Welcome to the 10th annual review of telecommunications infrastructure development in Africa. Ten years is a bit of a milestone and perhaps a period worthy of some reflection, so for each section I’ll take a look back to 2014 and see what has changed. I think if anything has characterised 2023, it is the sense that investment activity in ICT infrastructure is back to pre-COVID levels if not even busier. This is reflected in part by the bumper crop of over 500 articles in this year’s review.
Undersea Cables
As in 2022, Google and Meta dominate the undersea cable new headlines with the 2africa and Equiano cables. Equiano went live in several countries in 2023 and the 2africa cable, although not yet operational, was successfully landed in many African countries and is scheduled to go live in 2024. Other regional cable systems went live including the Liquid-funded T3 cable connecting South Africa with Mauritius. Plans for even more regional cables emerged including the:
Amilcar Cabral cable connecting Cape Verde, The Gambia, Guinea, Guinea-Bissau, Liberia and Sierra Leone;
MEDUSA cable connecting Morocco, Algeria, Tunisia, and Egypt with European countries; and,
West Africa cable that will connect Maroc Telecom and its partners in Côte d’Ivoire, Togo, Benin, Gabon, and Mauritania with Europe.
It is hard to be sure of an undersea cable until it actually goes live. The SHARE cable, supported by the Senegalese government, which promised to connect Senegal and Cape Verde was much in the news in 2022 but completely absent in 2023. I see no reason why it wouldn’t proceed but the absence of news is curious.
2014
Turning the clock back 10 years takes us to a time when the first major wave of investments in undersea cables was coming to an end with several projects not making it beyond their fundraising PowerPoint presentations and others like the EASSy cable more than doubling their capacity from 4.7Tbps to 10Tbps. That seemed incredible at the time but pales in comparison to 2africa and Equiano, each of which which together represent a design capacity of 320Tbps. The arrival of Internet giants like Google, Meta, Microsoft and others into the world of submarine cables represents the single biggest shift in the last 10 years, bringing much needed investment resources but also shifting the power balance. Combined with geopolitical issues now clouding the undersea landscape, undersea cable landings have become a much more complex issue. African governments now must weigh issues of political alignment and digital colonialism when considering landing station agreements.
If the number of news articles on terrestrial fibre projects this year is anything to go by (nearly twice as many as last year) this is a boom year for investment. Notable is the continued rise of regional infrastructure providers including Liquid Intelligent Technologies, Paratus, and MTN wholesale operator spin-off Bayobab. I should probably include Orange in that list but they are typically less public in their dealings making them harder to assess.
2014
According to AfricaBandwidthMaps, in 2014, 371 million Africans lived with 25km of a fibre optic point of presence. In 2021, that figure reached 669 million and it is likely now more than double the original figure from 2014. That represents about 1.2 million kilometres of terrestrial fibre as of June 2022.
The one thing that hasn’t changed substantially since 2014 is that utilisation of terrestrial fibre capacity remains comparatively low, which can be attributed to the fact that prices for terrestrial fibre generally remain high. Private operators tend to extract whatever the market will bear and state-owned networks are often beholden to loan repayments and resist lowering costs. I have argued that a low, geographically flat rate on wholesale fibre access would actually lead to greater revenues for operators but the economics of scarcity continue to dominate. A notable exception is the Ugandan government which has taken deliberate steps to lower costs on their state-owned backbone. Also worth mentioning is Anambra State in Nigeria which has taken the bold step of cutting Rights of Way charges, the cost to operators to deploy fibre across private/public property, to zero.
A little more pricing transparency would go a long way. Little has been done to benchmark fibre pricing in the region. Regulators might consider asking fibre network operators to publish a basic rate card as the regulator in Botswana does, at least occasionally, for their state-owned fibre network.
