Future networks will be built by evolution or revolution from the institutions, companies, and driving forces prevailing today, if only because of the huge mass of knowledge and investment dollars that these represent. It follows that, if we want to gain a clear projection of future possibilities, it’s worth examining the currents driving today’s market. 6GWorld™ sat down with Ray Dolan, CEO of Cohere Technologies and industry veteran through his work at companies such as Flarion and Sonus, to talk about some of these currents.

Geopolitics in 5G and 6G

The most obvious trend in telecoms and technology is the global tension between regions aiming to lead in technology.

“There’s not a lot of Western leadership in 5G. A lot of it got decimated while China continued to invest, really through a national industrial policy. Now we in the West are starting to explore whether we need, as individual nations and in collectives, a 5G+/6G policy,” Dolan commented.

“Some of the options that exist for the West to respond to China are propping up Nokia, Ericsson, and Samsung. There’s a lot of logic to that,” he observed. However, this is where economies of scale and the ability to invest in R&D come in. “They’re just not large enough at this stage to keep up with the investment required in 5G, let alone 6G.”

This is a reasonable concern. Major telecom players Ericsson and Nokia each spent over $4bn on R&D in 2020, but Huawei spent over $21.5bn. What’s the alternative, then?

“Let’s have a significant infusion from a coalition of willing partners. Once you start thinking like a coalition, you start thinking ‘Do I have any other options?’ One of the other options is to use this open RAN architecture which has become such a pervasive movement in [industry collaborations] O-RAN and OpenRAN,” said Dolan.

There certainly are arguments for an open RAN architecture, but they tend to focus on flexibility, competition and open playing fields within the telecoms environment. How does this translate to geopolitical competition? 

“If you put the network together in a cloud architecture, into the supply chain comes Amazon, Microsoft and Google and companies that have trillion-dollar valuations, that can step up to the requirements for research,” Dolan explained. “You enable the entrance of the compute supply chain, the datacentre supply chain… companies like Azure, Facebook, Google and VMware. Suddenly the Telecom Infra Project seems a lot more viable.”

It’s an interesting alternative to government-controlled industrial policy, but is there a risk that governments become overly dependent on a few large companies that take a strategic role and become too big to fail? Not so, according to Dolan, who sees the benefits of openness to foster competition within the sector. For example, “A platform can run on any cloud company’s infrastructure like AWS, Facebook, Azure, Equinix, Apple Cloud… so there would be way more than just one cloud platform provider. Nobody becomes indispensable.”

Beyond 5G in a Cloud-Based World

Looking ahead to the implications of a possible cloudified, open world, there are certainly some advantages, such as multi-access edge compute, which could simplify future “G” roll-outs.

“The industry was successful in 5G in making cellular and WiFi appear to be a common solution… and that was one of the fundamental platforms that tied things together,” Dolan observed.

On the other hand, “5G took the 4G radio and enhanced it; it really didn’t change the radio. So in a way it’s set up for 6G to eclipse it.”

In addition, the historical development cycle for generations of technology has long been compared to other digital elements, Dolan commented. “People came along saying ‘5G doesn’t support autonomous vehicles, that’ll be 6G’. Well, if you believe in a 10-year cycle then that translates as ‘Oh, we missed autonomous vehicles, we’ll catch them in 10 years.’ The world is moving way too fast for that.”

Taking those elements together – an open, multi-access platform, greater speed to market, and new radio – what might emerge as the next generation?

“When I think of 6G, I think of it as a platform architecture – along the lines of what Cohere has built – that enables spatial multiplexing, cloud RAN, lower latency and a lot of other principles, but also allows 4G to work, 5G to work, 6G to work,” Dolan mused. “Instead of taking 10 to 15 years to get to a generation, since everything on the platform works the same you’ll start to see people saying ‘Well, instead of having one 6G that works for 15 years, how about if we have a 6G flavour that works for autonomous vehicles, one flavour for remote medicine, then a 6G flavour that works for streaming Netflix and Amazon?’”

6.1G, 6.2G…

As we saw with 5G, the idea of meeting very varied requirements with one ‘G’ is problematic. At the same time, it is clear the demands on the network and radio resources are very different depending on context. “Some of these technical requirements are at odds with each other. You might want certain latency and reliability characteristics, certain caching characteristics to exist for streaming but you have hyper low latency and reliability for remote medicine and something else for remote vehicles,” said Dolan.

One possible solution that Dolan sees is to take a new approach to radio waveforms. Rather than trying to deliver all services on a ‘G’ over one waveform, why not take advantage of their varying properties to deliver specific services that suit them? “If all of those can be just slightly different waveforms then I think you’d have 6.1G, 6.2G, 6.3G as opposed to 6G.”

This would, of course, require a flexible platform capable of handling multiple waveforms, which could be compute-intensive. At the same time, it would create some new ways for the telecoms industry to interact with, and deliver value to, other industries. As the radio network and the waveforms it supports become more tied into the performance of a particular service, it may create a new market.  

Dolan commented, “I think of the different waveforms almost like different applications. If you create a platform for everything to ride on, it’s much less disruptive to have 6.1G, 6.2G etc… You can shorten the cycles for releasing different waveforms from 10 years to five years or even a year.”

The implications of approaching new generations of telecoms like this are far-reaching. Historical cycles of investment and renewal of infrastructure would need to be re-thought. Meanwhile, the telcos’ approaches to marketing need to be significantly different in such an environment, moving away from an emphasis on supplying the new ‘G’ towards selling a new portfolio of network-based services.