Standardized Architecture Key to 5G ORAN, Says AT&T CTO

“In the traditional approach, the RAN is built in a very modular way … split into components,” Fuetsch said. “There are a lot of details between the interfaces that connect these components together, and these are typically vendor proprietary.” An ORAN requires standardization for interoperability, he said. A VRAN is “really taking advantage of how you can actually run a lot of the software … in a much more general, multipurpose” way, he said.

ORAN and VRAN are “very complementary,” Fuetsch said: “We see a way here that we can drive much more innovation, much more competition into this new evolving ecosystem.” A properly configured ORAN “will support a wide range of deployment flavors,” he said: “It’s really important that the ORAN specifications support a broad set of capabilities and configurations.” Competing architectures could lead to “fragmentation” and “dilution,” he said.

In 3G, carriers had two standards, CDMA and GSM, and the alliance doesn’t want a similar dynamic now, Fuetsch said. “Our goal in ORAN is to make the specifications as inclusive, as capable, as interoperable as possible,” he said, noting the alliance includes more than 27 operators and 200 total members throughout the supply chain. “There’s strength and power in numbers,” he said: “We knew we couldn’t just do it alone.” AT&T is pleased with the results of its ORAN trials, but changing the architecture of networks “is not an overnight transformation,” he said.

“There’s a lot of work that has gone on and that’s involved in creating a good set of open RAN specifications” and work with other groups like the Small Cell Forum and Open Networking Forum, Fuetsch said. The alliance is trying to be compatible with the 3rd Generation Partnership Project “wherever possible,” he said. Where 3GPP doesn’t have specifications, the alliance is working with other groups, he said.

With ORAN, “telcos now can develop future-proof infrastructure,” said Alex Choi, Deutsche Telekom senior vice president-technology strategy and innovation. “There’s an important benefit of enabling a broader vendor partner pool.” Choi predicted two to three years for the first carrier-grade open networks to be deployed. An increasing number of developers are participating in the O-RAN Alliance and other forums, with more than 2,500 experts participating this year in the alliance, he said.

Many ORAN concepts come from the IT sector, said Colin Bryce, CommScope director-mobile network engineering. The difference between the sectors is “the level of robustness” required, he said. There’s a difference between a server in an air-conditioned data center and atop a cell tower, he said. “We’ll see a large number of vendors” offering software, but the real need is for hardware that can make ORAN work “out in the field,” he said: “The integration … is where the challenge comes.”

The focus mostly has been on making network components interoperable, but that’s not the most important or challenging part of open networks, said Oguz Sunay, Open Networking Foundation vice president-R&D. Bringing in programmability and RAN intelligent controllers will be “more challenging and interesting when we get there,” he said. Industry could have “control and configuration … app stores” with an “explosion” of developers offering apps, he said.

Big telecom players have been “notoriously bad” at supporting startups and smaller players, Sunay said. In venture capital circles, “there is now this common belief that telcos actually kill startups,” he said: “With open RAN, this is changing.” New wireless providers like Dish Network are “betting on new, innovative, smaller companies,” using them as vendors and funding them, he said: The big carriers are active on ORAN, and “the appetite is there and things are changing.”