Verizon shares vision for transition to 5G, expectations to meet high-reliability, low-latency thresholds
Fueled by massive amounts of millimeter-wave spectrum, true mobile 5G services will deliver ultra-fast download speeds and two performance characteristics that promise to be crucial in the critical-communications arena: 99.999% reliability and end-to-end network latency of less than 5 milliseconds, according to Verizon officials.
Verizon CEO Hans Vestberg initially announced these performance thresholds last month during his keynote speech at the CES event in Las Vegas. The “five 9s” reliability and sub-5-milliseconds end-to-end latency were two of the eight “currencies” that Vestberg cited as performance metrics that should be met before providers describe a wireless broadband offering as a 5G service.
While most attention from Vestberg’s presentation focused on the promise of peak data rates of 10 GB/s, the high-reliability metric may be most significant feature of 5G for the critical-communications community, whether it is serving public safety, smart cities, smart grids, remote telemedicine or autonomous vehicle.
Mike Haberman, Verizon’s vice president of network engineering, said that 99.999% reliability—a threshold traditionally reserved for only the most robust public-safety LMR networks—is achievable with 5G, because of dense networking, complementary spectrum use and virtualized network elements housed in a distributed cloud platform.
“We have the most reliable network in the industry,” Haberman said during an interview with IWCE’s Urgent Communications, noting that Verizon has not published a reliability metric for its current 4G network “But I think [Vestberg’s] point was that it’s going to be taken to the next level with all of the new technology in our cloud platform, just because of its inherent design.”
At the heart of this design are software-based network elements that can be distributed in many locations throughout the network, Haberman said. By leveraging the power of mobile edge computing (MEC)—typically housed at key macro cell sites—the network becomes much more resilient, he said.
“It’s about putting the resource applications where they need to be—you can dynamically assign them around,” Haberman said. “If you have a problem at one C-RAN location, you can move the function to another. It might be slightly farther away, but these C-RAN locations are pretty close. It might change your latencies a little, but you’ll still have the ability to provide the functions.
“So, you’re insulating yourself to hardware failures. Essentially, hardware goes down, and you get to it whenever you can get to it. But it’s not urgent to do it, because [the network function] is moved around automatically by an orchestrator. So, it’s sort of a self-healing network that takes advantage of the compute and storage that it sees at all of these distributed locations.”
Haberman said this architecture is very different than the approach traditionally employed by commercial wireless providers, thanks to developments in cloud technology and processing power at the edges of networks.
“Pre-cloud, you had a primary and a secondary,” Haberman said. “You either shared the traffic at a 40% load—so, if one failed, the other could handle it—or you did a hot-standby configuration. This is a markedly different, because we’re moving to a cloud environment.
“The beauty of it is that we’re doing a true cloud environment. Some other carriers out there are calling it cloud, but they have a lot of separate platforms underneath it. We’re actually putting it all on one cloud platform—all of the elements are on one cloud platform that’s orchestrated.”
Introducing intelligence at the edges of the network not only improves reliability, it helps enable the low latencies that so important to many critical-communications applications, Haberman said. Air-interface efficiencies in 5G reduce latency significantly, but leveraging mobile edge computing is the other key feature that enables end-to-end latency of less than 5 milliseconds, he said.
“Now, you can take functions that are in our network equipment center and move them down to the edge of the network,” Haberman said. “So, you can get whatever the device needs really quickly now, because you’re not going to a centralized location somewhere in the middle of the country. Your control can be coming virtually from as close as the cell site, so that’s the other piece of it where you can get those ultra-low latencies.
“Now, obviously, it all depends on where your application is … For example, if you’re trying to do gaming, and your game is in an Amazon cloud or somewhere else, then you need to get to the Amazon cloud to run the game and get back to you, so those are the other things you’ll run into. But where you get mobile compute at the edge, it’s a totally different game at that point, because you can get rid of all of that back-end latency that normally would be there—it’s gone.”
One of the key differentiators between 5G and traditional cellular technologies is the spectrum. Historically, commercial wireless carriers have operated on licensed frequencies below 3 GHz. With 5G, carriers like Verizon will be lean heavily on millimeter-wave spectrum in the 28 GHz and 39 GHz bands. Verizon averages about 1,000 MHz of millimeter-wave spectrum in markets nationwide, giving the carrier the kind of bandwidth needed to deliver “crazy speeds,” according to Haberman.
Through the use of 5G beamforming and multi-input, multi-output (MIMO) technology, the millimeter-wave signals have proven to be more resilient than many industry experts had projected, he said. While some early industry estimates projected the need for millimeter-wave sites to be less than 100 years apart, carrier trials have reported ranges of more than 300 yards under the right conditions.
