All the players are clear that broadband for all is a national requirement – but, in the absence of clear leadership, the private sector is finding ways to advance the agenda. Kathy Gibson found out more
The National Development Plan (NDP) is clear that broadband is going to be critical for the economic future of South Africa.
The bold vision outlined in the NDP is for 100% of South Africans to have access to broadband services by 2020, at a cost of 2,5% or less of the average monthly income.
This is the same vision shared by the Future Wireless Technology Forum, which held its first discussion meeting bringing together stakeholders in the implementation of the SA Connect policy.
“But what does this vision actually mean?” asks Ellie Hagopian, chairman of WAPA. “How are technologies disrupting the established models? And how can the industry work together in collaboration with government and regulators to achieve the agenda?”
Jens Langenhorst, deputy chair of WAPA, points out that solutions to achieving SA Connect have to bear in mind that there is no single solution that will work for the whole country.
“Just looking at population density, it is clear that one solution cannot be scaled across the whole country,” he says. “And the average salary is not the same either; 2,5% of an average Gauteng salary is $20; in the Western Cape it is just $7.”
By 2020, 2,5% of the average South African’s salary will be $16, Langenhorst points out. “As industry and government, if we were to serve 90% of the population, this would result in a monthly revenue opportunity of $6-billion.”
The three main technologies that will enable the industry to reach these users are fibre, microwave and satellite, he says. Copper is almost a non-starter, with just 1-million ADSL lines available throughout the country.
Langenhorst believes that industry players must tailor their technology solutions to circumstance.
“In the north of Johannesburg, it’s easy enough to have fibre to the home. It makes sense to look at wireless in other areas, and there are areas where satellite is the obvious choice. And there are areas inbetween where the solution is not so clear.”
Putting some context around the discussion, BMI-TechKnowledge’s Tim Parle points out that South Africa has seen massive growth in mobile connections, and this has been the main driver for broadband connectivity.
“Almost 96% of the population can get some kind of 3G coverage; and we are already at 25% or 30% with LTE,” he says.
Individual mobile penetration is at about 60%, with 32,6-million of South Africa’s 54-million people having a cell phone. Households fare better, with 14,9-million of the 15,6-million households having access to a mobile phone, and 1,7-million with Internet access.
However, South Africa is still quite low on the scale of average Internet speeds, with an average download speed of 7,9Mbps.
More significantly, says Parle, the country has slipped to 75th on the WEF e-readiness index.
The good news, he says, is that we have a lot of fibre already installed, with a lot of infrastructure ready for use.
What the policy says
The SA Connect policy was adopted in November 2013 and envisages different phases, with specific targets. The four pillars are digital readiness, digital development, digital future and digital opportunity.
• Digital readiness – In this phase, policies that curtail competitiveness need to be removed, while markets and institutions are restructured to create an environment conducive to broadband investment and service-based competition. This requires that the regulator be autonomous, accountable and well-resourced, capable of creating a fair and competitive environment.
• Digital development – The policy calls for the immediate pooling of public sector demand for broadband so that high-quality connectivity and services can be procured to meet all public sector needs, including health, education, safety and security, while reaching beyond the current areas of connectivity. Indicators of success will be: speed, quality and cost of network capacity at government facilities; speed of rollout and quality of service; network reach and price of access; increased investment by network operators; and take-up of services in the public sector, schools and clinics.
• Digital future – A roadmap has been promised dealing with open access wholesale fibre and wireless broadband networks, including how existing private networks and new investors can be incorporated into a multi-layer entity offering cost-effective and non-discriminatory services. Indicators of success will be: network reach; cost to communicate; speed, quality and cost of network capacity; speed of rollout; increased investment by network operators; and improved penetration.
• Digital opportunity – E-readiness programmes should be rolled out in schools and clinics, along with formal skills development for high-level skills acquisition. Programmes and incentives should also be in place for the development of local content and applications. Indicators of success will be: demand stimulation through local content funding; supporting apps to market; e-government services; ICT start-ups; registration of ICT-related patents; increase in demand-side skills such as ICT specialists, engineers, lawyers, economists; increase in the number of PhDs in ICT; increase in the percentage of spend on R&D; increase in demand-side skills through e-literacy campaigns and ICT in the school curriculum.
