By Alex Lawrence
CEO of Cohere Technologies, Ray Dolan comes across as something of a paradox. While he leads a company that is highly invested in the future of wireless radio communications, he doesn’t believe that questions about the radio will be central to 6G. In a recent interview with 6GWorld, Dolan outlined several arguments underlining why that is the case.
Wrong 5G, Wrong Time?
A lot of the debate about 6G is ‘We haven’t really even done 5G yet, guys, wait your turn.’,” Dolan began. “It’s a very natural response, but I would point out that the shortest lifecycle of the wireless industry up until today was 3G, because it really didn’t do anything except extend 2G. When the world pivoted significantly into mobile internet it needed something different.”
Dolan underlined the fact that, in 3G, iPhones caused a strain on a system that was designed to be a narrow “mobile web” as a subset of the full web experience.
“When Apple was bold enough to put a computer in your hand with a phone app and call it an iPhone it broke that premise,” Dolan explained, and the result was a drive to a much richer mobile broadband in 4G.
“It is a very good possibility that we’re all in the same situation – that 5G as it’s known today will in fact have the shortest lifecycle because it really doesn’t do anything new except throw more resources at a problem that’s already reasonably well-solved by LTE,” Dolan suggested.
“So this is how I’m thinking about the chronology: it’s either very far-off because there are no use cases that strain 5G, so 5G muddles along and continues to be built. More spectrum gets issued, more people do Ultra WideBand, but the user experience doesn’t change much. Or the kicker is a use case like enterprise or industrial automation, autonomous vehicles, AR/VR… and then, when change happens, it will feel like it happens overnight.”
Private Networks Point The Way to Radical Change
Dolan appears to be a fan of the saying that “History never repeats itself, but it often rhymes”.
“I’ve seen this movie before, specifically in the transition to CDMA and in the transition to OFDMA.” Dolan, the former CEO of OFDMA pioneer Flarion, is arguably well positioned to tell this story. “It’s slow, and then it feels like it happens overnight – and this is why.”
A story emerges of 3G standards championed by different regions, such as CDMA and WCDMA. The challenge that these faced was simple: they couldn’t provide the performance needed to support true mobile broadband services. However, there was a significant political and vendor push to adopt these technologies. This opened the door for Worldwide Interoperability for Microwave Access (WiMAX).
“WiMAX was a new supply chain. You didn’t have to deal with the incumbents. It was going to be on new, unlicensed spectrum. So it answered all of the issues, except that it didn’t work.
“WiMAX tried to use an unlicensed spectrum and they tried to do it without mobility. They tried to succeed without a whole bunch of ingredients. When the iPhone broke CDMA, the world went OFDM on the wide-area cellular network.” OFDM is the signal format used in LTE.
“Parallel this to today,” Dolan continued. “The world is going to go cloud native. It’s going to go Open RAN. It’s going to go in Kubernetes containers, with a new supply chain. And what makes that clear is – that’s exactly what private 5G is.”
In other words, just as WiMAX was a bellwether for technologies and approaches that transferred into “mainstream” mobile once a certain tipping point was reached, so private 5G will pave the way for the next iteration of the mobile market. Dolan’s prediction?
“People will explore private 5G networks. And they will have an impact because small cells are relatively mature right now. Enterprise players are dying to get into this space. If they stay inside the four walls of the enterprise some portion of unlicensed spectrum or CBRS will work, and you will see a ton of buzz about that. But my prediction is all of that work will transfer to licensed spectrum when operators move to this new architecture, and the trigger for that will probably be global multinational industrial automation or AR/VR.”
The Power of Cloud
The appeal – and the risk – of cloud-native telecoms is clear. From a government perspective it offers many countries an opportunity to play a strategic part in the supply chain in ways that China, Sweden and Finland now do.
“If you break up the [existing] market it’s the best time to enter. You’re going to see a lot of push around Open RAN and virtualisation, and that’s the opportunity for the technology architecture to move to cloud.”
Commercially there are some great opportunities to start engaging with enterprises in ways they already understand. However, it does dramatically change the competitive landscape, as Dolan pointed out.
“It opens the supply chain to people that really understand the enterprise and have the global support to drive industrial automation. Carriers are regionalised – even though they’re big, they’re regionalised. The enterprise’s app will be able to run through a cloud and the SLAs will be able to be driven by the application providers.
“In turn that will open up the opportunity for system integrators beyond Ericsson, Nokia, Samsung. And you’ve recently seen companies like Microsoft buy Affirmed and Metaswitch, trying to really understand the virtualised core EPC and trying to get from the EPC into the virtual RAN.”
This paints a picture that can seem quite intimidating to the established telecoms players. Arguably the telecoms service providers have never quite managed to crack the enterprise market in the way they would like, and 5G had been promoted as the opportunity to gain relevance with this rich market. Opening the door to companies which are already major global powerhouses threatens to undermine that.
However, the other side of the coin is the opportunity for radically different network economics achievable by going cloud-native. Dolan pointed to Rakuten and Dish as a new generation operating under very different economics: while still relatively young and unproven – in Dish’s case, yet to launch commercially – he is confident that the financial advantages they bring will be disruptive.
