Microsoft announced general availability (GA) of its telecom-focused Azure Operator Nexus platform that underpins the cloud giant’s 5G telecom aspirations. The GA launch also includes updates that are expected to be adopted by its “lighthouse” customer AT&T.

Microsoft showed a preview version of the platform at this year’s MWC Barcelona event. It provides operators with a pre-validated telecom platform that can run from the cloud to the edge and allows operators to select where to run their network functions in support of their 5G or 4G LTE network. The cloud is Microsoft’s Azure public cloud, with the edge being an operator’s on-premises location housing validated server hardware.

That platform has now been updated with new features designed to ease management and increase efficiency.

These include a new network fabric automaton that provides a single point of control to build, operate and manage carrier-grade a network fabric from the platform; a network packet broker that can capture, aggregate, filter and monitor traffic to help with service planning and troubleshooting; and updates to Microsoft’s Azure Kubernetes Service (AKS) to improve resilience, cluster availability and the capability to isolate and manage the performance of specific workloads.

Operator Nexus also gained native integration with Microsoft’s Azure Monitor service to help with management; a new service manager feature from Microsoft’s AIOPs platform; and a new security focused updates that include the Azure Lockbox, which allows operators control over access to data by Microsoft support engineers tasked with support issues.

Yousef Khalidi, corporate VP for the platform at Microsoft, explained in a blog post that AT&T was in a lab using the platform’s ability to provide more control over the placement of network function workloads, “with the intent to deploy them into production environments.” This taps into initial work in certifying network function vendors into the Operator Nexus ecosystem.

Khalidi cited work with Nokia in conducting an end-to-end 5G standalone (SA) call using the vendor’s 5G core container network functions (CNFs) running in Operator Nexus, and Ericsson getting ready to trial a cloud-native 5G core running on the Microsoft platform.

“Of equal significance, partners are now familiar with the process of what it entails to bring updated or additional network functions to this platform in the future,” Khalidi wrote. “Continued relationships with partners through the program will allow for a robust, trusted collection of network functions that are deployment-ready for Operator Nexus.”

Khalidi also noted specific features Microsoft is looking to add to the platform. This includes more network function providers to support radio access network (RAN), packet core, transport and enterprise services; a smaller hardware and software form factor to support RAN deployments; and the ability to integrate generative AI to help design, deploy and operate networks.

Microsoft has been working on its Azure-based telecom plans for several years. That initial work was to link its private Azure Edge Zones service, its Azure IoT Central platform, virtualized evolved packet core (vEPC) software it gained by acquiring Affirmed Networks, and cloud-native network functions it brought onboard when it acquired Metaswitch Networks.

These efforts were further boosted when Microsoft struck a deal to acquire AT&T’s extensive Network Cloud technology business in mid-2021.

Igal Elbaz, SVP and network CTO at AT&T, explained at the recent MWC event the carrier was currently running its entire 5G network on its initial virtualized cloud-based platform operating inside of its own data centers and on its own hardware, but plans to begin tapping into the new Microsoft iteration as it bolsters its 5G SA service.

Elbaz also noted that the Azure Nexus platform brings in automation and artificial intelligence (AI) capabilities that are becoming increasingly important to manage complex 5G network architectures.

“It’s table stakes today,” Elbaz said of the need for automation. “It’s critical to the success of building large-scale networks. We are continuing to introduce large-scale technologies into the network and when we design and we architect these we always think about how we’re going to operate this. What’s the level of automation, what’s the level of possibly AI over time, we all have this vision of autonomous networks.”