Innovation isn’t an option, it’s the expectation. That principle has kept us at the forefront of the technology for over 150 years. From the invention of the phone to the transistor and beyond, engineers at AT&T don’t follow trends; we set them. That’s the secret to serving millions at scale.
The development of next gen cellular networks is well under way, and once again we’re leading the charge with a few key partners. The key is Open Radio Access Networks (ORAN) which promise to boost speeds and reliability while driving down costs and making major strides in future proofing our networks. Not to mention increased opportunities for AI implementation and other emerging tech.
To better understand how ORAN will change telecommunications how we know, let’s first take a look at how networks used to operate.
The problem with how networks used to work
Traditional Radio Access Networks (RANs) are closed systems. The hardware and software come bundled from a single manufacturer: proprietary, integrated, and opaque. That coupling makes them stable, but it also makes them rigid. Upgrades are slow and expensive. New capabilities for AI automation, flexible capacity, custom security tooling require replacing entire stacks rather than adding a layer.
For a network operating at AT&T’s scale which are hundreds of millions of connected devices, tens of thousands of cell sites have that rigidity with real consequences. Every year of locked-in architecture is a year of delayed innovation.
What disaggregation actually means
The core idea behind Open RAN is disaggregation: separating the radio hardware from the software that controls it. Instead of buying a complete, sealed system from one vendor, AT&T engineers can now mix and match components across suppliers creating best-in-class radio units from one partner, virtualized baseband processing from another, AI optimization software from a third.
That might sound like a procurement change. It’s actually an engineering sea change. When the software stack is decoupled from the hardware, teams can:
- Swap out aging servers without replacing working software
- Deploy AI-driven optimization layers without hardware changes
- Run network functions in the cloud, applying the same principles as modern software infrastructure
- Upgrade individual components on their own timelines, rather than waiting for a full system refresh
The result is a network that operates more like a software platform than a piece of infrastructure - extensible, composable, and built to absorb what comes next.
AT&T’s Open RAN journey: Key milestones
AT&T didn’t arrive at Open RAN overnight. The shift began with a strategic decision in the early 2010s to virtualize the network and to move away from hardware-defined systems and toward software-defined ones. What followed was a decade of deliberate investment, partnership, and proof-of-concept work that is now producing operational results.
What this means for the engineers building it
Open RAN isn’t just a network story. It’s a story about what kind of engineering problems become possible when the constraints of closed systems are removed. We’re already seeing the benefits in scalability and efficiency as we expand implementations for our teams, our customers, and the industry at large. It’s already encouraging innovation, too, as more collaborators join the O-RAN Alliance focused on creating open, fully interoperable solutions to enhance user experience. More robust, customizable, and automated solutions are in development to take advantage of the power of AI. Wider choices in hardware and software, greater personalization in security tools, and Internet of Things (IoT) integrations are being worked on by our engineers right now.
This demo highlights how AI capabilities, powered by our next-generation Cloud RAN platform, can be deployed seamlessly to drive innovation and deliver superior customer experiences.
Rob Soni, Vice President, RAN Technology at AT&T