From virtualization to automation–the march to 5G service revenues
Rakuten’s virtual RAN and Elisa’s lights-out NOC
The flexibility 5G enables requires a similar level of flexibility in the underlying network infrastructure. Proprietary, single-purpose hardware is giving way to general-purpose hardware running virtualized network functions. This allows operators to lower capital and operational expenses while also gaining dynamacy in capacity provisioning, spectral resourcing and service management.
As these networks become virtualized, with functionality moving into the cloud, the way they are operated is also changing. The sheer complexity of a distributed architecture processing a huge volume of data produced from myriad sources requires automation. “The automation piece is extremely important particularly as they scale up services,” Kevin Shatzkamer, VP/GM of Service Provider Solutions, Dell Technologies, said. He continued: “First we see cloud automation systems replace proprietary stacks, including the introduction of more DevOps/ NetOps tooling, powered by human intervention and decision-making. Over time, we will see the introduction of AI and ML technologies, automating the decision logic itself.”
So what might this process look like in practice? Finnish operator Elisa was an early-mover in 5G, launching limited commercial service using its 3.5 GHz spectrum in 2018, prior to commercial device availability. Last year the company’s CTO said its subscribers use around 25 GB of mobile data per month and 172 GB of fixed broadband data per month, among the highest usage levels in the world. To keep up with this increasing capacity demand, Elisa developed network automation tools to streamline its operations.
“We have developed automation capabilities which enables us to do this and succeed,” CTO and VP of Technology and Architecture Kalle Lehtinen said. “We have built capabilities in network management processes.” For instance, he said needs-based analytics are used to inform network capex strategy and he described the operator’s network operations center as “zero person. For years now we haven’t had a single person in our network operations center.”
Another interesting exploration of virtualization at scale comes from Japanese operator, and new market entrant, Rakuten Mobile. A subsidiary of the e-commerce giant, Rakuten Mobile CTO Tareq Amin has overseen the greenfield build of a fully-virtualized network comprising around 4,000 5G-ready cell sites, multi-access edge computing data centers and the core network. Amin, speaking during a press event in February, said it was an intentional choice to take a new approach rather than undertake a more conventional legacy-type build.
“Telco networks of today are very complex no matter what the Gs are. It has no software-centricity; it’s all about hardware migration as you go from one generation to the other. If you look at our architecture, our architecture today is truly the world’s first open RAN deployment today across any telco. It is running at scale. It is absolutely real; it is not pie in the sky.” He said the network demonstrates the ability to lower capex by 40% and opex by 30%.
The network modernization piece is just one transformation needed to realize 5G service revenues. To learn about the key role of edge computing and innovative service creation, read this article. Read about how OSS/BSS transformation is another necessity in the 5G era here.
For the big picture on IT/OT converge in the 5G era and how operators can position themselves to capture the 5G enterprise revenue opportunity, download this report.
The post From virtualization to automation–the march to 5G service revenues appeared first on RCR Wireless News.