VB Survey to share how your company is implementing AI today.
Just as the earliest 5G networks began to go live two years ago, a handful of scientists were eager to publicize their initial work on the next-generation 6G standard, which was at best theoretical back then, and at worst an ill-timed distraction. But as 5G continues to roll out, 6G research continues, and today top mobile hardware developer Samsung is weighing in with predictions of what’s to come. Surprisingly, the South Korean company is preparing for early 6G to launch two years ahead of the commonly predicted 2030 timeframe, even though both the proposed use cases and the underlying technology are currently very shaky.
Given that the 5G standard already enabled massive boosts in data bandwidth and reductions in latency over 4G, the questions of what more 6G could offer — and why — are key to establishing the need for a new standard. On the “what” side, Samsung expects 6G to offer 50 times higher peak data rates than 5G, or 1,000Gbps, with a “user experienced data rate” of 1Gbps, plus support for 10 times more connected devices in a square kilometer. Additionally, Samsung is targeting air latency reductions from 5G’s under 1 millisecond to under 100 microseconds, a 100 times improvement in error-free reliability, and twice the energy efficiency of 5G.
The obvious question is “why,” and it’s here that Samsung is either grasping or visionary, depending on your perspective. Some of 6G’s potential applications are clearly iterative, it notes, including faster broadband for mobile devices, ultra-reliable low latency communications for autonomous vehicles, and factory-scale automation. Better performance of 5G’s key applications will appeal to some businesses and consumers, as will support for next-generation computer vision technologies that well exceed human perception: Samsung suggests that while the “human eye is limited to a maximum resolution of 1/150° and view angle of 200° in azimuth and 130° in zenith,” multi-camera machines will process data at resolutions, angles, wavelengths, and speeds that people can’t match, eating untold quantities of bandwidth as a result.
On the human side, Samsung also suggests that 6G will be needed to enable “truly immersive” extended reality, as next-generation XR headsets will need around 0.44Gbps throughput to power human retina-matching 16 million pixel displays; that’s more individual bandwidth than what 5G networks can guarantee. Similarly, Samsung expects that mobile and larger displays will begin to display actual volumetric holograms, requiring “at least” 580Gbps for a phone-sized 6.7-inch display, and “human-sized” holograms at several terabits per second.
To the extent that holographic displays are a known concept to many people, another “key” 6G application — “digital twins” or “digital replicas” — isn’t. Going forward, Samsung expects that people, objects, and places will be fully replicated digitally, enabling users “to explore and monitor the reality in a virtual world, without temporal or spatial constraints,” including one-way or two-way interactions between physical and digital twins. A human might use a digital twin to visit their office, seeing everything in digital form while relying on a robot for physical interactions. Duplicating a one-square-meter area in real time would require 800Gbps throughput, again well beyond 5G’s capacity.
Like other 6G researchers, Samsung is pinning its hopes for the massive necessary bandwidth on terahertz radio spectrum, which has a wide array of technical challenges ahead of commercialization, just as millimeter wave did before 5G launched. It’s also expecting AI to take on a more significant, system-level role in 6G, becoming explicitly required for every 6G component to enable distributed AI and split computing efficiency throughout every piece of the network.
At this stage, Samsung expects international standardization of 6G to begin in 2021, with the earliest commercialization happening “as early as 2028,” followed by “massive commercialization” around 2030. That would roughly parallel the accelerated timetable that saw 5G take eight years to go from concept to reality, as compared with 3G’s 15 years of development time. Between now and then, however, engineers will need to figure out ways to create even more massively dense antennas, improve radio spectral efficiency, and handle other novel or semi-novel issues introduced with terahertz waves. Only time will tell whether those challenges will be summarily overcome, as with 5G, or will hold 6G back as an engineering pipe dream until later in the next decade.