3GPP openly standardized 3G, 4G, and 5G wireless. But the software and hardware used to implement the base station have historically been closed and proprietary, with all market participants repeating the same development investments. In the same way, before Linux, every serious computing company maintained its own operating system. The same work, repeated behind closed doors, by isolated teams. Linux provided a shared foundation and a new business model that allowed the industry to stop rebuilding the same thing and start competing at the application layer on top of it.
As Paul Sutton, SRS Founder and CEO, has noted: ‘In the same way the Linux kernel transformed computer operating systems, an open source RAN kernel can transform wireless.’ This is the thesis behind OCUDU: the RAN is ready for the shift to open collaborative development.
A problem worth solving
The radio access network (RAN) is an enormous software system, tightly coupled to specialized hardware and subject to demanding timing and reliability requirements. It includes schedulers, protocol stacks, interface handling, timing control, and real-time signal processing. Yet across vendors, much of this software performs the same fundamental job: implementing the 3GPP and O-RAN specifications reliably and securely.
Today, each RAN vendor builds and maintains its own implementation of this software. That means the industry repeatedly invests enormous amounts of engineering effort to solve essentially the same standards-compliance and systems-engineering problems, while operators remain dependent on a small set of vendors for the resulting products.
The result is a highly concentrated market for critical infrastructure. A relatively small number of companies supply much of the world’s RAN equipment and software, limiting operator choice and concentrating control over technology that networks and the people and businesses that depend on them cannot readily inspect or independently modify.
The requirement for this duplicated effort is also a barrier for innovation. If entering the market means years of foundational engineering before you can add that bit of innovation, the market stays narrow. For operators the limited choice can increase prices and slow roadmaps. For researchers it acts as a wall preventing new ideas from being developed and evaluated.
The specifications themselves are open. Anyone can read them. Open implementations have existed too, srsRAN among them. What has never existed is a genuinely shared and open foundation that is carrier-grade and held under neutral governance.
What Linux actually proved
The lesson from Linux is not that open source is cheaper. It is that an elevated common foundation enables innovation and changes where competition happens.
Once the operating system became shared infrastructure, nobody won by having a marginally better kernel. They won by what they built above it: applications, services, hardware, support. The foundation stopped being a moat and became a floor. And because thousands of organisations depended on the same floor, it was inspected, tested, and hardened more thoroughly than any single company could manage or afford to do alone.
The claim is smaller than it sounds. A shared foundation does not make products for anyone. It removes the need to rebuild the part everyone needs, and nobody differentiates on.
What OCUDU is
OCUDU is carrier-grade, open-source RAN software: the base station Centralised Unit and Distributed Unit of a 5G network, the software functions that sit between the radio and the core. It is hosted at the Linux Foundation, licensed permissively, and developed in the open at gitlab.com/ocudu.
It is designed as exactly the kind of common baseline described above: a foundation the ecosystem can build on, adapt, and extend. It is one of the most complex parts of a wireless network: handling control, data, and timing. A shared implementation of the parts of the RAN that every network needs, and because it’s fully defined through 3GPP standards, it has served as a barrier to entry but not a competitive platform.
OCUDU provides the RAN software but, in order to build a functioning wireless network, you will also need a radio with an antenna (that transmits and receives radio signals), a core network (that manages SIM cards and interfaces with the wireline network), and high-performance compute hardware (Intel, AMD, or ARM based embedded computer or server). You bring the radio unit, the core, and the integration. That is not a limitation of the model; it is the model. The CU and DU are logical breakpoints allowing the foundation to stay focused, and the pieces above and around it stay open for others to provide, which is where an ecosystem comes from.
How a shared baseline changes the work
For a vendor or integrator, OCUDU shifts where engineering time is spent. The baseline, the standards-compliant CU/DU that every network needs, already exists in the open, tested against O-RAN interfaces. Effort moves to the layers where products actually need to compete: radios, deployment tooling, management, optimisation, and applications.
For a researcher, the wall between idea and network gets thinner. The same codebase that runs in deployments is the one you can read, modify, and test against. A concept does not have to be reimplemented from a paper into a proprietary stack; it can be built where it will eventually run.
For an operator, transparency is the practical benefit. A foundation anyone can inspect is one whose behaviour can be checked directly.
And because the foundation is shared, it compounds. Every integration exercised against it, every bug found in one deployment, every hardening pass, accrues to everyone building on it. That is the property that made Linux what it is, and it only emerges when the baseline is genuinely common.
Built to evolve
OCUDU serving as the RAN foundation cannot stand still, because new needs, use cases, and applications require innovation. Open source software is built to evolve with the ecosystem: toward AI-native RAN, where machine learning runs inside the radio network rather than beside it, to integrate new algorithms and technologies like integrated sensing and communications (ISAC), and toward 6G, where much of the research that will define the standard needs a real, open platform to build on.
Research and innovation on OCUDU is unique because it’s repeatable and deployable. This opens the door to contributions from anyone, anywhere, and at any time. Proprietary software is no longer the barrier, with the direction defined by those who contribute. You can choose to watch it take shape, or shape it.
Where SRS sits
SRS is the founding lead maintainer of OCUDU: we built the RAN software, continue to lead the development work and run the Technical Steering Committee (TSC). Technical direction and major decisions go through the TSC and the project charter, in the open. As part of the transition from srsRAN to OCUDU, we are expanding the ecosystem and adding voting TSC members based on contributions. We will continue to fulfill our role as a steward of the project, ensuring its growth as a neutral and community-driven project.
See the vision
If the idea of a common RAN foundation is one your organisation has a stake in, whether you would build on it, deploy it, or research with it, the place to start is the project itself.
Read the OCUDU Community Vision and explore the documentation at docs.ocudu.org. The code, the roadmap, and the community are all in the open at gitlab.com/ocudu.
We are at the beginning of the journey, with a significant head start. Linux was built over time, led by visionaries with thousands of users and contributors who built on it. The RAN version of that story is just beginning, with srsRAN providing a solid O-RAN-certified foundation, and it is now open to anyone who wants a hand in using, deploying, and writing it.
