefits make it worthwhile to put some serious effort into exploring how the difficulties might be overcome.

In conclusion, the report highlights the importance of local approaches to broadband strategy, both in terms of addressing particular local requirements and also to leverage local resources and assets as fully as possible. This is something that the four superfast broadband pilot areas are actively pursuing. Balanced against this is the clear importance of scale, to ensure commercially viable and sustainable approaches. As suspected, solutions will involve a range of technologies, depending upon circumstances, and there is much to do to explore the re-use opportunity to provide cost-effective backhaul:
“The re-use of public sector available networks is frequently referenced, but the cost and practicalities of converting the potential of these assets into a service that can be productised and sold commercially has not been adequately dealt with in any of the responses. Each owner of public sector available networks resources should make clear their intentions as to the extent to which they are able to and intend for their asset to be used in offering services direct to end users.”
A fair challenge I think, and one that the public sector should respond to?

No comments:

Post a Comment