Innovation has pushed down site costs and increased site autonomy to make ultra-ultra rural sites increasingly viable. Historically, sparse populations and low ARPUs in rural areas have meant mobile coverage in rural Africa lagged behind urban Africa, but the gap appears to be narrowing. TowerXchange’s rural roundtable, hosted by Jocelyn Karakula, Deputy CTIO of Orange, was packed, while our closing panel focused on the challenges of providing commercial connectivity in unserved and underserved regions of Africa. The panel was chaired by our own Head of Research, Matthew Edwards and featured Neil Bartlett, Access Network Program Manager at Facebook, Dion Jerling, the Co-Founder of Connect Earth and Albert Gardiner, CEO of Fairwaves.
The challenge
Africa’s population is growing at 2.7% per year, and is projected to hit 2.4bn by 2050 and 4.1bn by 2100. At the moment only half a billion are online, with most of the easiest to connect already connected. To achieve 100% SIM penetration by the end of the century would require more than 3bn net new connections and a significant increase in geographic coverage.
That lack of coverage is now the key barrier to getting more Africans online – the networks are just not available where many people need them. We need a new approach to bring connectivity to these areas, and at the 7th Annual TowerXchange Meetup Africa, MTN Group’s rapid rural roll-out RFP was highlighted as a positive example of innovation. Other MNOs are also exploring new technologies, and importantly, new business models to find a way to make the economics of rural connectivity work. But the economics will remain challenging at sites serving subscribers with monthly ARPUs around or below US$2.
Many African countries are landlocked, and even those with an undersea cable landing point can be large enough that their interior regions are effectively cut-off from the outside world, especially during rainy seasons. Another barrier is the cost of sites. A useful rule of thumb is that installing a typical cell tower costs US$100k, with amplified construction and maintenance costs in areas lacking good transport infrastructure, so you cannot deliver a return on that investment in conventional cell sites in Africa’s most remote, often poorest, neighbourhoods. Areas which lack a robust business case for mobile coverage also lack reliable energy and sometimes also lack even the distribution networks necessary to run a remote genset.
The energy challenge
As at all African cell sites, energy was seen as a primary challenge for rural connectivity, and just like the rest of Africa, there is no single answer. Rural connectivity is a broad category, taking in settlements of 1,000 which can only just support a 2G connection to 10,000 person settlements. Each site requires its own solution, even if some level of standardisation is necessary to achieve scale. The main priority in our discussions was the importance of minimising base station energy demand and correctly sizing generating assets. Demand can be anything from as low as 100W up to a few kW, depending on the density and size of the population being served by the cell site.
While one strategy for rural cell site power is to strip things back for maximum simplicity, another approach was also proposed at the Meetup, one which suggested a much more complex approach. Instead of stripping back energy generation to the minimum necessary, cell sites could instead become the power source for the local community; either acting as the generating site for a local minigrid, including connecting and billing other customers and maintaining a wider grid, or acting as the anchor tenant for another company that wants to offer distributed generation. These proposals remain largely untested, but point to a route to potential scale for ESCOs or towerco entrepreneurs.
Backhaul options
Where one lacks line of site microwave backhaul, a satellite connection remains the only real game in town. Our panel discussed some of VSAT backhaul’s real and perceived drawbacks. At remote sites with low ARPU the cost of a VSAT link can be prohibitive. For 2G and 2.5G sites VSAT’s bandwidth is more than sufficient, but for 3G and 4G sites the bandwidth can be too low. Huawei’s RuralStar’s non line-of-site wireless backhaul can help link up rural sites which are not too far from an operators main network, but this limits the solution’s range and still creates bottlenecks on existing backhaul. Other alternatives like Google’s Loon balloons have been cleared for launch in Uganda and ambitious low earth orbit satellite constellations are planned by the likes of SpaceX. However, major disruptions to traditional satellite backhaul remain 3-5 years away, so current backhaul solutions will remain a core element of any rural connectivity programme for at least half the lifespan of any site built in the next year.
OpenRAN and active sharing
Active infrastructure sharing has generally been resisted by MNOs and towercos: MNOs because they view their network as a source of competitive advantage and a new cell site as an opportunity to acquire long-term customers; and towercos because a shared tower is more valuable to them than a shared tenancy. However, our panel agreed that these old resistances will melt in the context of the push into rural areas and the desire to push down costs.
