How 5G and Open RAN are changing the economics of wireless
When the U.S. military drops into a hot zone, one of the first orders of business is to set up a communications network. These networks look and work a lot like a traditional wireless network, including a powerful antenna, switching hardware, and other infrastructure, all designed to be used privately and securely. Until recently, once any intelligence was captured, it would have to be beamed back to the U.S. for processing, then returned to the field with analysis attached. And that's an awfully slow process. By the time all that information was shipped halfway around the world—twice—it may have become outdated and the potential target may have moved on.
A new set of policies and standards collectively known as Open RAN is now changing all of that. Rather than having to send data elsewhere for processing, the U.S. Department of Defense—and anyone, really—can swap out formerly proprietary hardware that was part of the telco network with their own equipment. The upshot is that the DoD can now install its own servers and process data on the fly and at the edge, wherever its forces happen to be, reducing latency, increasing reliability, and—the kicker—doing all of this for a lot less money.
The RAN is the edge of the telecom network, consisting of the cell tower you see on the side of the highway as well as the computing devices on it and connected directly to it. All device access to the network, both cell phones and the Internet of Things, goes through the RAN.
RAN hardware has been extremely expensive and available only from a small number of vendors. Because 5G networks use radio frequencies that require a large increase in the number of cells and therefore the number of RANs, a new set of policies and standards collectively known as Open RAN has emerged to standardize the design. The result will be lower costs and a burst of innovation.
Open RAN is good for carriers
Private network applications, like the military field comm network, are exciting, but most of the motivation behind Open RAN comes from the mobile carriers.
Historically, says Guy Daniels, director of content for TelecomTV, the RAN has been "impenetrable and proprietary," a tightly guarded set of technologies unique to each equipment vendor. Vendors did this on purpose, of course, to keep customers handcuffed to their hardware and, as a result, to keep prices high. If one vendor's processing unit installed in a RAN in the field failed, it would have to be replaced with another (costly) same vendor device. Even replacing one vendor's equipment wholesale with another was difficult, because your choices were so limited. "It's basically Nokia, Ericsson, and Huawei," says Daniels. "That's your selection."
For years, wireless operators haven't been thrilled about any of this, but with the rise of 5G, the situation has become untenable. The faster performance, powerful features, and better coverage of 5G don't come by simply flipping a switch. These advances require additional hardware, and lots of it, much of it in the form of RAN systems.
"A typical operator like Verizon or AT&T has about 65,000 to 70,000 cell sites," says Geetha Ram, global leader of 5G telco edge solution strategy for Hewlett Packard Enterprise. With 5G, carriers have begun to add smaller cells like picocells and femtocells to their network, literally by the truckload. When all is said and done, she says, "the 5G densification could see the number of cell sites expanding five to seven times," to somewhere on the order of 400,000 locations that have to be installed and actively managed.
This edge of the network is not a high-tech IT environment but rather a gritty, industrial one, adds Ram. Cell sites can be found anywhere—on the tops of buildings, strapped to a tower, or even, as Ram notes, hidden in a cave in India. Servicing these locations is difficult, and the people who are tasked with doing so are rarely sufficiently trained in the intricacies of RAN troubleshooting. Ram muses that one frontline technician with whom she once personally interacted had been a pizza delivery driver only a week earlier.
Further complicating matters is the fact that operators have started realizing that with IoT-centric use cases and device proliferation, more processing needs to happen at the edge of the network, not on servers in the network core. Those applications cannot run on proprietary hardware but require more traditional computing technology, like the servers the DoD would need in the example above.
The solution to all of this: Open up the RAN and create specifications that allow for hardware and software to be used freely and interchangeably, regardless of who designed it.
Please listen to: 5G: What it's good for―and what it won't fix
"With Open RAN, you're not locked into a vendor," says Ram. In fact, you're not locked into using a vendor at all. "You can even upgrade the system with different software. Or, if certain cell sites are no longer at capacity, redeploy them to do other things." For example, a remote cell site that's only being used at 30 percent of its capacity could be programmed to use the excess capacity for a variety of unrelated edge computing needs, reducing strain on the core network and saving bandwidth.
Costs will decline (but not your phone bill)
So, what are we going to get out of Open RAN as it evolves?
"It all boils down to money," says Daniels. "Operators are faced with ever-tightening investment criteria and restricted spending power, and they look at the radio access network and it's huge. All those endpoints are just hugely expensive. So with the dawn of 5G and the realization that we have to densify networks and operators have to roll out even more cell sites, the economics just don't stack up." In other words, opening up the RAN means more competition among equipment providers and, by extension, lower capital expenditures for operators.
That said, Open RAN doesn't really hold the promise of any meaningful cost savings for the consumers and businesses who subscribe to wireless services. Carriers pressed with questions about passing savings on to their customers are likely to note that any savings on a particular piece of hardware will be more than outweighed by the sheer quantity of new equipment they're deploying.
Please read: The latest on 5G: What you need to know
"5G isn't really about the end consumer. It's about the enterprise," says Ram. "Enterprises want automated factories, robots running through stores, everything running at the edge. And they want to do all of this locally, with nothing having to run back to the core network, with no latency."
All of that said, we're still early in the process. Open RAN's main trade group and standards body, the O-RAN Alliance, was just established in 2018, and the bulk of its technical specifications are still being refined, with its most recent versions published at the end of 2020.
Daniels says a lot of this will shake out and normalize over the next six months, but he expects to see big things. "It does appear at times that Open RAN came out of nowhere," he says. "While it's come out the starting block incredibly fast, it's been mulled for a long time."
Lessons for leaders
- The RAN is the edge of mobile networks and critical to their performance.
- Telcos have huge investments in their RANs and will have to multiply those investments to make 5G work.
- Just as in other areas of tech, open standards are allowing the RAN to advance where proprietary vendor lock-in held it back.
A typical operator like Verizon or AT&T has about 65,000 to 70,000 cell sites. … 5G densification could see the number of cell sites expanding five to seven times.
This article/content was written by the individual writer identified and does not necessarily reflect the view of Hewlett Packard Enterprise Company.