Progressive communications service providers (CSPs) are considering the benefits of transforming their current appliance-based networks into cloud-native alternatives that use containers and microservices on public, private or hybrid clouds, to run their high-performance service workloads.
Migrating from established network architectures, where network resources are locked in silos of vertically integrated technology stacks, to decomposed network functions deployed on a cloud-native infrastructure is both challenging and exciting. But unlike in other industries, where IT departments have quickly embraced the public cloud, typically risk-averse CSPs are taking their time to work through a set of common cloud prerequisites. Ultimately, it is clear that a hybrid public and private cloud architecture that can support the demands of real-time voice and data network functions along with the delivery of mission-critical applications is required.
Embracing cloud benefits
In some instances (e.g. 5G SA), a cloud-native architecture is mandated by industry standards, but there are many reasons why operators are now looking at the benefits offered by a hybrid cloud, carrier-grade platform for deploying all network workloads. These include:
- Simplified deployment of workloads and services using intent-based configuration
- Support for best-in-class software solutions at all levels of the technology stack
- Automated lifecycle management for network functions
- AI-driven orchestration and optimization of network resources
- Fine-grained scaling of resources, on demand
- Ready integration of cloud applications for new service creation
- Developer-ready platforms for accessing network APIs
- Simplified system management and monitoring using powerful cloud tools
- Ability to quickly analyze massive amounts of network data
- AI-enabled fault identification and resolution
- Ubiquitous computing from cloud to edge supporting low latency applications
These benefits were hard to realize as CSPs embarked on building their own private cloud solutions. While significant advances were made, the tasks of managing and maintaining a cloud infrastructure proved onerous. So as operators started to move forward with their 5G SA core deployments, many started to consider the benefits of using a hyperscale-based infrastructure and automation framework developed by the public cloud providers.
As such, CSPs now view hyperscalers like Microsoft as close collaborators on their cloud transformation journey, looking to take advantage of established, and rapidly evolving, cloud infrastructure and services to fully deliver on the promise of software-defined networks and operations automation. Microsoft is filling the role of a trusted advisor that can help CSPs achieve their cloud goals, and not as a competitor in the battle for customers.
Collaborating with hyperscalers
In June 2021 AT&T announced plans to move its 5G mobile network to Microsoft Azure, an alliance that would provide a path for all of AT&T’s mobile network traffic to be managed using Azure technology. The agreement allowed Microsoft to acquire AT&T’s carrier-grade Network Cloud platform technology, which had been running AT&T’s 5G core at scale since launch. Key to the announcement was that AT&T would continue to operate its network and retain its customer relationships.
In addition to the acquisitions of both Affirmed Networks and Metaswitch, Network Cloud gave Microsoft both the “telco DNA” and a proven operator cloud solution to integrate under the Azure umbrella. Microsoft committed to making the platform applicable to other CSPs through their Azure for Operators initiative, and to continue building an ecosystem of partners who could deploy their network functions and OSS/BSS solutions on the cloud platform. As a result, the platform was recently announced as generally available under the rebranded Azure Operator Nexus moniker, accompanied by the expanding “Nexus Ready” program for certified network functions from ecosystem partners.
Azure Operator Nexus
Microsoft’s Azure Operator Nexus is a carrier-grade, hybrid cloud platform that enables CSPs to lower total cost of ownership, drive operational efficiency and resiliency with advanced AI and automation, and improve the security of highly distributed, software-based networks. Azure Operator Nexus is not a public cloud solution, but a hybrid cloud platform that includes dedicated hardware for deployment on the operator premises. Managed through familiar Azure services, the on-premises hardware is configured from the cloud, as are the network functions that are deployed on Azure Kubernetes clusters at the edge.