Most major operators now offer FTTH services in larger cities in Africa. While the investment cost is higher than wireless technologies, the absence of upfront or ongoing spectrum fees balances out that equation to some degree. There is also the fact that fibre technologies have a longer lifespan than wireless and are much higher in overall capacity, as well as being readily upgradeable. This makes fibre an attractive investment for any operator that can afford the initial capital outlay. FTTH (along with Fixed Wireless Access – FWA) is changing the landscape of African telecommunications, which has historically been almost entirely dependent on mobile communication technologies. Now we see two types of services: mobile broadband which is typically capped and FTTH / FWA services offering uncapped broadband to homes and businesses. The capped versus uncapped distinction is not by definition but it does seem to be how the market is playing out. Demand for uncapped services is largely driven by a range of OTT streaming services from social media giants and online movie/television streaming services.
OTT news in 2023 revealed a sector in the midst of change. The massive growth of OTT as a means of media delivery has some African governments contemplating whether they should be taxed. In Nigeria, streaming veteran iRoko has had to reinvent itself, pivoting away from its African audience to more global integration with smart TV services. Meanwhile, new players are entering the market. South Africa’s Multichoice is taking another run at the streaming market with ShowMax 2.0. A group of French footballers are backing a platform called StarNews Africa in six West African Countries. And the Ethiopian News Agency announced plans to launch pan-African media platform.
This makes for a complex time. OTT services are gaining ground over more traditional media delivery platforms including satellite. Broadcasters are racing to adapt to the new reality. Meanwhile, giants like Netflix who, to their credit, have invested substantially in the development of African movies and television shows are seen with mixed feelings by African operators lamenting lost revenues and governments concerned about the disproportionate influence of international media platforms.
2014
All this has happened in the space of 10 years. 2014 was the year that FTTH really began to take off. Frustrated with ADSL speeds, a neighbourhood association in a Johannesburg suburb banded home-owners together to solicit a tender for FTTH services. This triggered a tidal wave of FTTH competition in South Africa which continues to lead the continent. Elsewhere Liquid Telecom (now Liquid Intelligent Technologies) announced FTTH roll-outs across Eastern and Southern Africa. In Kampala, Uganda, Project Link, a pioneering wholesale metro fibre initiative (sponsored by Google) had just launched. Now known as CSquared, it operates in Uganda, Ghana, Liberia, and Togo.
Netflix didn’t officially launch in African countries until 2016 but YouTube and other streaming media services were popular long before that.
If news stories on radio spectrum licensing are anything to go by, 2023 might be called the year of 5G with operators across the continent announcing 5G launches and deployments. Namibia and Uganda successfully auctioned off 5G frequencies. It is worth looking at what these 5G launches really mean though. To the average user, 5G’s advantages over 4G is much less profound than previous 3GPP generational changes. As long as networks are not oversaturated, 4G offers adequate speeds for the majority of broadband users.
Given that anyone wishing to use a 5G service would have to purchase a new phone that supports 5G frequencies, it may be hard to see the value proposition over 4G for the average consumer. It is not surprising then that many operators have chosen to implement their 5G frequencies using Fixed Wireless Access (FWA) technologies. FWA involves the use of Point to MultiPoint wireless technology that directs communication to fixed devices at users’ homes and business, where they would then access 5G locally via WiFi. This is an extremely valuable service but it represents a very different kind of service from blanket mobile access networks. And yet 5G continues to be hyped in ways that are hard to credit. Consider this excerpt from coverage of a 5G launch in Nigeria this last year.
“While we may think that 4G is a remarkable improvement in our internet experiences, 5G completely opens a new vista where anything and everything is possible. Right now, with 5G, we can only be limited by the limits of our imagination. 5G ushers in a new era of collaboration that breaks boundaries and presents a new wave of economic benefits for this generation and the next, and we cannot wait to see how young people all over the country explore this new world of instant and unbroken connections to unleash a revolution in creativity.”
