Inside MTN’s plan to turn its towers into AI hubs

MTN Group’s investment as part of a grouping that includes Nvidia, Nokia, Cisco, AT&T, Booz Allen and Telecom Italia into AI RAN specialist ORAN Development Company (ODC) is part of the pan-African mobile operator’s ambition to pivot from being a traditional network operator into a platform business.

“We are looking at this opportunity as very much integrated into Ambition 2030, which really does include and cover AI as a significant contributor to the overall strategy,” Mazen Mroué, CEO for digital infrastructure at MTN Group, said in an interview with TechCentral on Monday.

“This very much aligns with the overall strategy to introduce cost-effective and affordable solutions by bringing AI into the RAN side of the network, which speaks to some of the problems we are facing in Africa.”

A RAN, or radio access network, is the portion of a mobile operator’s infrastructure that connects to user devices using radio frequency spectrum. The RAN sits at the edge of a mobile operator’s network, with other layers such as the transport layer – usually reliant on backhaul technologies like fibre – and the core network critical to completing the connectivity loop between devices and the rest of the internet.

MTN’s Ambition 2030 strategy focuses on three core pillars: fintech, digital infrastructure and the connected home. The digital infrastructure portion includes terrestrial fibre builds and an expansion into AI data centres planned for launch in South Africa and Nigeria.

AI in the RAN

According to Mroué, putting AI in the RAN layer will drive down the cost of deploying services and applications for third parties in use cases where data centre capacity may be too expensive or wasteful.

“We cannot continue building data centres without looking at ways to optimise capacity and efficiency, because every single workload requires connectivity, transport and many other elements. At the RAN layer, we can address some of the capacity and cost pressures by avoiding transport and other core processing responsibilities,” said Mroué.

Read: The hidden drag on South Africa’s mobile networks

The ODC investment allows MTN to position itself at the forefront of mobile network evolution, not only as it evolves its network for 5G, but also in anticipation of 6G and the high-bandwidth applications it will enable.

Mroué did not disclose how much MTN contributed to the $45-million total ODC raised in the funding round. MTN could have chosen simply to buy ODC’s AI RAN technology off the shelf instead of being an investor. But according to Mroué, part of the rationale for investing directly is so MTN can have a say in how AI RAN develops, especially regarding its suitability in the African markets where it operates.

“Not everything that’s being developed in the West can really suit and meet the needs of our continent. That is why we are ensuring that our voice is loud and is being heard. We also find it useful that there is a group of participants that represent different parts of the industry including chipsets, network equipment manufacturers and mobile operators.”

Uses for AI RAN technology are not limited to third parties. Optimising the network for efficiency in the face of fluctuating workloads and increasing complexity is one of AI RAN tech’s main objectives.

In a white paper titled “Autonomous intelligence in RAN”, network equipment specialist Ericsson argues that the complexity of network management grows exponentially with each successive generation of mobile technology.

‘Unique environment’

2G networks had around 500 tweakable parameters per base station, 3G had 1 000, 4G had 2 000, 5G has 3 000 and 6G will have more than 5 000 – more than any traditional rules-based system can handle.

MTN rival Vodacom has also moved to incorporate AI into its network management strategy.

TechCentral reported in November 2023 that Vodacom had partnered with chip maker Nvidia to build an AI-powered digital twin of its RAN in the Cape Town metropolitan area. Vodacom’s managing executive for group technology strategy, Ryan van den Bergh, said at the time that managing the complexity of multiple generations of connectivity protocols in a single network was one of the project’s key objectives.

“People have a variety of devices, ranging from 2G feature phones, to 3G and 4G smartphones, as well as high-end 5G phones. The network caters to each type of device using different components and in different ways … it gets very complicated very quickly,” Van den Bergh said then.

Read: MTN invests in AI network start-up alongside Nvidia

MTN faces similar challenges. According to Mroué, AI RAN will also help with optimisation in the “unique environment” in which MTN operates, where grid power is not a given and the company must power its own towers as efficiently as possible.  — (c) 2026 NewsCentral Media