Azure Operator Nexus is integrated with a portfolio of AIOps offers that help CSPs drive efficiency, speed and accuracy in network operations with automation and AI. Azure Operator Insights provides unified visibility from disaggregated data sources for end-to-end analytical and business insights, while Azure Operator Service Manager automates and orchestrates the deployment of cloud-native network functions. Microsoft has also introduced “Copilot in Azure Operator Insights” to enable natural language interactions with Azure Operator Insights.
Azure Operator Nexus acts as a common underlying platform on which CSPs can not only host the technologies that deliver their current network services, but also build new revenue-generating applications.
A new operator RenAIssance
CSP modernization strategies are also being heavily influenced by the rapid acceleration of narrow and generative AI solutions.
Generative AI is starting to play a crucial role in telecommunications and is fueling the opportunity for a true industry renaissance. Indeed, many telcos are poster children for the early adoption of AI technologies. From enhancing network optimization and predictive maintenance to creating personalized customer experiences and intelligent virtual assistants, “GenAI” drives efficiency and innovation, in turn leading to reduced operational costs and increased revenue.
One example: AT&T has introduced a novel tool called Ask AT&T, powered by Azure OpenAI’s ChatGPT technology. Running in an AT&T-dedicated Azure tenant, AT&T employees can bring company data into Ask AT&T without worrying about the material leaking into the public domain. Early use cases currently being explored include using AI to help optimize the network, making care representatives more effective at supporting customers, helping coders become more productive and reducing employee meeting time through automated meeting summaries.
Likewise, Lumen Technologies is an early adopter of Microsoft 365 Copilot to empower their nearly 30,000 employees. Customer service teams at Lumen are using Copilot to surface relevant policies, summarize tickets and easily access repair instructions. Sales and customer experience teams are using Copilot to add depth and context to customer communications and summarize actions and next steps.
While AI-enabled applications can be run independently from the network, further integration of generative AI across telecom services and customer-specific large language model training will bring significant benefits to CSPs, as they seek to boost employee productivity, network automation and fault resolution, service security and the creation of new customer applications.
Monetizing the modern telco cloud
It’s hard to justify modernizing CSP networks without an accompanying path to new service creation and monetization. Operators are already looking for ways to monetize the significant investments made in 5G networks (that have not yet resulted in a corresponding uptick in revenue).
For operators to achieve these goals, their networks must become more “open for business” by providing programmable access for independent software vendors and app developers; something that hyperscale clouds have included since day one. Previous attempts to open operator network application programming interfaces (APIs) have failed in part due to the lack of coordination across operators, monolithic platforms not supporting external configuration and control, and a developer community that can create network applications in a familiar coding environment.
The GSMA’s Open Gateway initiative is a framework of network APIs designed to provide universal access to operator networks for developers. GSMA Open Gateway APIs are defined, developed and published in CAMARA, the open source project for developers to access enhanced network capabilities, driven by the Linux Foundation in collaboration with the GSMA. Working in CAMARA, APIs between telcos and developers can be delivered quickly, using developer-friendly tools and software code.
The Open Gateway initiative is currently supported by 39 mobile operator groups worldwide. Hyperscalers are also joining this API ecosystem and operating as aggregators. Microsoft has brought the GSMA Open Gateway initiative to Azure with its Azure Programmable Connectivity service and is already demonstrating compelling solutions with operators like Telefonica, DT and AT&T. Continued exposing of open APIs from the network will make it easier for developers to create modern connected applications that can operate over globally interconnected 5G networks.
Moving down the path
A modern, cloud-native network architecture is the key to unlocking many of the benefits that operators currently seek. By building on carrier-grade, hybrid-cloud platforms, CSPs will be better positioned to embrace best-in-class solutions at all levels of the technology stack, enjoy the advantages of scalable cloud resources and operational efficiencies, easily integrate cloud services and move more quickly down the path to complete network automation.
By taking a first-mover approach to embracing both generative AI and open APIs, CSPs will also be able to enhance current processes, improve employee efficiency and embrace the established developer communities that might finally help deliver the new service revenues that operators desire.