Latest / The 5G Podcast with Fexingo: Wireless Networks, Carriers, and Mobile Infrastructure / Why Open RAN Is Still a Long Bet for Mobile Operators
Transcript
- Lucas: So there's this idea that's been floating around telecom circles for the better part of a decade now — Open RAN. The promise is simple: break open the proprietary black box that Nokia and Ericsson sell to mobile operators, let carriers mix and match hardware from different vendors, and slash costs by up to forty percent. It sounds almost too good to be true. And so far, it mostly has been. Luna: Right, because we've been hearing about Open RAN for years, but when I look at actual deployments, it's still a fraction of global 5G networks. What's the holdup? Lucas: Great question. Let's start with the poster child for Open RAN's ambition — Rakuten Mobile in Japan. Back in 2019, Rakuten announced it would build a fully virtualized, Open ran based 4G and 5G network from scratch. They called it the Symphony project. The idea was to source radio units from a bunch of different vendors, run the baseband software on standard servers, and prove that you didn't need Nokia or Ericsson to build a world-class network. Luna: And they actually did it. I remember the early coverage — they launched 4G in 2020 and 5G in 2021. But then things went quiet. Lucas: Yeah, they launched. But the network had serious teething problems. Coverage gaps, performance issues, and the costs ended up way higher than projected. Rakuten spent over six billion dollars building that network, and their customer acquisition targets fell short. In 2023, they basically admitted defeat on the pure Open RAN approach and started partnering with traditional vendors. Luna: So the supposed cost savings never materialized. But was that a failure of the technology itself, or just the execution? Lucas: I think it's a bit of both. The technology has genuinely improved — the Open RAN standards from the o ran Alliance are more mature now, and there are dozens of vendors offering compliant gear. But the integration challenge is enormous. When you have a radio unit from vendor A, a server from vendor B, and software from vendor C, getting them all to work together at carrier-grade performance is brutally hard. And carriers have an incredibly low tolerance for downtime. Luna: So it's not just plug and play. And that means the total cost of ownership might be higher than just buying an integrated system from Nokia. Lucas: Exactly. A report from Dell'Oro Group last year estimated that Open RAN still accounted for less than ten percent of total radio access network spending in 2025. The incumbents — Nokia and Ericsson — have actually been gaining share, not losing it. They've responded to the Open RAN threat by making their own products more open and flexible, undercutting the competitive argument. Luna: That's a really interesting dynamic — the threat of disruption actually made the incumbents better. But there are some places where Open RAN is working, right? I've heard about smaller operators in Japan and parts of Europe. Lucas: Absolutely. The most successful Open RAN deployments so far have been with smaller operators, greenfield networks, or in rural areas where coverage is thin. For example, a company called 1&1 in Germany built a greenfield Open RAN network using Mavenir's software and Rakuten's Symphony platform. They're a tiny player — less than a million subscribers — but they've shown it can work in a controlled environment. Luna: And in the U.S., Dish Network tried a similar approach with its Boost Mobile network. That's had its own challenges, but they're still going. Lucas: Right. But even Dish, which is now called EchoStar, has scaled back its ambitions. They originally promised to cover seventy percent of the U.S. population by 2023, and they're nowhere near that. The point is, Open RAN seems to work best when you're building from scratch with a small number of locations and a captive subscriber base. For a major operator like Verizon or Deutsche Telekom, ripping out existing gear and replacing it with an unproven multi-vendor system is a massive risk. Luna: So what would need to happen for Open RAN to really take off? Is it a matter of time, or is there a fundamental technical hurdle? Lucas: I think there are two key things. First, the chipset ecosystem needs to mature. Most of the Open RAN baseband processing relies on general-purpose CPUs from Intel or AMD, or on custom chips from companies like Marvell. These are improving fast — Marvell's Octeon processors are now used in several commercial deployments — but they're still not as power-efficient or performant as the custom ASICs that Nokia and Ericsson use in their proprietary gear. Luna: So the hardware isn't quite there yet for massive scale. But also, there's a software integration challenge — making all these pieces talk to each other reliably. Lucas: Exactly. The o ran Alliance has defined standard interfaces, but they're not always implemented consistently. And when something breaks in a live network, you