Still, operating at such high frequencies means that signals propagate from cell sites with limited range, driving the need for proliferation of small cells in the network—an architecture that also optimizes spectrum reuse throughout the network, Haberman said.
In addition, the use of higher-frequency spectrum allows physical components of the network to be much smaller and more flexible than is possible when operating on spectrum below 3 GHz, Haberman said.
“With beamforming, I have 120-degree sectors, and if I have 128 elements in there, and each element can tilt up and tilt down, and … each element could have four transmitters on it—you’ve got 4×4 MIMO in individual channel elements,” he said.
“The only way you can do that … is to be in millimeter wave. You can’t do this in lower frequencies, obviously, because the little cell would fall over, if you put this much technology on it with larger elements [needed to support operation on lower frequencies]. So, you have antenna technology that really is the game changer with 5G—that bandwidth. It’s also the design of it, because it’s designed for low latency.”
Verizon is offering 5G fixed-wireless offerings in several markets, but the carrier has not announced a timetable to mobile 5G.
“We haven’t given out an exact launch date,” Haberman said. “Clearly, we’ve been working on the implementation, because you’ve seen our 5G Home product—that’s the tee-up version. The underpinnings of the deployment are the same between NR [5G New Radio] and tee-up, because you need dark fiber, you need real estate, you need a core—which we have—and you need some equipment on, for lack of a better term, the pole.”
Haberman mentioned AT&T’s decision to market the deployment of equipment that AT&T officials have said is designed to enable a smooth transition to 5G via a software upgrade—a development that AT&T calls 5G Evolution, or 5G E.
“We’ve had that out for quite a long period of time, but we’re not renaming 4G LTE as 5G,” Haberman said.
Although 5G technology technically can be delivered in low-band spectrum, millimeter-wave spectrum—lots of it—is needed to make 5G deployment result in the kind of quantum-leap performance jump that the wireless industry has promised users for years.
“Let’s be very clear here: If you put this [5G technology] at low band—if I say, ‘I’m going to take 5G NR, and I’m going to put it at 700 MHz,’ or say I displace CDMA and put it at 850 MHz—and say I put up a 10×10 MHz carrier,” Haberman said. “If I did that, the system would be worse than 4G LTE. I wouldn’t be able to aggregate carriers together, so I’d be sitting there with an NR 10×10 MHz carrier, and 4G—given the area—could be a 50×50 carrier. So, I’d be sitting at a severe disadvantage on the 5G NR side.
“We could do that—our radios are capable, so we absolutely could—but the only thing that would really do is light up an icon on the phone, but it really wouldn’t change the customer experience. It might give me a slightly better latency—that’s possible—but, if you don’t have the bandwidth, what applications are you talking about?”
Verizon is not providing a timetable for its mobile 5G rollout, but Haberman offered some insights about how the transition period to 5G will look to users.
Perhaps the most important aspect of the 4G-5G transition—as opposed to previous generational wireless technology advances—is that both 4G and 5G are part of the same Long Term Evolution (LTE) standard approved by 3GPP. Most wireless experts acknowledge that LTE Release 15 as the first 5G standard. Because it is part of LTE, 5G can continue to leverage many of the network functions and equipment that are being used to deliver 4G services, which is expected to result in a smoother transition for users.
For instance, the first 5G phones also will support 4G connections simultaneously, allowing for seamless switching between the technologies, based on the network technology available in a given coverage area, Haberman said.
“It’s different than 3G to 4G,” he said. “We did have 3G and 4G in a device at the same time. The only difference is that you can be dual-connected between 4G and 5G; 3G and 4G were not dually connected.”
This feature also helps contribute to the promise of greater reliability, because connectivity is not dependent solely on one technology or spectrum band, Haberman said.
“You’re going to be able to take advantage of the different reliability characteristic of all the different locations,” Haberman said. “For instance, if you lose a 5G ultrawideband node—the high-bandwidth, high-frequency stuff—in most cases, you’re going to have either 4G LTE, or you’re going to have a 5G NR carrier at a lower band—that’s sort of the way the technology is going to roll out.
“In the next few years, you’re going to have 5G ultrawideband in the device, and you’re going to have 4G in the device. The way it’s configured is you’re going to be dual-connected. So, if something happens to that local 5G-ultrawideband node, you’re not going to have the ability to transmit over that node, but you’ll still have the 4G LTE carrier available to be able to transmit on. So, you’re going to get an additional layer of capability in the device.”