FEATURE

Making the connections

There have been several attempts to find the right model to drive deployment of superfast broadband, but what is right for one country or city may not be right for all, as Robert Roe discovers

The benefits of superfast broadband are well established, but the best way of installing the fibre is still being debated. Established models range from fully government-owned fibre, through to fully privatised models that rely on service providers to get services up and running. Choosing the right model, and rolling out fibre deployments effectively using the chosen model, is a task that varies with geography, subscriber density, demand, and the willingness of telecommunications companies and government to invest in the backbone of the optical communications network.

Stokab is a city-owned company based in Stockholm that provides ‘dark fibre’ – passive infrastructure in the form of optical fibres that do not have active termination equipment. Anders Broberg, director of communications for Stokab said: ‘There was a really good copper network [in Sweden] and it still is a really good copper network, but in 1994 there was extremely little fibre. In 1994, the household market was using copper, not fibre, so it was the businesses that were interested in fibre in the beginning.’

Research has shown that there is a positive correlation between fibre deployment and economic growth. In particular, high-speed networks strengthen and drive diversification of economies, as small and medium businesses (SMBs) are often the quickest to adopt and benefit from improved online business. High-speed telecommunications networks drive innovation, but also improve GDP and provide jobs.

Some of these benefits can be quantified; a report by the consultancy Arthur D Little states that ‘for every 10 percentage points increase in broadband penetration, the isolated economic effect on GDP growth is  about one per cent’. On job creation, the report estimates that:  ‘For every 1,000 additional broadband users, approximately 80 new jobs were created.’
The report published by Arthur D Little goes on to explore the potential models used for fibre distribution, the benefits and disadvantages of each model, and provides some insight into why there is no universally accepted model for fibre distribution.

It notes: ‘The industry has spent almost 10 years debating which network architecture best fits the complicated puzzle of return on investments, technological soundness, and competitive position.’ The challenge is rolling out fibre effectively once a selected model has been chosen, so that national economic benefits of the latest fibre infrastructure can be realised without the considerable investment adversely affecting the delivery of the service.

This is a challenge that Stokab has faced since the early 2000s, when it began rolling out FTTH solutions to household customers. ‘We have not used any tax payers’ money to build this [network] and it is possible to do this,’ said Broberg. He went on to explain that a key strategy for them was ‘not building everything at once; you just start with the businesses, and then follow the money so you can generate cash flow.’

This kind of strategy may seem more in line with a private investment model, but during the initial rollouts for Stokab, in the early 2000s, there was much less demand for fibre-based broadband. ‘Our experience is that in the first case people say “I don’t need fibre” but once you have broadband you just want to have more and more capacity, and very quickly a demand for high-speed broadband will develop,’ said Broberg.

The Arthur D Little report goes on to outline five major models to encourage the deployment of fibre-based broadband. The five models listed in the report range from: Private investment- unregulated – where the network is completely installed by private companies; to, at the other end of the spectrum, government-controlled fibre – where a fibre network is installed by government and leased to service providers.

The report observes that these models have been implemented both successfully and unsuccessfully in many cases. ‘In a global survey of National Fibre Strategies, there were leading and lagging examples for each macro model. Hence choosing a National Fibre Strategy is not just a simple case of good model versus a bad model. Rather, it is a case of identifying the best model for specific national market conditions, and applying that model well.’

‘While it is tempting to look at the top and bottom of the list to find winners and losers in the fibre race, this would be an oversimplification that ignores many markets’ specific characteristics,’ claims the report. It goes on to explain that the UK for example, ‘due to its relatively compact geography and proud, 100-year history of significant network investments, has a good copper- and coax-based access network.’ This copper network holds back the distribution of FTTH/B solutions because consumers already have access to, on average, 50 Mbps broadband and therefore are generally unwilling to upgrade to more expensive, higher-capacity services. ‘This situation has impacted the timing, and reduced the pressure to move fully to fibre solutions,’ the report concludes.