SA Connect aims for a universal average download speed across South Africa of 100Mbps by 2030. To reach this, an average user experience speed of 5Mbps is targeted by 2016, available to 50% of the population and to 90% by 2020.
“As of today, there are some concerns about whether the policy is still alive,” says Parle. “But it needs to be alive.
“Our message is don’t panic. SA Connect is well-intentioned and put together. It is a structure and a start: at least we have a plan. There will be different technologies that go into making it a reality and we need to look at which are relevant for specific purposes.”
Paying for it
Perhaps the biggest challenge to the realisation of SA Connect is the amount of money that will be required to make it a reality.
Envir Fraser from Convergence Partners estimates that the total investment needed to take South Africa to a broadband future could be between R50-billion and R100-billion, spent on a combination of national long-distance networks and mobile solutions.
Part of the investment in infrastructure should be made by the public sector, he says, but there are good returns to be made by the private sector as well.
“Most of the market so far have been driven by private sector investment and we see that continuing,” he says. “In terms of infrastructure, there is a significant gap that needs to be filled, with the understanding that there is a market to support it.”
Margin and price declines for telecommunications services, however, are making it more difficult to justify big infrastructure investments as market trends move towards a service-oriented culture.
“The market environment is changing significantly from being technology-centric with a clear differentiation between mobile, fixed, ISP and broadcast networks. It is changing to a combined service offering where the customer is looking for services regardless of where he is in the enterprise, or which wholesale service he is consuming.
“SA Connect emphasises this change.”
Fraser thinks we will see more consolidation in the market. “To reach the price points that we need to get to, scale is needed at a regional, national and international level.
“We will see more services delivered at significant scale and at lower price points. And the operator space is going to change significantly.”
But this offers an investment case, with opportunities for at least the next five years, and possibly 10.
“In the public sector environment, policies and regulations will be key to enable the delivery of more extensive shared infrastructure. We need competition, and the development of a more widespread open access environment will be critical.”
For the private sector, Fraser says creating a buyers’ club type of environment and aggregating demand will create opportunities. To reach the target of 100% broadband coverage, he says partnerships will be key and intervention from the public sector inevitable.
“Importantly, there will be a significant investment required in fibre, wireless, mobile and other access technologies like satellite. The bottom line is that there is significant opportunity.”
Rural broadband
The cost of supplying broadband in South Africa has come down dramatically since the first submarine cable was landed just a few years ago, with the cost of local transit dropping from about R10 000 per Mbps in 2009 to R1 000.00 per Mbps in 2014.
William Stucke of Stucke Associates points out that up to two dozen private companies are doing a good job of rolling out fibre to suburban neighbourhoods – but villages in rural areas aren’t receiving the same attention.
“And under-serviced areas get a double whammy in that they tend to combine a low population density with a low income per capita,” he says, “The total revenue to be had from them is very small so it’s hard for private companies to justify an investment in fibre.”
He does not believe the incumbent telcos will succeed in running fibre to the underserviced areas either. “Their business model is highly capital-intensive and they are used to getting high returns on their investments. Particularly the mobile network operators (MNOs) don’t really know how to run a business with a margin below 10%.
“So spending million for peanuts in return isn’t a viable business opportunity.”
Smaller companies will start making a difference in remote areas, he believes, with small cells able to connect small populations within a limited area.
But to start making inroads into rural broadband connectivity, Stucke says an enabling environment is necessary, along with high-demand spectrum and the completion of digital terrestrial television (DTT) – which hasn’t started yet.
“We also need a rapid deployment policy to be published, regulation drafted and to get TVWS (television white spaces) and dynamic spectrum assignment in place.
“We need more spectrum to be allocated to licence-exempt companies. And we need investors with big balls.”
Not just about the ARPU
While the rapid roll-out of fibre to the home in parts of our cities is cause for celebration, it doesn’t really help to bring connectivity to the unconnected.
Andre Hoffman, president of the SAIEE, points out that fibre to the home (FTTH) initiatives are in the high-income areas – so the users who already have access to ADSL and 3G are now simply getting fibre as well, while those who don’t have ADSL are still getting nothing.
“This perpetuates the situation where, economically, we are creating a bigger and bigger gap,” he says.
