Summary
- UbuntuNet is a non-profit organisation registered in Malawi and governed by its members, serving the National Research and Education Networks (NRENs) of Eastern and Southern Africa; its documents use "association" and "Trust".
- It assembled regional capacity through contracts with operators, points of presence, member routes, and AfricaConnect programmes, including 10 Gbps backbone links and larger national connections.
- The portfolio now extends beyond transport to include identity, cloud and storage, open science, cybersecurity, and NREN development, with services of varying maturity.
- Aggregating demand and expertise creates value, but weak campus links, arrears, vendor power, conflicts, grant cycles, and incomplete information can prevent capacity from reaching researchers.
Network 3 brought the core to a 10 Gbps operational class
The 2023 report describes the move from links near 2.5 Gbps to main links of 10 Gbps. It also mentions at least 20 Gbps total capacity through redundancy. It is not correct to turn both statements into a universal speed.
The most coherent interpretation is a design with redundant 10 Gbps links on core segments, not 20 Gbps on every path or for every member. Rates vary by route, provider, and national access.
The change was significant for price and operating margin, though it remains far from networks of hundreds of gigabits or terabits. The comparison must reflect regional purchasing conditions, not present speed as the sole indicator of success.
Unit price reductions were an infrastructure outcome
The 2023 report cited reductions of between 50% and 71.5% compared to 2022. Some NRENs moved from services of 130 Mbps–4 Gbps to 500 Mbps–8 Gbps, depending on the case.
The improvement came from collective purchasing, long-term contracts, and a more efficient core. It shows that the objective was not just a capacity number, but delivering more per budget unit.
By passing savings to members, UbuntuNet keeps little cushion for currency, failures, or arrears. A low price is only sustainable when everyone understands and pays the recurring cost.
BotsREN and ZIMREN show how regional improvements reach NRENs
BotsREN initially connected at 10 Gbps in 2023 and upgraded to 20 Gbps in January 2026. ZIMREN connected at 10 Gbps in June 2025. These are member access speeds, not the capacity of the entire network.
Still, they show that backbone investment translates into identifiable national services. The final benefit depends on how many institutions lie behind, on domestic circuits, and on each campus’s equipment and processes.
A 20 Gbps national access can co-exist with universities limited to 1 Gbps, unstable power, or saturated firewalls. Connection figures must be accompanied by usage, loss, availability, and coverage.
The regional layer begins where national networks reach their borders
Whoever studies or researches rarely sees the name UbuntuNet Alliance. The first visible network is usually that of the campus or institute. Traffic then passes through a national NREN, such as KENET, RENU, TENET, ZAMREN, TERNET or MoRENet. UbuntuNet becomes decisive at the next boundary: when that national network needs to exchange traffic with another country, reach a scientific facility, connect to GÉANT, or use a common service too expensive to reproduce locally.
This position explains why the Alliance can be essential without resembling a retail provider. It does not sell residential broadband, does not connect all universities directly, and does not control campus Wi-Fi. It provides the regional layer of routes and services atop autonomous national systems. Other regional networks, carriers, clouds, exchange points, and the destination system complete the journey.
The user experiences a single session; operational authority is fragmented. UbuntuNet can modify its routes, expand a core link, monitor a regional service, and coordinate an incident. It cannot restore a university’s electricity, fix an institutional directory, force a national operator to repair fibre, or guarantee the capacity of a remote application. The task is to make multiple independent domains work together sufficiently coherently.
Federation is the design, not a temporary substitute for central ownership
UbuntuNet Alliance is an alliance of NRENs, not a parent company. Each member retains its own legal personality, staff, budget, domestic infrastructure, university relationships, government context, and pricing. The regional level exists because those organisations decide to purchase capacity jointly, exchange traffic at common points, operate shared services, and represent converging interests.
The federation suits a region where telecoms markets, currencies, university models, procurement rules, languages, and technical maturity differ. A centrally owned academic network across all countries would require political authority that UbuntuNet neither has nor needs. The model combines national autonomy with regional scale, and allows more mature NRENs to transfer experience to those still consolidating.
