We’re entering a new year, and for telecom that raises the question of whether “open” initiatives might finally take hold. As we ended 2025, it was clear to most that Open RAN was not going to succeed, so it’s fair to ask whether the problems that it faced would confound other initiatives as well. Would they? Depends, perhaps, on why Open RAN didn’t, and so we have to start back almost a decade ago and unravel that answer.
Telcos, back then, were rattled by the combination of continued pressure on their profits, their return on infrastructure, and the consolidation in the telecom vendor space. With competitors leaving the market, many had visions of one of their old bugaboos, vendor lock-in and price gouging. While there were plenty of potential suppliers for packet gear used in mobile networks, the RAN was supported by a few giants, and it was this group that seemed to be consolidating.
So, Open RAN. Promote an open-source community solution that would attract a lot of players, and that would make the market highly competitive, promote innovation, and do all the things that competition and open strategies were supposed to do. Simple, right? Obviously not.
The big problem, some veteran telco types tell me, was that telcos really didn’t want an open solution. What they wanted was the threat to their comfortable vendor choices driving down pricing. There were a number of early trials and announcements that seemed to validate the strategy, but these veterans say that there was always a major barrier to adopting an open solution. First, it raised risk by introducing new and often small vendors. Even when a major vendor pushed it (like NEC) they were never able to gain significant market share. That might be due to the second piece—novelty. Telcos had bet on a small stable of vendors for decades, and so anyone recommending an alternative vendor faced an uphill battle to adoption, then a career risk if anything went wrong.
Then there was integration. None of the new players had a full solution to mobile infrastructure, and that meant that several might be needed. None had been integrated with packet equipment from that vendor community, so that was another integration novelty. And of course, of anything went wrong, out came the pointy fingers.
Greenfield mobile operators got some splash here for a bit (think EchoStar) but if you have a market where incumbent giants with a low internal rate of return are unhappy with profits, how good a market will that be for a new player? And, as the article cited above points out, Vodafone decided to use Nokia and Ericsson instead of an Open RAN player. There have been some successes, but it doesn’t look like there’s any momentum to the concept at this point, and it wouldn’t surprise me (or Wall Street, or telcos themselves) to see further retreat from Open RAN over time.
The big problem here is simple. Nobody can really expect to prosper in a market that’s already low-margin, with an open solution that by its very nature is commoditizing things further. Or at least not with half-hearted approaches to cost management that ignore any real opportunities to maximize profit.
From the first, my own view of 5G was that it didn’t go far enough in what I’ll call “mobile feature isolation”. The simple truth is that the only real difference in mobile networks is mobility. Thus, the goal of a mobile standard or architecture model should be to handle mobility as an overlay on standard packet technology. If we had location-independent routing at the IP level, integrating mobility management with it would be the right solution. Absent that, an overlay model the leveraged perhaps a new feature in routers in a standard way would be the answer. Virtual networking technology, including some of the SD-WAN implementations, illustrate that you can build an overlay network that at some point integrates with a baseline VPN. Couldn’t a similar thing have been done with mobility?
Then there’s the whole issue of the RAN Intelligent Controller as a kind of resource orchestration tool. It was a step, a potentially valuable step, toward hosted features and that concept of mobile feature isolation I mentioned above. But only a step. If you’re going to have RIC, and to host near- and non-real-time elements through its mediation, you need to think about expanding what those elements are so you could open up new revenue opportunities.
Open RAN should have been about feature hosting as its primary attribute. It should have embraced that goal not only in areas proximate to the RAN, meaning tower locations, but also through the entire metro area. It should have pointed out that metro-area-hosted features were the future of telecom services if you believed in anything beyond basic broadband Internet. It should have sought integration features with routers, cooperation with router vendors. It should even have encouraged a hosted-element view of things outside the RAN proper. It did none of this, and it’s not doing it now. Not that it matters, because it’s too late for any 5G-related strategies; the investment period is winding down.
And for 6G. Nothing could possibly be sillier than expecting 6G to revitalize Open RAN. Any attempt to do that would foreclose forever the chance of doing the right thing. There’s nothing more boring, less likely to be covered by the tech media, than a re-launch. Open RAN cannot be fixed now, it has to be totally replaced, and this time it has to get it right from the first.
But, just like 6G overall, getting an open strategy right is not an easy proposition, even though it’s more important for 6G than for 5G. Why? Because 6G can only succeed if it accepts that real-world, real-time, features are the only salvation to the challenge of getting acceptable ROIs. Because telco pressure is already focusing on making 6G more an evolution, which would almost surely confine it to the same profit sinkhole that sucked in 5G.
The only hope of Open RAN, or open-anything in a telecom sense, is to broaden the goal set. It’s not about RAN or even mobile infrastructure, but about the converged-infrastructure future. It’s not about connecting future applications, but componentizing them and hosting some of the components. Decompose the grand whole of this into a bunch of tiny pieces, and you don’t have a justification for openness, you have an infrastructure anarchy that only a giant vendor would have the credibility to integrate for telco buyers. Sound familiar? It should.