The reason the fibre providers target the leafy suburbs is because they offer the right combination of density and ARPU (average revenue per user), Hoffman says.
But they could be missing out on the opportunities to be found in the townships.
“There is plenty of density in the urban townships,” he says. “And, yes, there is money for SIM cards and for DSTV. Content and communication is important to township residents and takes up 20% or more of their disposable income.
“So I’m not sure if the traditional arguments of density and ARPU hold up. It may be true in the rural areas, although there is fibre running through the rural areas as well.”
The gap in connectivity concerns Hoffman, and he believes we need to start thinking differently about connectivity, which can be used to improve content and its delivery.
“There is a lot of infrastructure out there, and we just need to think about different ways of getting it to people. And we need to get the costs down to R250 per month, which will require huge effort.
“We could continue to spend all our money in the right areas, but if we don’t extend the broadband enabler to more people, the general economy will no longer live. It’s not just a question of ARPU.”
Achieving the vision
While SA Connect sets out specific targets for broadband coverage, the true value of connectivity will be in the services that are delivered.
Dr Angus Hay, chief technology officer at Neotel, points out that the policy optimistically calls for broadband to reach all South Africa households by 2030.
If this is to be true broadband, he says it should offer high definition video speeds of 10Mbps or more, low contention, be unshaped with low latency; offer multiple sources of content; run peer to peer; and deliver QoS (quality of service) for at least some services.
Technical requirements would be high bandwidth, low contention access; L3 functionality close to the edge; very high bandwidth and a highly available IP core; low latency Internet for both local and global access; and distributed content and caching.
By 2030, 60% of South Africans will live in urban areas – and this means that 40% will still live in rural areas.
“Solving the urban and rural problems are different,” Dr Hay says. “And you need to understand why people are moving from rural to urban areas in the first place. Part of the reason is that the things they want for economic development are available in the urban areas – things like broadband, jobs, development and education.”
However, Dr Hay points out that the fallacy propagated by the telecommunications industry is that coverage is the same as connectivity.
“True broadband Internet access is not the mobile Internet,” he says, adding that the architecture for achieving maximum reach and access in widely dispersed towns and villages could be a set of low-cost high-bandwidth fixed wireless or optimal fibre links.
“This approach is both affordable and practical to implement within a reasonable timeframe.”
What needs to be determined is the correct mix of technologies that will be used to connect all South Africans.
“Whatever we do, fibre is always going to be part of it,” says Dr Hay. “And actually rolling out the fibre isn’t that hard. There are two ways of lighting the fibre: either active Ethernet or passive optical networks which can offer speeds from 100Mbps right up to 40Gbps.
“In terms of fibre, there is no question which is the way to do, and we have a good roadmap.”
LTE is another technology that can offer fixed broadband at speed of up to 20Mbps. “LTE offers a predictable speed and a way to predict performance,” says Dr Hay. “There is a nice roadmap for this well; and with decent spectrum there is a good future for LTE.”
LTE is also available in unlicensed spectrum, he adds. “LTE technologies will become an important part of the ecosystem, whether it’s used in a licenced fashion or not.”
Fifth generation (5G) technology is still under development, but could increase the speeds we see today by a factor of 1 000 and radically increasing availability.
“We will get to a point where 5G will be pervasive, but it will be largely in urban areas and won’t really make a difference in rural areas,” he says.
WiFi has a big role to play in rolling out broadband because it is rapidly becoming ubiquitous, with 60% of wireless traffic from devices running on WiFi.
“WiFi will continue to be the bearer technology for a lot of access,” Dr Hay believes. “When we’ve connected the schools and the hospitals, WiFi is going to be used for the end connection.
“The trick is to make it work – but it can be very cost-effective, offer efficiency and not require mass mobile rollouts to set up.”
Other technologies that will play a role in making SA Connect a reality include microwave and satellite, he says.
Dr Hay cautions that SA Connect points to specific coverage targets by a certain date – but that we might be chasing the wrong targets.
“We need to try to understand what really drives broadband growth,” he says. “It won’t be driven by having stuff available, but by people wanting and and needing it. And that is about education, jobs, social development, the way we work as a society, and the urban rural divide. We don’t need to target access, but how many people are using it and making a difference.”