Complexity does not disappear; it concentrates at the interfaces. A slow transfer may involve a campus firewall, a national circuit, a regional PoP, an international carrier, and the destination system. Each actor sees only a part. That is why service boundaries, shared telemetry, escalation contacts, change rules, clear contributions, and real exit options are necessary.
The legal framework works, although public terminology is not uniform
Current material describes UbuntuNet as a non-profit association registered in Malawi. The 2023 report and transition documents also speak of a non-profit Trust governed by members and Trustees. The research pack did not retrieve the current statutory deed, so it is not appropriate to erase the difference.
The practical structure is better documented. UbuntuNet has members, not ordinary shareholders. Governance is exercised through a Members’ Assembly, a Board of Trustees, and a Secretariat led by the CEO. The organisation hires staff, receives grants, signs agreements, owns or leases assets, and operates services. There is no evidence of distributable equity or a conventional valuation.
Calling it a private telecoms company would suggest investor control and profit distribution. Calling it merely a voluntary community would underestimate contractual and fiduciary obligations. The safest wording is non-profit membership organisation registered in Malawi, noting that its sources use “association” and “Trust”.
The foundational problem was economic and geographic, not the absolute absence of internet
Universities already had internet before UbuntuNet. What was missing was a stable regional mechanism to aggregate academic demand, exchange research traffic, and connect to global R&E networks under terms suited to science and education. Institutions bought small amounts of international capacity at high prices, and sometimes traffic between African countries travelled via Europe.
A commercial connection gave web access, but did not automatically create direct routes between neighbouring NRENs, predictable capacity for large data, federated identity, shared security, or joint negotiating power. Small communities bought with little leverage, and emerging NRENs struggled to demonstrate value before having enough institutions.
The Alliance changed the unit of demand. Campuses aggregated at national scale, and NRENs at regional scale. That collective buyer could plan PoPs, sign long-term contracts, organise operations, and negotiate with global partners. UbuntuNet did not eliminate the commercial market; it entered it as a coordinated buyer and a specialised operator with a public mission.
Five NREN initiatives formed the first coalition
UbuntuNet Alliance was conceived in the second half of 2005 by initiatives from Kenya, Malawi, Mozambique, Rwanda, and South Africa. The founding basis is associated with KENET, MAREN, MoRENet, RwEdNet, and TENET, according to the institutional forms of that time. They did not start from the same level.
TENET and KENET brought operational experience and established university communities. MAREN provided an anchor in the country that would later host the Secretariat. MoRENet represented a state-backed national model. RwEdNet would show both the promise and the fragility of an NREN: its collapse and subsequent reconstruction proved that a regional network cannot indefinitely substitute for a sustainable national institution.
The mix was useful. A coalition of only mature networks might have remained an interconnection club; one formed only by weak initiatives would have lacked traffic and credibility. The combination allowed building production while at the same time developing national capacities.
Francis Tusubira helped turn a network proposal into an institution
Francis “Tusu” Tusubira is closely linked to the founding phase and to the African NREN movement. The task was not to choose routers. It was to convince universities, ministries, regulators, carriers, funders, and engineers that a regional research network must be a permanent institution, not a time-limited project.
That required a legal entity, membership rules, accounting, contractual capacity, an annual forum, and a common explanation of why an NREN is not simply a cheap ISP. Tusubira’s contribution must be understood within that coalition; it does not replace the work of the founding networks, boards, technical communities, and international partners.
The lasting result was an organisation capable of receiving programme funds, signing long-term capacity, and representing a regional community. That institutional capacity made AfricaConnect possible.
Amsterdam gave a first legal vehicle; Malawi became the regional seat
Registration documents were filed in Amsterdam in 2006. For an organisation still forming, a Dutch entity offered a framework recognised by European partners and facilitated contracts and projects before a stable base existed within the region.
As the organisation matured, a European legal seat became less coherent with the mission. In 2012 a new constitution was adopted in Malawi, registration followed in 2013, and historical documents describe the transfer of assets and liabilities. The Secretariat is now in Lilongwe.
The move brought governance closer to members and strengthened the African character of the institution. The correct timeline distinguishes conception in 2005, initial formalisation in the Netherlands, and subsequent transition to Malawi, without turning the whole process into a single founding date.