Sharing the radio side of a network allows for capex and opex to be shared on marginal cell sites, at the cost of exclusivity in a local market. One technology which will enable active sharing is OpenRAN. Encouraged by Facebook’s Telecom Infrastructure Programmme, OpenRAN solutions are based on general purpose vendor-neutral hardware and software-defined technology, which enable the cost of a base station to be reduced. Reducing the cost of active equipment on a cell site is not normally a concern of towercos, which have historically managed only passive infrastructure assets, but activie infrastructure will have to become more important if towercos want to be significant players in rural connectivity.
Quality of service
Rural sites need to be cheaper to deliver. If they are to be commercially sustainable, that comes with trade-offs. Those trade-offs are inevitable when bringing 3G to a site which will have only hundreds of subscribers at ARPUs of US$2 or less. Quality of service will have to be sacrificed at some rural sites to enhance the economics. However, MNO attendees at our rural connectivity roundtable were wary of building rural sites in some areas, knowing that regulators would penalise them for sites which fell below certain quality of service indicators. This meant some areas which could be served were not served because fines would make individually feasible sites uneconomic at a network level. Regulatory KPIs need to change if rural connectivity is to be able to expand as rapidly and as far as is technically and economically possible.
Where sites are needed
Finding the right location for a rural site is difficult. Urban sites are generally areas of current or future high traffic, but by definition the rural sites we are targeting will not be serving large amounts of voice or data traffic, but they still need to be near a critical mass of potential subscribers. Censuses in many African countries are now very out of date. For example, data from some regions of the DRC was reportedly being last reliably collected in the 1930s. Therefore, alternative methods for initial site surveys are needed.
We heard that in the early days of network building, rural sites would often be built near the local chief’s house, but because the chief often lives kilometers from the main population centre the site would be largely useless. If rural connectivity is going to be a success then site selection needs to improve so that the maximum number of people are covered at the lowest cost.
Facebook has had success combining satellite imagery of housing with nightlight imagery and census data to find appropriate locations to survey. Site visits are then necessary to review conditions on the ground to check local support and community engagement, access for commissioning and maintenance, and local vendors to support development locally.
Site design
40m towers can of course cover a wide area with a level of service, especially 2G, but are relatively expensive to build, maintain, and require a lot of power to transmit power across their service area. Even 20m towers may be excessive for some areas where a lower antenna, with a lower power requirement can cover a wide enough area to cover most people in a rural settlement.
Some rural sites will one day be less rural and can support the installation of a higher tower ahead of simple current commercial viability, but at many rural sites a 12m monopole will be sufficient to provide the necessary initial coverage. A short tower with simple foundations is easier to erect, and a quicker erection will be cheaper. Using guyed-masts also makes for more rapid and lower cost site build, a strategy which has been deployed by Helios Towers in Tanzania.
MTN Group have been asking for sites which cost no more than US$10,000. Cost reductions of this magnitude are elusive, but reducing steel content, simplifying foundations, accelerating erection and matching brackets to equipment needs will all play a role in cutting costs.
Making it pay
Rural sites have lacked investment not only because total revenues are low, but because it has been difficult to divide up the lower revenues so that everyone investing in a site can earn a return on their time and money invested. Business model innovation and a shift from capex-led investment by MNOs to opex-based or revenue sharing models led by network experts is enabling more rural sites to receive investment.
Africa Mobile Networks (AMN) was pointed to by our panelists as a key rural connectivity enabler (see infobox). AMN deploys its own capex to build a cell site at locations it thinks are commercial viable and connects them to a mobile network’s core network via VSAT on a revenue share basis. At sites it does not want to own it will still build sites for MNOs on but on a fixed priced opex-basis. 500 sites have been built and AMN is targeting 20,000 sites by 2030.
This model allows MNOs to retain use of their core systems and extend access to value-added services, while improving network efficiency. It also allows AMN to share in value generation and earn a return on its investment. Other firms offer similar models in rural areas. Alternative revenue streams can also support connectivity. In rural areas ISPs could support a Wi-Fi connection by supplementing it with “Freemium” access upgrades, or through selling advertising space online or on billboards.
There may also be other combinations of capex sharing, opex sharing or revenue attribution which could open up more sites to investment. Experimentation will be essential and each panelist at the TowerXchange Meetup Africa 2019 was experimenting in their own way to try to advance this cause.
How to move forward
The ultimate challenge is economic. How cheaply can sites be installed, and how cheaply can they be operated? How many subscribers will they serve in the immediate and long term, how much can those subscribers afford to spend, and how will revenues be shared over the life of a site? Sparsely populated and very poor areas will remain hard to cover, but innovations in site design, cheaper radio-side technologies, and innovations in infrastructure sharing business models means that the other challenges are beginning to melt away.