need to know which vendor's component is responsible. That creates a coordination headache that operators hate. Some have argued that Open RAN actually increases operational complexity rather than reducing it. Luna: So it's a chicken and egg problem. Operators won't adopt it at scale until it's proven, but it can't be proven at scale without operators adopting it. Lucas: That's exactly it. And I think that's why the timeline has kept slipping. A few years ago, people were saying Open RAN would hit mainstream adoption by 2024 or 2025. Now most analysts are pushing that to 2028 or 2030. Vodafone, which was one of the biggest proponents, recently scaled back its Open RAN deployment targets in Europe, citing integration difficulties. Luna: I want to push back a little, though — isn't there a geopolitical angle here? Governments in the U.S. and Europe have been pushing Open RAN as a way to reduce dependence on Chinese vendors like Huawei, and also to create a more competitive vendor ecosystem. Lucas: That's a really important point. The political push has been significant. The U.S. government has poured billions into Open RAN research and incentives through the CHIPS Act and the Public Wireless Supply Chain Innovation Fund. The idea is to create viable alternatives to Huawei and ZTE, especially for smaller operators in rural areas. And there has been some success there — operators like UScellular and Verizon have done small Open RAN trials in rural markets. Luna: But those are trials, not full-scale network swaps. And the political pressure doesn't make the technology work if it's not ready. Lucas: Right. And the incumbents have been very effective at making that argument. Nokia and Ericsson both offer products with open interfaces now, so they can say 'you can get openness without the risk of multi-vendor integration.' That's a compelling pitch. I think the most realistic future for Open RAN is not a wholesale replacement of existing networks, but a gradual adoption in specific segments — small cells, indoor coverage, rural deployments, and new spectrum bands. Luna: So the narrative has shifted from 'Open RAN will disrupt everything' to 'Open RAN will be another tool in the toolbox.' That's less exciting, but probably more accurate. Lucas: Exactly. And I think that's healthy. The hype cycle peaked around 2021 when everyone was talking about Rakuten and Dish. Now we're in the trough of disillusionment, which is actually where the real work gets done. Companies like Mavenir, Parallel Wireless, and Altiostar are still iterating on their software. Incumbents are improving their own open interfaces. And the chip companies are building better silicon. Luna: Speaking of Mavenir — they went public in 2023, right? How are they doing? Lucas: Mavenir listed on the NASDAQ in 2023 via a SPAC merger, and it's been a tough ride. The stock has fallen significantly as revenue growth has been slower than expected. They reported about five hundred million dollars in revenue in 2025, which is up year-over-year, but they're still not profitable. The market is basically pricing in their promise rather than their current performance. Luna: So investors are skeptical. And that makes it harder for these companies to raise capital to keep developing. It's a tough cycle. Lucas: It is. But it's not all doom and gloom. There's a really interesting development in Japan — a consortium of operators including NTT Docomo and KDDI have launched what they call a 'shared Open RAN' platform for rural areas. They pool spectrum and infrastructure, and use Open RAN gear to keep costs low. That kind of collaborative model might be where Open RAN finds its sweet spot. Luna: So instead of each operator building their own network, they share the access network and compete on services. That's a fundamentally different business model. Lucas: Exactly. And Open RAN makes that easier because you can have multiple virtual networks running on the same physical infrastructure. I think that's the unheralded story — Open RAN's biggest impact might not be cost savings for a single operator, but enabling new types of network sharing and competition. Luna: That's a more nuanced take. So to wrap up, where do you see Open RAN in five years — say, 2031? Lucas: I think it will account for maybe twenty to twenty-five percent of new RAN deployments, concentrated in greenfield builds, rural coverage, and specific enterprise use cases like private 5G networks. The incumbents will still dominate the big, dense urban networks. But the ecosystem will be more diverse, and the threat of Open RAN will have forced Nokia and Ericsson to keep innovating. That's a win for operators and ultimately for consumers. Luna: So the hype was premature, but the direction is still right. Open RAN isn't dead — it's just growing up. Lucas: Exactly right. And as with any infrastructure technology, the real test is time. We'll see how the next few years play out. For now, it's a long bet that still has plenty of potential — if the industry can get the integration right.