ITWeb Africa – Airtel launches 5G network in Nigeria – 20 Jun 2023
This kind of disingenuous hyperbole is intended to help drive adoption but is frankly unhelpful to anyone. It is great to see spectrum becoming available for 5G but, equally important are 4G deployments which have become dramatically less expensive as the technology has matured. Policy makers and regulators should be thinking about spectrum from the point of view of inclusion and affordability as much as from the point of view of competition.
Notable by its absence in African countries has been the introduction of licensing frameworks for private 4G/5G networks, offering access to 5G spectrum on a more localised basis. Regulation for private networks has been spreading around the world but African regulators have been slow to consider it, perhaps out of concern at potentially undermining revenues from national 5G spectrum licenses. An exception here is South Africa which launched a consultation on ‘Dynamic Spectrum Access and Opportunistic Spectrum Management’ in 2023.
2014
Turning the clock back 10 years and we see LTE were going through a period of strong early growth with twice as many LTE networks launched in 2014 as the year before. Looking back, I see that I made similar comments about LTE and the affordability of LTE smartphones in 2014 as in my comments about 5G and handset affordability above. Perhaps affordable 5G handsets are just around the corner.
WiFi continues to be the technology that keeps on giving with every successive generation of WiFi offering greater and greater performance and ease of deployment while remaining remarkably inexpensive. The affordability of this technology and its license-exempt nature have made it a natural choice for governments, startups, and communities to choice to deploy as an affordable access solution.
As in 2022, 2023 saw several governments announce free-WiFi access initiative, including Uganda, Tanzania, South Africa, and Kenya. There have been many free-WiFi initiatives over the years but not much has been said about their sustainability. WiFi networks need proper maintenance and access to backhaul broadband. It is not always obvious how free-WiFi initiatives are intended to be sustained.
In some cases free-WiFi networks are the result of corporate sponsorship such as this planned network of 4000 WiFi hotspots being rolled out by Coca Cola in Kenya. In other cases, operators are attempting to sustain free-WiFi through an advertising model. ThinkWiFi (which took over Google Station’s network when they abandoned ship in 2020) has just launched an ad-driven WiFi network in Kenya. Whether an ad-driven free-WiFi network can sustain itself remains to be seen. A 2017 study by Caribou Digital investigated the question but did not reach a definitive conclusion.
Ten years ago, WiFi was still struggling for legitimacy as an access technology in many African countries. In countries like Zimbabwe and Namibia, WiFi was restricted from commercial use without a license. In Malawi, the annual device fee was almost as much as the cost of the equipment itself. Public WiFi initiatives were just getting off ground:
In South Africa, a freemium public WiFi initiative called Project Isiswe had just launched;
Rwanda had deployed some public WiFi hotspots as part of its Smart Kigali initiative; and,
Kenya was struggling with some initial government-funded WiFi deployments in Nakuru.
Meanwhile, hopes were high among TVWS pioneers with both Microsoft and Google supporting TVWS pilots across the continent although formal regulations to permit TVWS use were proving elusive. Mawingu, a thriving rural wireless ISP in Kenya, got its start as a TVWS operator but abandoned it for WiFi when regulations were not forthcoming.
It is increasingly rare for a mobile network operator to own and operate its own towers. Orange is a notable holdout but it said to be considering the sale of its tower portfolio. A small number of tower companies dominate the landscape including American Tower Corporation (ATC), IHS Towers, Helios Towers, and relative newcomer EastCastle. Tower outsourcing makes sense from the point of view of infrastructure sharing in that tower companies are incentivised to offer tower access to multiple operators which should lead to cost-savings. However, very little detail seems to make it into the public domain regarding the actual level of infrastructure sharing that goes on with tower companies.
2014
Tower outsourcing was already happening in 2014 but I wasn’t paying much attention to it. I didn’t grasp its importance at the time.