While this may appear to be at variance with what Broberg said in regards to copper networks in Sweden, it is the timing that was crucial in the case of Stokab. In the early 2000s, when Stokab began rolling out fibre to household customers, copper technologies were far behind what can be achieved today using technology such as G.fast.

No single model is right for any given situation; for example, the solely private investment model relies on private investment coupled with heavy regulation to encourage competition from service providers. ‘This is often constrained by heavy investment requirements and thus long pay-back periods. This is coupled with the uncertainty of service take-up rates due to competing infrastructures,’ the report explains. This is especially true in less-densely populated areas or where getting permission to complete groundworks is difficult.

An example of this kind of model is Western Europe, which lags behind other regions overall in terms of broadband penetration. Generally speaking, countries that have achieved higher penetration have done so by adapting the model to better suit their own needs. ‘Where individual countries within Europe have achieved higher penetration, they have done so by adopting attributes of quite different models in their local market,’ the report by Arthur D Little states.

Another example given in the report is the government-controlled fibre, which was the model implemented in Stockholm. ‘The well-intended government-controlled fibre, while probably resulting in the most uniform and widespread infrastructure, will do so at a slower and perhaps less financially efficient way.’

Choosing the right distribution model for fibre rollouts is important due to the investments required. The Fibre To The Home (FTTH) Council estimates that re-wiring Europe on a similar reach and scale as the historic networks will cost approximately 200 billion euros.  Similarly, the Arthur D Little report highlights that Goldman Sachs estimated at least US$ 140 billion would be required even partially to cover the USA.

Equally important is that the strategies for deployment need to be implemented with a clear goal in mind for them to be effective, this can be seen in established networks where multiple strategies have been implemented to varying degrees of success. The report observes: ‘A typical developed country may have an incumbent fixed network; a challenger fixed network, four mobile networks, and cable TV networks.’ This scenario can amount to five or six separate national telecom infrastructures, something unheard of for other infrastructure, for example water supply. The report states:

‘This mosaic of infrastructures typically results in densely populated areas being served with multiple solutions of variable standards, and the rest of the nation being under-served.’

Stockholm was an early adopter of a government-controlled fibre distribution model, removing some of the problems with multiple layers of telecommunications infrastructure.
In a study conducted by the Acreo Swedish ICT research institute and published in 2012, it was reported that Stokab invested 5.4 billion SEK (more than 600 million euros), whereas the economic gain to the region was found to be around 16 billion SEK (around 1.9 billion euros). The study, co-authored by Forzati and Crister Mattson, looked not only at the revenues gained by Stokab from renting out its fibre, but also cost savings for the city and county council, and also the economic benefits derived from having the available infrastructure.

According to Forzati, senior scientist at Acreo Swedish ICT and one of the authors of the study, the key decision was to create Stokab as provider of ‘dark fibre’ so that anyone could operate or deliver services across the infrastructure: ‘The fact is that this infrastructure is there for anyone to use – it’s an infrastructure for society – whereas in many cases elsewhere, the network is owned by one of the players that is delivering the services.’

‘In 2012, Stokab’s customers included more than 100 telecom operators and more than 700 other companies and organisations,’ according to the Acreo Swedish ICT report. This meant that operators could compete against each other by leasing fibre direct from Stokab rather than, as is the case elsewhere, having to lease from a company that not only owns the infrastructure but is itself delivering its own services.

The Arthur D Little report explains: ‘It is this free-market approach that drives competition within Stockholm creating better services for the subscribers.’ Moreover this has meant that ‘the cost of leasing fibre in Stockholm is less than half what can be expected from other capitals around the world,’ according to the report.

This translates into lower costs not only for operators and enterprises that have a need for fast and reliable communications. ‘Lower prices propagate down the value chain and stimulate new services and entrepreneurship’ the report states.