Early transit to GÉANT demonstrated usefulness before a full backbone
Historical sources indicate that in 2008 R&E transit towards GÉANT was established. Even limited, it gave members a route to European and global scientific networks. It showed that a regional institution could deliver operational value before assembling a mature backbone.
The relationship also brought purchasing power, project management, experience in identity and security, and access to EU funding. That cooperation accelerated development, but tied part of the expansion to European timelines and priorities.
It was not an infrastructure simply imported nor complete independence. UbuntuNet used global partners to build a regional capacity governed by its members. The outstanding question was what portion could be sustained after the projects ended.
Long-term carrier contracts turned aggregated demand into infrastructure
The fifteen-year agreements with WIOCC in 2013 and SEACOM in 2016 gave UbuntuNet a more predictable base than short-term low-capacity contracts. A long term can reduce unit cost, secure routes, improve resilience, and facilitate upgrades.
It also fixes obligations. The Alliance is exposed to provider performance, technological change, maintenance, currencies, and recurring payments. It may have to pay an international bill even if a national NREN falls behind.
These contracts show how the federation becomes infrastructure: one entity aggregates needs, signs on behalf of several members, installs equipment, and coordinates restoration. But it does not mean UbuntuNet owns the fibres, landing stations, or transport networks of the providers.
The 2014 service launch turned promise into operation
In July 2014, UbuntuNet Alliance and DANTE announced regional links at 622 Mbps. The figure is small by today’s standards, but it was relevant for networks that were buying much less at very high prices.
From that moment there was traffic, incidents, configurations, shifts, and expectations from institutional customers. A mistake ceased to be a project problem and began affecting universities and laboratories.
The first production created a reusable base: PoPs, NOC processes, supplier relationships, and billing mechanisms that could be scaled in later phases.
AfricaConnect2 expanded coverage and strengthened eastern resilience
AfricaConnect2 was announced in 2015 with €26.6 million for several regions and included UbuntuNet, WACREN, ASREN, and GÉANT. That amount is not UbuntuNet’s exclusive income, but the budget of a multi-party programme.
For the Alliance, the phase expanded links and membership and funded the SEACOM Kampala–Dar es Salaam–Amsterdam route. The new path increased diversity in the east and reduced dependency on a single route.
It also consolidated a federated pan-African model: several RRENs could develop their own infrastructure within a common framework. Results remained uneven due to national and market differences.
AfricaConnect3 made the service portfolio a core part of the mission
AfricaConnect3, with €37.5 million at programme level, sustained the network, pricing, training, identity, cloud, and open science until April 2025. The evaluation could no longer be limited to the number of circuits; it had to consider whether members could use the infrastructure.
The phase produced Network 3, new PoPs, faster national links, and lower prices. It also strengthened identity federation, repositories, Utafiti Africa, AfricArXiv, and cybersecurity.
The increase in functions is logical because a network without storage, identities, or staff can be underused. But each service adds software, data, security, support, and permanent costs.
AfricaConnect4 opens another investment window and a sustainability deadline
AfricaConnect4 began in 2026 with €40 million of additional European support over four years for sub-Saharan African regional networks. It is a programme amount, not an exclusive allocation to UbuntuNet.
The phase includes national and campus components, cybersecurity, stations and climate data, EUMETCast, federated computing, GPUs, and open science. It recognises that a fast backbone has no value if campuses cannot use it.
The 2030 horizon forces resolution of who will operate and fund stations, repositories, clouds, identities, and clusters when the grant ends. Each service needs an owner, a recurring budget, charging rules, staff, and a closure or migration plan.
AS36944 marks UbuntuNet’s public routing boundary
UbuntuNet operates AS36944. PeeringDB and BGP observers confirm the identity, some peers, prefixes, and RPKI validity of observed origins. They reveal the public boundary, not all private circuits.
PeeringDB shows a self-declared traffic range of 5–10 Gbps, 1,200 IPv4 prefixes, and 150 IPv6. bgp.tools observes fewer directly originated prefixes. They measure different things: declared announcements, customers, origins, and collector views.
