Op-Ed: Is open RAN the next NFV?

On the face of it, open RAN should be a really good idea. The set of open standards creates interoperability between the different components that make up 5G radio access networks (RAN) and virtualizes wireless services over a cloud core. This should deliver two benefits: making it easier for communications service providers to develop, implement and monetize new services, and allowing them to mix and match products from different vendors — creating competition and bringing down prices.

That’s the theory, anyhow.

In practice, the telecom industry has been on a similar mission before, and it didn’t go well. I’m talking about network functions virtualization (NFV), the industry’s effort to create a standardized way to run interoperable virtual services in the cloud.

Ten years ago, I stuck my oar into the NFV pond by launching a not-for-profit industry organization called the NIA (New IP Agency) which was designed to chivy the process of creating interoperability for NFV along. NIA didn’t work and neither did NFV.

Why NFV didn't work

There are a few reasons why. First, NFV was really complicated (unless you were involved in it you have no idea how complicated it was).

Second, despite some cool technology, it didn’t really solve the problem of helping service providers make more money from mobile services. That turned out to be a cultural issue more than a science one, and cultures are like oil tankers: they take time to turn.

But there was a third reason why NFV didn’t live up to its promise: the fastest growing and most influential new service providers in the world at the time – the cloud hyperscalers – didn’t really give a crap about standards, or interoperability, or making life easier for other service providers.

While the telecom industry busied itself over its NFV scrolls, companies like Google and Amazon Web Services (AWS) sat on the fence and basically ignored what they were doing. They preferred a Burger King approach (a.k.a. “Have it your way”) and focused on doing what they were really good at: being arrogant and making tons of money from super-sized cloud services.

This was a hugely significant moment in the history of telecommunications. For 150 years, standards were considered an inviolable, essential component of the telecom ecosystem. And then the hyperscalers came along and were just like, “OK boomers, have fun playing cribbage.”

A decade later and the most obvious effect of this shift has been to create a new industry segment, multi-cloud, with products from companies like Aviatrix that are designed to solve the problem, created by the hyperscalers, of getting different clouds to play nicely together.

Which brings us full circle to open RAN. To their credit, the 300+ members of the O-RAN Alliance have worked hard to overcome the complexity issue that helped scuttle the NFV effort. (It helps that they’re just addressing one bit of the network, rather than trying to boil the entire telecom ocean, like the original NFV efforts.)

Early implementations, led by Deutsch Telekom, while not without problems, haven’t been completely sh*t.

That doesn’t mean that open RAN is going to be a success. Actually, there’s a real risk that it turns out to be a solution in search of a problem.

The biggest challenge faced by wireless carriers today isn’t interoperability, it’s working out how to make more money from 5G. In principle, having 5G solutions from different vendors working together will bring the cost of building 5G networks down.

In practice, operators outside the regulatory Zone of Stupid (ZoS) in North America and the U.K. that care about low prices can just buy their networks from Huawei, which regularly undercuts its Northern European competitors by as much 30%. And Huawei doesn’t support open RAN.  

The biggest beneficiary of interoperability isn’t actually service providers but smaller 5G vendors, like Mavenir, which should finally have a chance to compete on their merits to be part of major 5G deployments alongside products from Ericsson and Nokia. But I’m skeptical about whether this peaceable kingdom of interoperable 5G will ever come to pass.

It’s much more likely that mobile operators will continue to do what they have always done, what they are culturally attuned over a century and a half to doing, which is to prioritize finding one company that they believe can do everything they need to help them grow their currently underwhelming revenues from 5G network investment.

What's the alternative?

The alternative — creating a patchwork network out of a cool but slightly wonky new multi-vendor approach — is higher risk and voids the “one throat to choke” supplier accountability that mobile network operators feel comfy with.

As we roll into 2024 and 2025 expect to see the noise about open RAN increase, and then die away, just as it did with NFV. It will be replaced by a storm of activity around two new areas: One is modernization (tightly integrating business applications with cloud infrastructure, something Microsoft is laser focused on currently).

The other is the 5G-Advanced upgrade, also known as 5.5G, which will arrive with the 3GPP Release-18 update in March 2024. Crucially, this standalone upgrade, which works with a cloud-native core, will improve the uplink speed and stability of the standard, which will be key for enabling all kinds of high revenue industrial applications, from high-quality video surveillance, to upgraded computer vision automation and wireless robots on the factory floor.

Expect a blitzkrieg of coverage from Silverlinings on both areas. Open RAN? Not so much.


Op-eds from industry experts, analysts or our editorial staff are opinion pieces that do not represent the opinions of Fierce Network.