Africa’s connectivity revolution has already proven a bright spot in the last two decades of global development, and advances to rural connectivity will see millions more people brought online over the next decade. Our panel and roundtable were excited by the opportunities, but changes will be necessary: to support this investment we will need to see a change in the regulatory environment, specifically more flexibility from regulators on quality of service so that network operators can innovate. New business models need to be supported by MNOs and towercos, active sharing must be allowed, indeed encouraged, in a rural context, and investors need to recognise the huge opportunity that still exists in African telecoms even at sites with subscribers who can only support an ARPU of US$2 or less per month. Those people need connectivity, and we can now bring it to them with economic models that make sense for everyone.
Internet para todos
In Latin America, more than 100 million people in rural areas don’t have internet access. Facebook, Telefónica, IDB Invest and CAF have joined forces to bring them connectivity through Internet Para Todos (internet for everyone). IPT is a rural telecoms infrastructure operator that will offer internet to six million people in Peru while trying to create a new deployment model that can be replicated in other rural areas across Latin America.
Using 2,750 small stations received from Telefónica that offer reduced radio coverage, and 380 bigger stations Internet Para Todos is already able to offer 2G technology. Their focus is now to deploy 4G technology on those 3,130 sites, which will guarantee internet access to 3.2mn Peruvians. Most of those sites transport signal through satellite, so another priority is to deploy fibre or microwave backhaul, depending on the needs of each area, in order to improve services and reduce satellite dependence. Furthermore, between 2020 and 2021, they will deploy 1,000 new sites that will allow them to reach the goal of offering connectivity to six million people.
IPT is based on the Rural Mobile Infrastructure Operator (OIMR) concept, a unique business model defined by a law published in 2013 by the Peruvian Government that allows Internet Para Todos to offer infrastructure and connectivity services to the MNOs as a wholesale business despite not having any spectrum. Currently, Telefónica is IPT’s only client but their plan is to close deals with other operators by offering strong QoS and maintenance capabilities. This model allows operators to notably reduce their capex as they don’t need any up front capex, nor do they have to pay maintenance fees – they only pay for carried traffic.
MTN Group’s rapid rural roll-out RFP
At the end of 2020 MTN Group closed an RFP for its rapid rural roll-out programme. MTN Group would like to roll out up to 5,000 new rural sites by the end of 2020 through a network of new turnkey partners offering innovative low-cost solutions and new business models. These new rural sites will extend MTN’s network and bring the benefits of voice, data, SMS, and digital services to more people in all their African markets. Big winners in the RFP were Nigerian firms Raeanna and Hotspot Networks, and rural connectivity specialists Vihaan Networks Limited (VNL), Vanu and Africa Mobile Networks.
Huawei’s RuralStar rural solution
The industry’s OEMs are also responding to demands for solutions which work in low ARPU areas. Huawei’s RuralStar solution was developed to overcome the high cost of providing 2G, 3G and 4G mobile connectivity in rural locations. The solution combines a number of different elements to achieve a low total cost of ownership. For backhaul, RuralStar eschews VSAT and uses a non line-of-site wireless solution to link rural sites to a site on the main network which shares its spare backhaul capacity. It uses a simple pole tower to reduce component cost and to make erection cheaper and easier, and it combines a small solar PV array with a battery back-up to reduce refuelling and maintenance requirements. For example, the RuralStar Lite, launched at Mobile World Congress in 2019 features power consumption as low as 200 watts, a tenth of a traditional base station. The low cost solution economises on power usage and therefore covers a lower population and smaller area than a traditional site, but still enables enough people to be covered cheaply enough for operators to earn a return on their investment in around three years.
Africa Mobile Networks
Africa Mobile Networks (AMN) is a rural infraco. AMN now owns and operates 500 sites through a commercial model based on a revenue share with operators. AMN offer mobile operators the chance to expand their networks without capex or opex risk because AMN builds, operates and maintains the sites and connects them to the operator’s core network. AMN is now also offering an opex-model which shares risks for sites more evenly. AMN operate in ten countries in Africa. To remain economical, AMN keep a tight control over the equipment they deploy, importing all their equipment from China and assembling in Africa to their own specifications. This combination of site design and business model innovation is what will enable greater rural connectivity in Africa. AMN has so far achieved 99.8% uptime. While it predominantly deploys single-tenant monopoles, AMN also builds 20m lattice structures which will be capable of supporting multiple tenancies at its most profitable sites. AMN is already operating profitably, and plans to reach 20,000 sites by 2030.