Those data do not replace core capacity or a member’s access. The ASN confirms operational BGP participation, not availability, peak traffic, or physical diversity.
PoPs and peering create options, but a map is not an asset inventory
The 2023 report identifies locations in Cape Town, Mtunzini, Maputo, Dar es Salaam, Nairobi, Amsterdam, Gaborone, and Johannesburg, with Djibouti and Mombasa under development. Older pages mention London and other locations.
Differences may arise from dates, the distinction between core, handoff, and peering PoPs, or route changes. A point on a map does not say who owns the router, fibre, rack, port, or contract.
Johannesburg and NAPAfrica added options in the south. Real resilience depends on diversity of ducts, providers, power, data centres, and control planes; two drawn paths may share the same fibre.
The Lagos–Cape Town route made pan-African federation more concrete
Developments in 2025–2026 connected WACREN with the South African environment through ZAOXI, SANReN, and TENET, and from there with UbuntuNet. Partners described it as the first effective direct interconnection between African regional research networks on African soil without passing through Europe.
The claim must remain attributed and limited to RRENs. It does not imply that commercial or private links did not exist. The novelty lies in the institutional and operational relationship between regional research networks.
The route allows more scientific traffic to stay on the continent and reduces certain detours. Even so, it depends on carriers, exchanges, BGP, and national partners. No single entity controls the full path.
Membership counts reflect different categories and dates
The 2023 report recorded 15 active members, more than 3.5 million users, and over 1,000 institutions in 13 countries. An audit RFP in March 2026 listed NRENs from 18 countries; the website showed 16 profiles. They may measure legal membership, activity, connection, or website maintenance.
Lesotho was admitted in November 2025, and there were elections in 2026. The community keeps changing. Every total must be dated and defined.
Being a member does not prove an active connection, and a national connection does not imply equal service for all universities. Especially in conflict-affected countries, legal status and operation can diverge.
Campus infrastructure decides whether the regional investment is used
A fast backbone does not fix limited public evidence local fibre, old Wi-Fi, unstable power, a saturated firewall, or lack of staff. Experience is limited by the weakest link.
That is why AfricaConnect4 includes national and campus tiers. The regional network can remain idle if institutions lack storage, servers, transfer tools, or a prepared LAN.
Responsibility is shared: UbuntuNet coordinates the regional layer; the NREN builds the country; the university runs the campus; governments, carriers, and donors influence the environment. Measuring only national presence would hide the real problem.
Lilongwe hosts governance and Kampala concentrates network operations
The Secretariat is in Lilongwe, Malawi, while organisational material places the NOC in Kampala, Uganda. This separation is consistent with a distributed regional entity.
The Secretariat manages members, contracts, projects, and finances. The NOC monitors links, opens tickets, coordinates providers, and supports NRENs. The separation can improve resilience, but requires secure access, continuity, and precise responsibilities.
It also demands documented knowledge, shifts, credentials, and succession. A small organisation cannot depend on a single person to understand critical systems.
Shared operations must strengthen national capacity
Not all NRENs can sustain a 24/7 NOC. UbuntuNet can provide monitoring, preventive maintenance, and first- or second-level support. The national network retains customers, policies, providers, and domestic authority.
The service is valuable when it fills a gap and transfers knowledge. It is risky if the NREN stops understanding its configuration or cannot leave the agreement. Action rights, windows, logs, escalation, costs, and exit must be clear.
The regional NOC can diagnose and coordinate; it cannot repair national fibre without the provider. Its strength is shared evidence and operational memory.
Federated identity is a second network built on institutional trust
Connectivity moves packets; federation lets a service trust an identity verified by another institution. The university maintains the account, the national federation coordinates rules and metadata, and the regional and global layers exchange trust.
UbuntuNet does not store all passwords. It operates or supports proxies, federations, metadata, and access to eduGAIN. Quality depends on directories, registration and deregistration, MFA, certificates, and local attributes.
The advantage is being able to access libraries, clouds, repositories, and computing with home credentials. The risk is that a compromised identity can reach multiple services. Federation trust requires federated auditing and revocation.