Meanwhile, brownfield operators may struggle to change as quickly, in some cases due to major single-vendor dependencies. However, “others think that the overall economics of going to cloud, and the benefits of doing that in the next couple years, are going to be so profound that they’re exploring cloud partnerships while they’re talking with Nokia and Ericsson,” Dolan noted.
Bringing the Radio Into Cloud Evolution
All of this is interesting as an overview of the market, but on the surface it’s not directly related to Dolan’s company, Cohere Technologies. The company has been making a name for itself in the last few years for what it calls its Universal Spectrum Multiplier, and recently announced a further $46 million in investments from a diverse array of players. Essentially it allows for spatial multiplexing through a software implementation, opening up the ability currently to double capacity at cell sites.
“We’ve shown that it works on 4G; and we’ve shown that it works on 5G; and soon we’ll show it working on 4G and 5G together,” Dolan noted.
He likens the technology platform to a cathedral or other building with good acoustics, enabling easy conversation. As he points out, the building is agnostic to what language is being spoken. If Cohere is able to demonstrate their technology over 4G and 5G simultaneously, “it means two different tours are going on in the cathedral. And no one cares that they are speaking a different language.”
“Now if you have a platform that puts 4G and 5G in the same place, 6G radio would just lay on top of it in software. So this notion of ten years as a cycle for 6G is no longer relevant.”
As a side-note, it is unclear how 6G will end up being defined. Dolan here is referring to some new radio element forming the definition. If post-5G technologies are defined by new radio waveforms or capabilities, then Dolan’s statement may be true. However, while this is an entertaining conceptual game, that doesn’t answer the question of what it has to do with Cohere. Happily, Dolan is not short of opinions.
“Let’s assume that 6G is in fact a new language and a new architecture. Ericsson, Nokia, Huawei, Samsung are trying to keep it all localised at the base station. They’re trying to keep it closed in a new, open world. And while it’s virtualised, they’re trying to keep it on proprietary ASICs because it’s more integrated, it’s more power efficient and smart” he began. However, he says that Cohere will move its platform to a US datacenter in 2022 in collaboration with several major companies.
“When Cohere proves that we can create a cloud-based platform; and we give the operators an option to leave it local under a tower, or move it to a datacenter, or both; and we spatially multiplex 4G and 5G and even show OTFS as an option for a new waveform, we will usher in that debate on what 6G means.”
“Is what we’re doing in fact 5.2G? We’re fine with that outcome. We’re also comfortable with people saying ‘That’s 6G, because 5G is in fact perpetually a localised base station solution and 6G is in the cloud’.”
There is another element which, in Dolan’s view, separates 6G from any particular radio technology: the ability for platforms such as Cohere’s to support large numbers of different waveforms in licensed and unlicensed spectrum.
“Imagine a world where AR wants fat bandwidth at low latency; IoT wants a narrow pulse at low latency; automated vehicles need a medium bandwidth pulse at low latency that’s also connected to LIDAR and radar in a complimentary way because it ends up having multipurpose control plan.
“All of those things are very complex if you try to do that in a 5G or 6G world, but if you have a platform that hosts an additional waveform with no cost to do it, then waveforms can be created for cars that are fundamentally different than for goggles, and they are fundamentally different than for a meter that’s being read for Pacific Gas and Electric. And those markets by themselves are all large enough that they can support their own modems.”
A Cloud Centric Future… A Renaissance for Networking?
So far, much of the cloud “opportunity” seems to be oriented towards potential new players. While there may be some cost savings to be made, it is hard to imagine many companies with entrenched revenues wanting to open themselves up to the risk.
However, that’s not necessarily the case, according to Dolan. “I think there’ll be a bit of a Renaissance on the network side and a bit of a reduction on the importance of handsets. At least that’s a possibility. If it happens, it’s going to be a very big deal.”
It feels like that statement needs quite some explaining.
“Up until now, the multi-mode challenge has been handled on the handset side and networks have been largely single mode,” Dolan began.” The handset players like Qualcomm, Samsung have had to build chips that had every modem, and that’s so well established that people think that’s the only way it can happen. As a result, it’s concentrated the silicon solutions on the handset side. It’s also made for some very, very expensive handset implementations.
“Now imagine a world where the network becomes multimode through a cloud player, and a solution that’s needed for a device can be single mode. You could use a GSM modem that costs $3 to integrate, not $45 because it’s also trying to do everything else that you could possibly imagine.”
There is certainly an appeal to this, not least because lowering costs for modems would open up the range of connectable items and drive up the number of connected devices for the operators. At the same time, there are benefits for handling and decommissioning legacy networks.
“Right now people are thinking, ‘Do I decommission my 3G network? If I decommission my 3G network, all of these embedded devices that are relying on 3G service are going to go out of service.’ Instead people are going to be able to maintain their 3G network in software and just spin it up at the volume that’s required.”
To wrap up, Dolan comes back to his overall position. “Probably the move to cloud is going to require 6G. 6G is cloud. Cloud is a new economics and cloud is a new supply chain. And there’s no question that going to cloud will protect and probably enhance security.
“Despite all of the arguments against that thesis from an Ericsson or Nokia, going to cloud will usher in a new supply chain. That heavily favours the Western world, which has been virtualised and cloud- and internet-centric for a long time. So software companies and cloud companies will enter the global supply chain, if not in 5G in 6G.
“And the proof of this thesis lies in private 5G networks.”