‘Our experience is that the customer, especially operators, really like that we are not an operator, not a competitor. So they could choose fibre from the incumbent or fibre from Stokab, in that case they mostly prefer to choose Stokab because we are owned by the public and we only have dark fibre so we are treating everyone equally,’ said Broberg.

According to Forzati: ‘Beforehand, they were purchasing services from the incumbent, but when the network became available they were completely free to choose to buy services from whichever service providers they wished – and that led to a very significant reduction in price.’

Because Stokab’s investment policy has been sustained over two decades, the fibre network now connects almost all multi-dwelling units and commercial properties within Stockholm: about 90 per cent of households and almost 100 per cent of enterprises can sign up to a fibre-based connection.

Forzati drew a qualitative comparison with Copenhagen where different decisions had been made about the broadband infrastructure – the fibre in Copenhagen is owned by the incumbent telecoms operator. Although the two cities are relatively similar in terms of size, population and economic structure, ‘there is much less fibre to the premises: barely 20 per cent of multi-dwelling units in Copenhagen are connected, compared with more than 90 per cent in Stockholm.’

The availability of dark fibre in Copenhagen was also limited, he continued, and the price higher. Forzati remarked that it was not possible to attribute differences between the economic developments of the two cities directly to different strategies in their broadband development, because there were too many other confounding variables. However, he added: ‘You can suspect there is a link’.

Forzati noted that: ‘You can translate the higher rent as increased value of the property – a higher capital value returning higher rents.’ The study estimates that the increased capital value to the housing companies so far is around 1.8 billion SEK (approximately 20 million euros) and the positive effects have started to outweigh the investment costs since about 2012. There is a further advantage to the housing companies from installing fibre within their buildings – they can save on electricity, water, and heating costs due to the automation and ‘smart building’ solutions that fibre makes possible.

However successful the government-controlled fibre deployment model has been in Stockholm applying this model to other cities and regions does not mean that it will be successful. Broberg explained some of the difficulties of deploying this model in 2014, now that many areas have at least some kind of established fibre network. ‘You can copy the model, definitely, but if you have a city where you have a lot of operators that already have a lot of fibre it would be quite hard to establish a neutral operator,’ said Broberg.

He went on to explain that it could be used in some areas, mainly those that are less densely populated because ‘the bigger operators have not built as much fibre in the smaller cities’. The second reason is that ‘now the demand for fibre is much higher than it was in 1994 so it is easier to get business if you provide fibre.’

One clear benefit to this kind of model is that: ‘The goal for the city is not to get a lot of money for profit from this company. The goal is to provide a good ICT environment in the region,’ said Broberg.

This is in part due to the way in which government-owned companies work in Sweden, Broberg went on to explain how this has affected the creation of Stokab. ‘You must have some mission, such as provide water or provide fibre in our case. You must also be quite close geographically to your area. The third component is you are allowed to generate as much profit as you want as long as you take the profit back to the company and invest in the business.’

‘If you stop [re-investing] then you must have lower prices so this has a positive impact on the market in our case’ said Broberg.

As fibre deployments have already begun in the vast majority of countries, a fully government-controlled fibre deployment model would just increase complexity against established networks. However such a model can be valuable for rural areas that have very little coverage. Here a small region can achieve high coverage, using government investment, much faster than if left solely to operators. ‘We have a lot of visitors from across the world and the situations in all cities are different so it is not always easy to copy the whole Stokab model,’ said Broberg.

References
National Fibre Strategies—National economic imperative or just another private industry task? Arthur D Little, 2013.
Stokab, a socio-economic analysis, Acreo report acr055698, Stockholm, July 2013

Feature

CableLabs is spearheading efforts to develop a proposal that uses coherent optics to dramatically boost the capacity of hybrid fibre coaxial networks, reports Andy Extance

Feature

Systems vendors are using intelligent software to squeeze more performance from optical networks. Pauline Rigby reports on developments at OFC 2017