Eduroam turns campus identity into mobility
Eduroam allows using home credentials at another institution. The request may pass through local, national, regional, and global proxies before returning to the home institution.
The simple experience rests on RADIUS, certificates, realms, secure Wi-Fi, and correct accounts. UbuntuNet can provide the regional layer, but cannot fix an expired password or a misconfigured access point.
Success must be measured by active institutions, authentications, incidents, and resolution, not merely by installing servers.
EduID.africa and eduGAIN offer a path to broader trust systems
EduID.africa, operated with WACREN and ASREN, serves as a hosting and reference federation for NRENs that do not yet have a mature national federation. It reduces the need to wait for full implementation in each country.
MAREN and TERNET joined eduGAIN in 2025, connecting the Malawi and Tanzania federations with the global interfederation. Admission certifies interoperability, not identical maturity on every campus.
The regional layer should accelerate national capacity and enable a clear migration. If it becomes a permanent and inscrutable substitute, the federation loses autonomy.
Research4Life turns federation into practical access to knowledge
The 2025 pilot allowed institutions in Kenya, Malawi, Uganda, and Zambia to access Research4Life resources with home credentials. Follow-up reported 66 institutions onboarded in 15 countries.
This reduces the use of shared passwords, which are hard to revoke and attribute. Access can be tied to a person and their affiliation without creating separate accounts for each resource.
The number is a snapshot, not universal coverage. Each institution must maintain identities, and Research4Life retains its rules. Still, it is clear proof of the value of federation for the researcher.
The cloud combines regional infrastructure with commercial procurement
One strand includes community infrastructure: UbuntuNet Open Science Cloud, transfer, Drive, hosting, collaboration, and experimental environments. The other uses commercial frameworks, notably the Redington/AWS agreement of 2024.
The regional cloud can retain control and local expertise; a hyperscaler offers breadth of services. UbuntuNet aggregates demand and support, but does not own AWS or control its availability, pricing, or jurisdiction.
The 2023 report documented installation in Lusaka and equipment for Kigali, without current figures for capacity, usage, tenants, or SLA. Deployed hardware does not equal a mature multi-region cloud. Identity, orchestration, backups, security, support, and funding are essential.
Support services solve ordinary frictions as well as large scientific workloads
Utafiti Africa aggregates funding and training opportunities and exceeded 1,100 active users in 2024. It does not award grants; it reduces the cost of discovering them.
File transfer, hosting, Drive, Meet, and Learn can avoid duplicated purchases and facilitate collaboration. Their maturity is not uniform, and some public descriptions mix current operation with old plans.
Sending a large file is not equivalent to running continuous streams from telescopes or climate monitoring. Those workloads need tuned endpoints, storage, observability, and sometimes dedicated circuits. Each service must have an explicit scope.
Open science became a second infrastructure layer
Connectivity does not make a result discoverable, citable, or preserved. Repositories, metadata, identifiers, and staff also depend on hosting, storage, identity, certificates, and backups.
UbuntuNet and Access 2 Perspectives have hosted AfricArXiv since late 2023. AfricArXiv remains community-led; it does not belong exclusively to the Alliance. UbuntuNet is also the Southern Node of the Africa Open Science Platform.
Repositories have a long lifecycle. They must maintain software, rights, data, and preservation, and enable export if a relationship changes. Abandoning a repository can harm the scholarly record more than replacing a circuit.
Repository and identifier stewardship demands institutional work
The Research Repository Stewardship Programme trains librarians and data stewards. In July 2026 it held its first Portuguese-language workshop in Mozambique, a sign that localisation is part of the infrastructure.
Institutions need policies, metadata, rights, support, security, backups, updating, and a budget. A multi-tenant platform reduces duplication and creates shared risk. The Network Adoption Fund aims to turn NRENs into sustainable operators.
DataCite, DOI, ORCID, and the 2026 DOCiD alliance with TCC Africa add identity for authors, institutions, and entities. They do not fix poor metadata. UbuntuNet can coordinate procurement and integration without hiding global governance or local responsibility.
Cybersecurity resides in distributed teams
UbuntuNet runs bootcamps, national workshops, and training in identity and systems. There is no evidence of a regional SOC controlling all members. The model strengthens local teams, contacts, and common practices.
