5G is still working to deliver on the technology’s promise, and while early efforts to shape 6G have already begun it won’t mean a sudden goodbye to certain network foundations carriers are investing in and laying the groundwork for today.
Building and leveraging 5G infrastructure in a 6G world is one of several topics that will be discussed during the Fierce Wireless 5G Blitz Week, a free virtual event taking place March 28-31.
The typical technology cycle for a new G to be introduced is 10 years, with 2030 cited as a general timeframe for when 6G might enter the picture. And while that represents the start of standards setting, it doesn’t mean all of the work put into 5G simply falls by the wayside.
Prakash Sangam, founder and principal of Tantra Analyst, pointed out that while people think the next step of 5G is 6G – they’re missing a major evolutionary step in between – 5G Advanced. In December the 3GPP set out key projects to be included in Release 18, which could help underpin future systems, and 5G Advanced could possibly cover up to Release 20.
3GPP just finalized the third 5G NR spec, Release 17, marking the conclusion of the first phase of the 5G technology evolution.
Still, 6G, which is just starting to be explored and far from being fully defined, will make its way here eventually. And countries and groups are working to get a head start, such as ATIS’ Next G Alliance in North America and Europe’s Hexa-X project, among others. And carriers will want to continue to leverage investments that help make networks more future proof, across spectrum and infrastructure.
Fortunately, cellular standardization is an iterative process, Sangam said, noting that hardly anything new is introduced, deployed and commercialized in a major way in one release. Currently, commercial networks and devices are still at Release 15 and Release 16, he noted.
So what are some of the current 5G efforts that could likely make their way into 6G?
Monisha Ghosh, professor of electrical engineering at University of Notre Dame who has previously served as CTO for the FCC, pointed to network densification, fiber transport, and lessons learned from spectrum sharing, higher band spectrum like millimeter wave and antenna technology like massive MIMO as 5G work that will carry over to the next gen.
“I truly believe there will be a lot of time and investments made in building up a 5G infrastructure,” Ghosh said. For example, densification in deploying more sites and small cells, and getting fiber out to more places than it is available today. “And all of this infrastructure should be definitely used for 6G as well, and I hope it will.”
Virtualization, AI/ML
While it’s hard to say at this point what technologies will define 6G, and different groups have their own visions, some are thinking along similar lines, such as cloud-native, AI/ML, and energy efficiency aspects.
Ghosh suggested 6G will be what 5G doesn’t deliver on, while Sangam said new path-breaking concepts introduced in 5G Advanced will probably be what becomes mainstay for 6G.
Part of that ties into AI and machine learning, which 3GPP is looking into for 5G Advanced, both for the air interface and radio optimizations.
As for what part of 5G infrastructure can be used for 6G, Sangam said in parallel to the air link evolution helped by AI/ML, the most interesting thing in network architecture evolution is using disaggregated, virtualized, cloud-native software defined networks.
The move to more software driven networks could mean it’s easier to innovate while still leveraging network hardware, with architectures starting to evolve today serving as a building block.
Ghosh noted that, part of the open RAN concept is disaggregating the network so innovation can happen along pieces of the network instead of the whole thing.
“It allows you to access the network at different points, which allow you to do machine learning because now you can use data more efficiently from the software aspect,” she said.
Open RAN is easier to start from scratch, as is seen with Dish Network in the U.S., which is building out a greenfield open RAN-based standalone 5G network, using AWS to host its RAN and core.
“Hopefully by the time the 6G concepts mature, the open RAN concepts will mature, and they will be a lot more ready for deployment in 6G, while they continue to be deployed in 5G,” Ghosh said.
AT&T is one carrier that’s touted efforts on the core side, shifting its existing 5G mobile core and future standalone to Microsoft Azure’s hybrid cloud. Verizon, meanwhile, has a widescale deployment of vRAN for its 5G network using Samsung gear, and has been emphasizing an Intelligent Edge Network build.
The evolution is toward software defined networks, which Sangam said when one evolves from 5G/5G Advanced to 6G, “the underlying compute hardware can be utilized for future generations,” where the next-gen “technology” – is all about the new software.
“Another way to look at it is, as with most virtualized, cloud-native networks, the upgrade cycles between hardware and software are almost disconnected,” Sangam said. So that provides “extreme flexibility” from the perspective of capital and financial investment, next-gen feature and capabilities deployment, as well as timelines for when to do so, and so on.
On top of that, “all of that software will be AI/ML driven, which provides even more flexibility and optimization opportunities,” he said, including with the air interface.
Not all about Terahertz
Ghosh pointed out several pieces are already happening in 5G today, that can carry over, such as learnings from mmWave and massive MIMO tech.
