AT&T sold its Network Cloud 2.7 software stack and associated IP to Microsoft in 2021 as the operator moved to simplify the process of adding network functions while also accelerating its journey to cloud-native.

The deal allows AT&T to focus on serving its mobility customers instead of continuing to evolve its own network cloud, while Microsoft benefits from being able to sell what was initially called the Azure for Operators service.

“What we are seeing is all of the major carriers are moving a lot of their core network functions to the cloud, or at least to cloud-native architectures, as they’re modernising their network cores,” Bob O’Donnell, president and chief analyst at TECHnalysis Research told Mobile World Live (MWL).

While AT&T hit its goal to virtualise 75 per cent of its network by end-2020, onboarding virtual network functions (VNFs) became onerous as each vendor had their own version.

Operators needed to test and configure VNFs prior to onboarding them for specific uses.

O’Donnell noted moving network functions to cloud gives carriers the flexibility to scale them up or down very quickly.

AT&T’s mobile core network currently spans more than 60 cloud-native network functions (CNFs) and VNFs from 15 different vendors.

Dorota Blat, assistant VP for network integration roadmap at AT&T Labs, told MWL the operator is still picking the network functions it wants for its mobility layer, whether they’re VNFs or CNFs and certifying them to deliver services to its customers.

As part of its Azure for Operators service, Microsoft is collaborating with network function vendors including Ericsson and Nokia to ensure compatibility with its cloud.

Similar to Google Cloud’s Anthos for Telecom initiative, Microsoft is building an ecosystem of network function vendors.

“That makes the AT&T role, or the telco role, significantly easier,” Blat said. “We can now focus on mobility services, the customer experience and so forth. The cloud layer is the focus of Microsoft and that includes working with the ecosystem and network functions to make sure that they can run on the cloud.”

Cloud-native migration
Operators are moving to a cloud-native architecture to build and run scalable applications including network functions in environments spanning public, private and hybrid clouds.

In a keynote during the LF Networking Open Networking and Edge Executive Forum on 12 April, AT&T CTO Andre Fuetsch stated the carrier’s cloud strategy for its standalone mobility core made cloud-native a significant focus.

Futuriom founder and chief technology analyst Scott Raynovich noted large operators will continue to migrate to cloud-native functions and containerised platforms because the operating costs are much lower when it comes to maintaining software on their networks: “It also has the added benefit of faster upgrade cycles and service velocity.”

Shawn Hakl, VP of 5G strategy at Azure for Operators, told MWL some network elements will stay physical, others which will remain virtual and “hopefully” a number which “migrate to containerised.”

“We think the right answer is for things to go containerised, but we’re supporting virtualised network components as well,” Hakl noted.

While Blat and Hakl believe cloud-native network functions are the future for mobile operators, a hybrid world of virtualised network components and CNFs will co-exist for a few years.

O’Donnell stated because AT&T needs to provide uninterrupted services to its customers, there won’t be a hard shift to cloud-native. He predicts operators will deploy it in a single location of city to assess its performance.

Blat noted vendors still need to refactor their network functions to cloud-native from the ground up instead of containerising their existing functions.

“That’s really where the true benefits will be for operators like ourselves, where it’s easier to do software upgrades and things of that nature by taking advantage of cloud-native.”

“I think Microsoft is helping us accelerate that because not only do they bring the cloud technology by working with the network functions, they’re helping them get to being cloud ready.”

What’s next?
A glimpse of what’s on the horizon for combining AT&T’s network cloud with Microsoft’s technologies surfaced at MWC Barcelona 2022 with the unveiling of Azure Operator Distributed Services (AODS), the next-generation of services built on the operator’s former software technology.

At a macro level, AODS will enable operators to run workloads including core and RAN, mobile and voice, and OSS and BSS on a single carrier-grade hybrid platform.

AT&T and Microsoft are developing AODS with initial testing stages planned for later this year.

The editorial views expressed in this article are solely those of the author and will not necessarily reflect the views of the GSMA, its Members or Associate Members.