Laws and authority for incidents remain national or institutional. The Alliance can share tools, but cannot automatically remediate a compromised host on a campus.
The 2026 financial training, with entities from 15 countries, shows that resilience also means being able to pay contracts, manage currencies, comply with grants, and collect contributions. Administration is infrastructure.
Governance offers continuity, although the public record has gaps
The Assembly governs, the Board of Trustees oversees, and the Secretariat executes. Madara Ogot was appointed CEO for four years from 1 February 2022 and continues appearing as CEO in 2026, although no separate renewal announcement was found. Hellicy Ng’ambi assumed the chairmanship on 1 January 2024.
Miriam Chahuruva and Sabelo Dlamini were elected Trustees on 29 May 2026. The sources do not provide a fully reconciled list after that election nor identify all departures.
The lack of a synchronised page does not prove a vacancy, but prevents precise assertion beyond verified roles and reinforces the need for an updated record.
The 2023 accounts reveal a large mission with a narrow margin
UbuntuNet reported $2,404,779 in income, $2,393,911 in expenses, and a surplus of $10,868. Connectivity contributed $1,902,318, or 79.1%, and network operation costs were $928,032.
The margin is very small against a regional mission. A currency movement, a bill, or an arrear can consume it. Passing procurement savings to NRENs fulfils the mission but reduces the reserve.
The CEO noted that late operation and maintenance payments severely affected the ability to meet obligations. No later audited accounts were found; current status should not be inferred from AfricaConnect budgets.
Donors, carriers, currencies, and conflict remain within the model
AfricaConnect funds expansions; carriers supply the physical layer; clouds control their platforms; contracts are often in hard currencies; and some members operate under conflict or instability.
RwEdNet collapsed in 2018 and returned in 2024, proof that an NREN can fail even when a regional backbone exists. Membership from conflict-affected countries does not guarantee continuous service.
The work consists of diversifying, developing national capacities, documenting responsibilities, and preventing sovereignty from meaning isolation. UbuntuNet increases options; it does not eliminate the reality of providers and partners.
Public evidence is strong on programmes and weaker on current operations
There is abundant documentation on contracts, links, elections, training, and launches. Missing are a route-by-route inventory, availability history, full traffic series, per-link utilisation, incident database, current headcount, and post-2023 accounts.
PeeringDB, bgp.tools, access rates, and core capacity measure different entities. A PoP can represent technical or contractual presence without revealing the physical asset. An independent SLA cannot be calculated from those pieces.
The absence of a public incident does not mean absence of failure, and lack of data does not mean poor operation. Publishing availability, restoration, usage, and adoption would help members and funders assess whether projects have become sustainable services.
Regional autonomy is created with options, not by eliminating all external dependency
UbuntuNet gains control by aggregating purchases, keeping more traffic in Africa, training engineers, and operating regional services. It cannot eliminate global standards, commercial cables, equipment vendors, hyperscalers, or scientific partners.
The relationship with WACREN, ASREN, SANReN, TENET, and GÉANT shows a pan-African federation without a single owner. Lagos–Cape Town improves route localisation but still depends on carriers and shared policies.
Every option has a trade-off: AWS eases access and increases dependency; a regional repository strengthens governance and demands software and staff; a long-term contract lowers prices and increases provider and currency risk.
The test is reversibility. Can a member change provider, export data, move workloads, and keep its identity? Autonomy grows when it can do so without losing the regional community.
UbuntuNet transforms separate networks into a usable research environment
The achievement is technical and institutional. Independent national networks aggregate demand, share operations, connect internationally, and build services that would be expensive or non-existent separately. The backbone, identity, repositories, cloud, and training are distinct layers of the same answer to fragmentation.
The organisation remains incomplete. Its capacity is modest compared with climate, astronomy, and AI ambitions; public information describes milestones better than performance; national finances are uneven; and providers, donors, and services at different stages remain part of the system.
This does not diminish its importance. UbuntuNet is not a separate African internet nor a finished continental platform. It is the shared layer that makes a larger system possible. Success will consist in members being able to use, pay for, monitor, and leave services without losing autonomy.