AI + Machine Learning, Announcements
Important observations from Microsoft at Mobile World Congress 2023
By Yousef Khalidi Corporate Vice President, Azure Networking
8 min read
Mobile World Congress (MWC) was back in full swing this year, and so was Microsoft. To start, a recap of our MWC 2023 announcements includes 12 major product announcements made at MWC 2023 by Microsoft, with more than 20 demonstrations to highlight our latest developments. These demonstrations covered an array of topics such as live private and public multi-access edge compute (MEC) use cases, network API efforts with our operator partners, and the use of AI and analytics to improve operator efficiency and resiliency, among others.
Microsoft at Mobile World Congress 2023
As we often do, we have collected observations from hundreds of customer meetings (up nearly 35 percent from last year) and dozens of partner engagements at our demo booth, and I would like to share some of the important feedback that we heard while at MWC 2023:
- Operators are deeply interested in the monetization opportunities created by programmable networks and modern connected applications.
- Operators have a demand to see a strong total cost ownership (TCO) case when adopting cloud-native technology.
- There is ongoing, keen interest in lessons learned from real deployments at scale versus proof-of-concept deployments.
- Operators want to understand the impact that developments in AI can have on their ability to run networks that are more resilient, sustainable, and efficient.
- Operators are looking for an ecosystem that includes meaningful partnerships—both technical and commercial—in order to ensure holistic success.
- Operators are accelerating their plans for virtual RAN deployments.
In this blog post, we take a deeper look at each of these core pieces of feedback, providing additional insights, news, and links to further reading.
Creating value with network APIs and mobile edge computing
On top of the valuable opportunity created by the convergence of cloud computing and networking in a distributed fabric that spans 5G to space, Microsoft further believes that operators are uniquely positioned to create value as they modernize their networks to cloud-native technology.
The network itself is a rich source of intelligence and functionality that can be exposed—with the appropriate privacy controls and security—not just to end users but also more broadly to the developer community. This exposure will allow operators to find new sources of innovation in the developer community, and new approaches to support the goal of delivering value above and beyond pure connectivity. Network APIs will power a new generation of network-aware applications that put the flexibility and power of 5G to work, solving mission-critical business needs. Through our discussion of this exciting new development with our customers, a few key items emerged:
- Consistent implementation across operators is needed, as developers are unlikely to build apps that run on a single network.
- However, some localization will be required to ensure specific regulations on data residency, privacy, and controls are supported.
- Tying into existing developer communities is essential to accelerate the adoption of these capabilities.
- Developers require agility; therefore, we must trial new services in advance to create meaningful input for the standardization process.
- There is an outsized need to establish an appropriate business model that supports the investment that operators are making in these capabilities.
- Recognition that business models are likely to evolve rapidly as the technology trials with developers become more concrete.
- Not all network APIs are created equal, and the business model should be flexible enough to permit different monetization schemes.
To find out more about Microsoft’s leadership in the development of network APIs, read about Azure Programmable Connectivity.
Microsoft continues to work closely with our operator partners to deploy both private and public MEC solutions. A consistent theme that we hear from our operator partners is the need for a robust ecosystem of application solutions that highlight the value of 5G and edge computing. Several conversations with operators and ISV partners highlighted the progress that is being made in the deployment of solutions across the manufacturing, energy, and transport industries. Further, we understand the value of and need to move beyond individual proof of concept trials to deeper solution catalogs that cover multiple business processes enabled by a private network deployment.
We invite application ISVs to collaborate with us by joining the private MEC and public MEC programs.
Updating network architecture while reducing the total cost of ownership
As operators have built more sophisticated service implementations on top of traditional networking solutions, they have moved away from physical appliances to adopt disaggregated network functions based on standardized hardware deployments. While this has yielded some CAPEX savings, the operational complexity of these do-it-yourself (DIY) virtualization efforts remains high. This complexity directly drives up cost and risk. As a result, our customers have challenged us to show definitive benefits when adopting infrastructure based on cloud provider technology as they look to upgrade their network architecture to support the increased resiliency needs and new services, and to undergo the upgrade to a 5G stand alone (SA) core. As part of these discussions, operators have asked us to address:
- Demonstrating specific savings from the use of automation to accelerate deployment of capacity and to improve release quality.
- Improving standardization and consistency of network function deployments on top of a cloud platform.
- Eliminating the need to deal with managing hardware procurement and deployment.
- Moving from a traditional CAPEX model to OPEX-based consumption.
- Enabling the acceleration of and improvements to the quality of new service deployments within the network.
To learn more about the cost-effective deployment of 5G networks, read the Analysys Mason TCO report for Microsoft Azure Operator Nexus, February 2023.
Lessons learned from supporting a production deployment
While many operators recognize the potential of cloud technology to accelerate innovation and reduce costs, they are keen to understand the experiences of those early leaders who have already deployed similar cloud-based solutions at scale. For this reason, Microsoft invited our flagship customer, AT&T, to share their views on stage at MWC 2023. Igal Elbaz, Senior Vice President, Network Chief Technology Officer, and Rob Soni, Vice President of RAN Technology, AT&T Services, joined Microsoft to provide direct insight into their experience as AT&T continues to deploy Azure Operator Nexus to support AT&T’s mobility network.
