Sensing, deep learning and upgrades to massive MIMO and full duplex are emerging as some of the key technologies that will define 6G, experts say.

Jeffrey Andrews, director of 6G at University of Texas at Austin, says he expects 5G will be able to handle much of the data loads but the huge breadth and coverage of 6G meant it would likely operate as "a few systems within a system with perhaps different modes of operation."

Speaking at a Samsung-organized event last Friday, Andrews said sensing, which is essential for use cases like autonomous driving and robotics, would be one of 6G's important capabilities.

Figure 1: 6G's dense and tower-mounted basestations would be ideal sites for radars and lidars.
(Source: Kirill Ivanov/Alamy Stock Photo)

"Driverless cars don't really work unless they can sense things and know about things outside their own field of vision," he said. "This is where the network comes in, both in collecting sensor data from devices as well as actually sensing itself."

Andrews said he expects 6G's dense and tower-mounted basestations would be ideal sites for radars and lidars. The 6G network would be able to offer sensing-as-a-service in addition to communication, he added.

He said deep learning would be needed to help design and operate 6G networks.

Sharing, cooperation and AI

To help control costs in running the ultra-dense next-gen network, he anticipated "an unprecedented amount of sharing and cooperation and reuse of infrastructure." This complexity would require some kind of AI to manage it.

Additionally, deep learning could help solve problems in optimizing 6G basestations – another complex problem. Because each cell has a unique propagation environment and serves users with their own distinct behavior patterns, custom human-based tuning was not very effective. Deep learning could have a big impact on site-specific optimization, Andrews said.

He said UT researchers were also working on improving massive MIMO performance in high spectrum bands.

Currently, locating the user device with a narrow beam is a laborious process and a key bottleneck that had to be solved for mmWave and 6G.

"If we're going to have any hope in going to the higher frequencies we need to be able to quickly align beams and then track those beams. It is key to being able to adapt to the cell propagation environment so we don't spend a lot of time sending beams off in directions that are fruitless."

Samsung Research is also focusing on upgrading massive MIMO for 6G, said senior vice president Charlie Zhang.

He said his team had developed a distributed, massive MIMO technology to boost both capacity and coverage. So far they have been able to increase capacity by 230 Mbit/s.

Zhang also cited the development of XDD duplex technology, which has the ability to combine downlink and uplink traffic in FDD and TDD modes.

It's now been included in 5G Advanced standard, but "there's a chance of moving to full duplex, where in the same slot of time or frequency you can support both downlink and uplink simultaneously," he said.

"It's very exciting because if we can achieve that we can almost double the capacity."

Want to know more? Sign up to get our dedicated newsletters direct to your inbox.

Offering an operator perspective, Takehiro Nakamura, a senior vice president at NTT DoCoMo, identified coverage and spectrum as key 6G issues.

"We really want to extend the coverage drastically," he said. Currently, DoCoMo's 5G network, like every other telco network, is built where people are, he pointed out.

"But for the future, IoT devices will be everywhere so we need to provide service not just where people are, but in the mountains, in the sea and sky and maybe even space, too," he said. "We need to achieve 100% geographic coverage."

Related posts:

— Robert Clark, contributing editor, special to Light Reading