5G's Future Is Broken. Here's Why We Need 6G
Think 5G is going to deliver VR, AR even eXtended Reality experiences? What about 8K VR, 6DoF and volumetric video? Think again. Telcos, scientists, and governments have reset expectations on 5G and found it wanting.
"When you do a deep dive on road maps for XR and volumetric video and industrial IoT it very quickly points you to what is possible with 5G and that it won't be able to support longer terms roadmap of these use cases," says Alan Carlton, VP of InterDigital Europe. "6G must finish the work 5G set out to do."
5G isn't broken but nor is its design sufficient to follow through on its potential, he argues. Despite incredulity in some quarters not so long ago that a new generation would ever be needed after 5G, an international 6G standard is already on its way.
What is concerning from a telco point of view is that their ability to return the investment in 5G was predicated in part on selling new video-centric applications to consumers—VR, augmented and mixed reality, 4K streaming. While "entry level" versions of this will be possible, there will be a lot left on the table if 6G doesn't finish the job.
"You may say 'Wasn't all that supposed to be given to us by 5G?' but the truth of the matter is that 5G will only open the door on this roadmap of XR," Carlton says. "If you look at the baseline KPIs of 5G—the ambition is 100Mbps downlink and 50Mbps uplink as the average ubiquitous capability. The long-term roadmap of XR is to get to a truly pervasive everywhere experience. [To get there] you have to bump up those average experience data rates from gigabyte to multi gigabyte delivered to your devices."
According to Carlton, "Current generation 360° 4K video needs 10-50Mbps [to reach the consumer] and next generation 360° 8K needs 50-200Mbps. That is already more than 5G can deliver.
"Looking to the future of 6 Degrees of Freedom (6DoF) or free viewpoint video—technologies which are critical for full immersive experiences—this needs 200Mbps to 5Gbps. You can see how 5G will be challenged. You will get an XR experience on some level with 5G but the roadmap on XR and visual processing technology already breaks 5G in the future.
"In my view you will have great experiences of XR in industrial IoT private networking applications [running at near-peak data rates in 5G] and decent entry level XR experiences for consumer purposes but to get the full immersive 'everywhere anytime' experience, then 5G is at least an order of magnitude short."
XR is not even the most demanding use case for video. Holography is the application usually mentioned as a futuristic use case but a conservative estimate of what holographic video needs is closer to a Terabyte in data rate.
"Holographic video is still very much at the end of the volumetric video roadmap. 6G plays a part to open the door to the roadmap in a way that 5G is fundamentally not designed to do."
Aside from being incapable of delivering enough data fast enough for applications like volumetric video, it is becoming apparent that 5G is simply not designed for the more ultra-precision real-time uses cases outlined in 5G scenarios (remote surgery, precision remote control of robotics in industry).
"What we [need] is not just an enhanced mobile broadband (eMBB) service or a massive machine type communications (mMTC) service or an Ultra-Reliable and Low-Latency Communications (URLLC) service. In a nutshell with 6G we're going need eMBB ++, mMTC++, URLLC ++ and some sort of new service capability which is likely to be a fusion of physical and virtual technologies as manifest in things like ubiquitous deployment of XR capabilities in the world.
Carlton continues, "The point is that all the talk of network slicing is a kind of brute force way of doing things. Having to deploy two or more network slices is fine for a limited number of deployments but to get to a massive scale it is not the most elegant architecture. That's what people are now thinking."
The seeds of change already exist in 3GPP release 17. NR-Lite offers hybrid capability, a mix of URLLC and mMTC. "It is anticipated that we'll need something more to get us to a truly fully flexible system that is able to support all these hybrid services. This will necessitate a 6G evolution," he says.
While 5G networks have been designed to operate at extremely high frequencies in the millimeter-wave bands, 6G will exploit even higher-spectrum technologies. Exploration in this area is also revealing cracks in the 5G architecture.
"If we're going to get anywhere close to Gbps capabilities everywhere, the only way to go is up," he says. "We have push into new spectrum. The 3GPP has already tabled a work item to explore 72 Ghz spectrum bands and to support preliminary XR applications.
"But even when you start pushing 80Ghz or 100Ghz a fundamental problem begins to appear with the digital design of OFDM (Orthogonal Frequency Division Multiplexing) methods. It is related to the underlying physics of dark silicon. There are already whispers and discussions among those exploring high level spectrum that we need to move to a more hybrid digital/analog type of design. There is a reasonable case that in going from 5G to 6G we will need a genuinely new radio design."
Another potential technology playing into 6G architectures is intelligent metasurfaces. Electromagnetic metasurfaces can be characterized as intelligent if they are able to perform multiple tuneable functions.
"This is proving quite promising in delivering a turbo boost to line speed and data rates," says Carlton. "It involves sticking a smart antenna in the radio channel that can manipulate Snell's law (a formula for light waves used to describe the relationship between the angles of incidence and refraction). So, we are talking about introducing a whole new RAN element into the network which has not existed before."
Carlton describes this as "the final frontier of network softwarization—where we are literally able to turn the radio channel into a software programmable entity.
"The basic physics and material science for this technology is pretty mature, but we're only at the point of experimenting with it in the 5G era now. This technology has the potential to inject another break point down the line in the 5G Ran. It may be a necessary technology that emerges in full force in the 6G era."
6G has moved out of academia into forums and into the industrial research and consensus building phase.
InterDigital is one of the organisers of the 6G Symposium (an event which will bring together industry heavy weights such as AT&T, Facebook, Samsung, U.S. Department of Commerce and Verizon). It announced FCC Chairman Ajit Pai and NIST Director Walter Copan as speakers.
Other initiatives include the University of Oulu in Finland; a Sony, Intel and Nippon Telegraph and Telephone Corporation (NTT), collaboration and another from Samsung and LG.
"All of these activities will likely converge into an ITU 2030 requirements type doc with standardization anticipated to start on 6G around 2025," he says. "You might see commercial launch of 6G from 2030 onwards."
All of which begs the question, why stop at 6G?
"I see the relationship between 5G and 6G as quite symbiotic. 5G lays the groundwork and 6G carries it over the line. The only argument I can muster for a 7G is if we all want to have holographic wireless experiences. That would take crazy TB per-second data rates just do that over fibre. But maybe that will be the use case that drives us beyond 6G."
5G holds promise for smart cities, smart vehicles, and industrial automation. For video streaming? Not so much.
An ambitious demo for IBC's Showcase brought together multiple broadcasters and vendors to demonstrate 5G's potential for remote production, but showed that it's a long way from ready for primetime
Speeds of 4.3Gbps could download an hour-long 4K episode in just 14 seconds; the old record was 2.92Gbps