Watch this video to hear about AT&T’s journey in their own words.
While many of the lessons learned from our deployment at AT&T have already been built into the Azure Operator Nexus platform, these lessons are also reflected in our blueprints for onboarding new network functions to the platform—our best practices for deployment and operations, our operating model, as well as our platform API design. To learn more about the Microsoft carrier-grade, hybrid platform, read the blog “Introducing Azure Operator Nexus.”
To learn more about how other customers are using Azure technology to test 5G technology, please see “MTN deploys one of the first 5G Standalone Core in Microsoft Azure.”
Harnessing AI-powered operations with new services
In light of the increasing complexity of disaggregated network architectures, larger numbers of devices connecting to networks, and a desire for rapid innovation, our customers continue to express interest in harnessing the power of cloud-based analytics and AI. To address this interest, Microsoft announced the availability of two new services—Azure Operator Insights and Azure Operator Service Manager—which were developed to simplify network management. A key part of our platform value proposition is that many of the capabilities we are building on with these new offerings are the same capabilities that power the management of Azure itself—so they have truly been tried and tested.
When we spoke to customers about Azure Operator Insights at MWC 2023, the following themes surfaced:
- On-premises or cloud-based data lakes have often been tried in one form or another, with mixed results. Data silos still exist as do the challenges of getting a systematic understanding of network health and customers’ quality of experiences.
- Customers were excited to see Azure Operator insights paving the way to remove the silos and enable data democratization for all users.
- GPT was of course at the front of everyone’s minds, and customers were also very excited to see the different ways in which Azure Operator Insights managed data would benefit from GPT—both in the short and longer terms.
Many customers also asked us how to automate network actions based on the insights AI can deliver. This is where Azure Operator Service Manager plays a role. Several key themes were raised consistently across many discussions on automation:
- Manual operating procedures are still prevalent within operators’ environments. These tend to be error-prone, costly, and typically delay the deployment of software or configuration changes.
- Where customers have tried to automate these procedures, the plethora of automation solutions for each underlying platform or network function vendor has resulted in fragmented tooling that fails to address the overarching service as a whole.
- Customers were interested to see how Azure Operator Service Manager enables a service-centric automation toolchain, addressing network services composed of multiple network functions and deployed across many sites and heterogeneous infrastructures. Customers were also particularly keen to learn how we have been able to reduce real-life deployments from days to minutes.
We also heard operators talk about their leading-edge experiences when adopting AI for fault management, customer service, and automation. However, it was the fact that Microsoft is ensuring AIOps is an integral part of all of Microsoft Azure for Operators offerings and the consistency in management that this guarantees that really got operators excited—as well as GPT integration.
For more information, visit Azure Operator Insights and Azure Operator Service Manager.
Establishing new partnerships and offering meaningful ecosystem support
Microsoft is proud of our unique acquisition of industry-leading, cloud-native network functions that provide us with the telco DNA needed to truly understand the unique requirements of carrier-grade solutions. We are equally committed to supporting the partners that operators know and trust today, with full and equal access to the Azure Operator Nexus platform capabilities. Our customers reinforced the expectation that we continue to work closely with the industry ecosystem in areas such as network function pre-certification, software DevOps, and security best practices to enable the successful delivery of the end-to-end service experience.
Watch our partners describe Microsoft’s efforts to jointly service the industry—including Monica Zethzon, Vice President and Head of Solution Area Core Networks, Ericsson, and Fran Heeran, Senior Vice President and General Manager of Core Networks, Nokia.
Register here to join the Azure Operator Nexus Ready program.
Adopting cloud technology to support virtual RAN workloads
As mobile operators begin planning for the next RAN upgrade after their current 5G new radio (NR) deployment, they are looking to better understand key questions such as:
- Will hybrid cloud platforms support various proposed acceleration technologies for both private and macro networks if the operators choose to adopt them? This includes the TCO and performance expectations associated with the use of disaggregated platforms to support RAN workloads, with a particular emphasis on energy efficiency and spectral efficiency in dense areas.
- Will cloud platforms provide the management of far-edge at scale, and enable fully cloud-native RAN workloads that are implemented on the proposed Open RAN interface specifications, such as the O2 interface?
- Can platform services be used to gain end-to-end visibility from applications, packet cores, RAN, and infrastructure? And further, can this be leveraged to gain true visibility into the network and to provide fully optimized and automatized CI/CD and AIOps experiences?
To better understand Microsoft’s vision for the adoption of cloud technology in support of RAN workloads, check out the Microsoft programmable RAN platform whitepaper.
All in all, MWC 2023 provided a fantastic and rich opportunity for Microsoft to connect with our customers and partners as we expand the use of our Azure for Operators portfolio to modernize and monetize the network.
Discover more about Azure for Operators
Learn more about the Azure for Operators portfolio.