In 6G those “might mature into sub-terahertz, and even higher number of antennas for massive MIMO antennas in 6G, reductions in latency, increases in throughput, better coverage of number served per cell,” she said. “All of those metrics will definitely improve in 5G and the infrastructure that is being laid down to get those improvements will also” help make 6G better.
Sangam too pointed to using a full range of spectrum bands as likely parts of 6G, including bands between 7-14 GHz and sub-Terahertz (if you’re interested in the debate about where the Terahertz range starts, read this).
Carriers, particularly Verizon, put an emphasis on millimeter wave spectrum as part of its early 5G strategy which has since been supplanted by a more significant focus on mid-band for 5G, but remains part of plans.
Still mmWave is important to the 5G picture, even if it has yet to be deployed widely. And while researchers are looking to how spectrum even higher up in frequency could be leveraged, in Ghosh’s view, 6G won’t be all about those super-high bands.
“I hope that people will really learn from the lessons learned in millimeter wave to truly understand where terahertz will fit into the 6G ecosystem,” she said. “It may be a part of 6G. Will it be how 6G is defined? I don’t think so.”
Repurposing and sharing spectrum
Another aspect of 5G she thinks can be leveraged in 6G will be spectrum sharing.
This has made strides, first in 4G in the CBRS band, which shares among three tiers of users including federal incumbents. That kind of sharing will continue to grow in importance because midband is scarce and offers good propagation characteristics, she said, and become a core expertise that can be applied for sharing among both like and unlike users as the industry looks for new spectrum in future generations.
She noted that may be difficult to get dedicated spectrum in higher megahertz and sub-terahertz spectrum that’s used for things like radio astronomy and passive sensing. The concept of sharing between different levels of users started in 4G, matured in 5G and she’s hopeful will continue evolve so that industry would “have learned how to do it right by 6G.”
As noted before, even when a new G is introduced, other Gs still stick around for a while. 5G spectrum auctions are still taking place, (the FCC just announced a July 29 start date for the 2.5 GHz auction) and U.S. carriers including AT&T and T-Mobile, are just shutting down 3G or making plans to sunset their legacy systems - as are UK operators like Vodafone and Orange. And as those legacy networks get shuttered, some spectrum can perhaps be repurposed for 5G, as Orange, for example has said it plans to do.
And by the time we get to 6G, some of the spectrum for 4G could be transitioned out and repurposed, Gosh said.
“You have to have this constant view in the future, of how, as new Gs get designed and implemented, that we don’t hold on to the old Gs for too long, and that’s a delicate balance,” Gosh said, noting some technologies are imbedded in not only smartphones and devices, but sensors and IoT in buildings and other applications.
One area where regulators can play a role in helping operators feel more confidence as they move into the 6G era is knowing that when they’re designing systems, there will be spectrum for them to deploy., she said.
It can’t wait until standards and technology are defined and ready for deployment, Ghosh emphasized, noting that the U.S. is still in the midst of 5G auctions even though 3GPP has had 5G defined for almost three years.
“There is certainty that comes with knowing there is a certain band of spectrum that will be available” knowing the characteristics of that spectrum so operators can start designing networks, devices and protocols to support it.
A Terahertz network is different than mid-band “and knowing what you’re looking at early on will definitely help in developing the best technology one can,” she said.
The network as the killer app
As new G’s come out, next generations discussions always start with “what is the killer app,” according to Sangam.
In his view, the “6G answer for that will be ‘network is the killer app.”
In 5G, the industry is still on the hunt for killer apps. This can be seen by carrier’s concerted efforts to stand up innovation hubs and support developer communities. It’s something T-Mobile just this week brought into focus with its 5G Hub and innovation program to help drive breakthrough innovations for 5G, which have so far, been underwhelming.
Ghosh acknowledged that when starting with a new G, it’s a bit of “grasping at straws” because it’s hard to predict what applications will come. An often brought up example is Uber, which 4G networks weren’t built for but helped usher in an app economy.
“There is a little bit of build it and they will come,” Gosh said. “I don’t think [6G] should be totally driven by that motivation, but some of that has to be there, otherwise we’re not going to be constantly improving the performance and where we are at.”
To her finding a killer app is very organic growth, noting with 5G applications have expanded beyond smartphones and opening up new areas and growth in deployment scenarios like private networks.
Asked if about if the need or technology will come first for 6G, Sangam said so far 3G-5G have been led by technology before need. “For 6G, it all depends on how quickly the later phases of 5G accelerate.”
One ask Ghosh has about 6G: let’s not hype it as much. In her view, it does a disservice when expectations aren’t met. She noted 4G never met its stated goal of peak 1 Gbps downlink, and 5G has yet to hit its goal of 10 Gbps peak download, though it still could.
“We don’t need the hype to keep technology moving forward, let’s aim high but not hype so much,” Ghosh said.