For fibre network operators, both undersea and terrestrial, data centres are a natural extension of their business model. 2023 saw many fibre operators, including Liquid, Paratus, MainOne, Angola Cables, and Orange announce new data centres. Alongside fibre operators, independent datacentre companies appear to be flourishing. Africa Data Centres (ADC), Open Access Data Centres (OADC), Wingu Africa, Raxio, iColo, and Teraco all announced datacentre projects across the continent in 2023. The distinction between fibre operators and independent datacentre companies is a little bit fuzzy however. For example, ADC are owned by Liquid and WIOCC is a major shareholder in OADC. The map of independent companies is made more complex as datacentre giants like Digital Realty have acquired Teraco and iColo and Equinix has acquired MainOne. In short, datacentres in Africa are both expanding and consolidating rapidly in the face of growing demand and international investment, with South Africa leading the way. Of the 121 datacentres across the continent, 55 are in South Africa, according to this article.
To build a proper picture of internet infrastructure in Africa, it would be lovely to have an integrated map of terrestrial fibre networks together with a map of IXPs, CDNs and datacentres.
2014
Ten years ago, carrier neutral datacentres were in their infancy on the continent. While some national operators had their own in-house data facilities, they were not what one would now think of as a commercial datacentre. Pioneers included South Africa’s Teraco, which opened its first datacentre in 2010 and Kenya Data Centres (now Africa Data Centres) which launched in 2011 in Kenya.
One operator completely dominated African satellite news in 2023: Starlink. It seemed that every week brought a new story of Starlink being licensed in country after country (or not licensed as the case may be). Ministers declared 100% broadband coverage and happily toured SpaceX / Starlink facilities. I have set out my own thoughts on the potential impact of Starlink and don’t need to repeat them here. Suffice to say that LEO satellite constellations are a new player in the access space and Starlink is a clear front runner. OneWeb completed its constellation in 2023 and was expected begin operations in late in the year but appears to be delayed, at least in Africa.
Many other interesting things were happening in the satellite world that did not, perhaps, receive the attention they deserved. HTS satellite broadband services expanded across the continent. YahClick began offering broadband services in Nigeria and South Sudan. Eutelsat partnered with Blueline in Madagascar to offer affordable satellite broadband. Angola launched its own satellite (Angosat-2) to connect more than 150 locations across the country. Gilat partnered with Ethio Telecom to modernise their satellite services. Both Tanzania and Cameroon announced their space satellite ambitions in 2023.
2014
In 2014, satellite broadband was only for the wealthy as most existing satellite offerings on the continent were not designed for broadband. Newer generation High Throughput Satellites were just coming online. HTS pioneer YahSat was available in a few African countries. Middle Earth Orbit (MEO) constellation O3B went operational in September of 2014 but was only financially practical as backhaul for operators in remote areas.
With over a hundred articles that fell in to the miscellaneous category this year it is hard to choose what to single out. I think if there is one thing I would focus on it is the evolving relationship between the internet giants and Africa. Companies like Google, Meta, Microsoft, and Amazon have brought connectivity, technology and investment to African countries. Often they have shaken up slower moving operators and spurred innovation. However, these companies have grown in size and dominance and are obviously in the business of serving their shareholders. African countries have begun to reconsider their relationship to Silicon Valley. Whether it is the Ugandan government introducing taxes on foreign tech firms or workers taking Meta, Uber, and OpenAI to court for unfair labour practices, Africans are beginning to evaluate and redefine their relationship to these companies.
As with last year, I feel compelled to single out the stealth success of Network-as-a-Service (NaaS) operators like Africa Mobile Networks (AMN), NuRAN, and Vanu who continue to grow in rural service delivery: Vanu in Liberia; NuRAN in Madagascar, Cameroon, and DRC; and AMN also in Madagascar.
All in all, telecom sector in Africa seems to have gad a very good year. From a competition perspective, it is encouraging to watch an ever growing shift from monolithic competition between mobile network operators to a wider and wider range of technologies and business models.