Mobile carriers have a lot of money and fantasies riding on their ability to virtualize networks, and many outstanding challenges remain. 

The first wave of network virtualization involved the management and orchestration of virtual machines (VM), but operators only gained a few of the automation and agility benefits promised in the original concept, according to Grant Lenahan, partner and principal analyst at Appledore Research. 

“About 24 months ago they hit the reset button and began heading toward cloud native,” he said, adding that most communications service providers have been making this “painful pivot” to actualize more benefits of virtualization.

Technical and engineering complexities notwithstanding, virtualization, in an ironic twist, also requires operators to balance human involvement and adoption. 

“I don’t think the operators have ever taken on something as ambitious as this,” LightCounting consultant Roy Rubenstein told SDxCentral.

“Even when there’s periods of significant technological change … the hurdle in all this is humans can only accommodate so much change,” he said. Moreover, “there’s always a lag in terms of how it gets embraced by society.”

The vision of fully virtualized, automated, orchestrated, and disaggregated networks is relatively clear, but the collective undertaking could last two to three decades, according to Rubenstein. 

Mobile network operators are relying on a confluence of activities and developments in 5G, mobile edge computing, cloud-native technology, and telco clouds to achieve their virtualization goals. These largely disjointed efforts are gaining momentum while carriers shift from virtualizing core network functions to more interesting and complex efforts at the edge, Rubenstein explained. 

“Mobile edge computing and telco cloud factors heavily for needed scale and performance,” Will Townsend, senior analyst at Moor Insights & Strategy, told SDxCentral.

Indeed, as software permeates every network layer and element and mobile network infrastructure evolves to adopt webscale architecture, hyperscalers are poised and well positioned to become the most important vendors — and potential threats — for carriers at large. 

Public cloud providers like Amazon Web Services (AWS), Microsoft Azure, and Google Cloud are increasingly gobbling up telco-centric workloads and services. 

Cloud-native technology is, at the very least, a “hand me down from the public cloud industry,” Lenahan said. “Many public clouds are in fact capable of hosting high availability workloads, and the investment model of the webscalers and the shared user model gives them some real economic advantages.”

However, this too requires carriers to embrace many internal mindset changes, he explained. “Cloud demands that communications service providers shift from building reliable apps on unreliable platforms to building unreliable apps on inherently reliable platforms” that are actively managed and self healing, Lenahan said. 

“Engineers must unlearn decades of experience that frankly served them well,” he said.

As such, accepting new operational models is likely the biggest obstacle facing most operators' virtualization goals, according to Lenahan. “Cloud native demands that operations teams trust intelligent automation to make constant changes automatically. It demands being hands off and comfortable with uncertainty. This is new territory for most telco operational staffs brought up on tight service-level agreements (SLA) and hands-on control to ensure those SLAs are met.”

Operators may not readily admit it, but these technical and operational challenges face most carriers in equal measure. “No operator can really pull ahead on this compared to the rest,” Rubenstein said. “All sorts of technologies need to be ready, and it’s not that an operator can do this with a single vendor. It’s an industry effort.”