By Ray Dolan, CEO, Cohere Technologies
While 5G networks are being rolled out in some markets across the globe, conversations around 6G are already ramping up, with large industry players beginning to raise questions about what it will look like, and which countries may take the lead.
There are many parallels to the early 2000s, when 3G was just getting rolled out but conversations were starting on 4G. Back then, the internet had become pervasive during the period of 3G’s development, and there was a technical and economic disconnect between two very large markets – the internet and wireless. Operators around the world had just paid massive fortunes for new spectrum at auctions with the intention of delivering “mobile broadband”. It quickly became obvious that CDMA was never going to be able to truly provide simple, affordable, seamless mobile internet access at scale. People developed WAP (remember that?) and BREW (another oldie but goodie) and proposed business models using “walled gardens”.
Yes, the idea of creating a parallel internet – the “wireless internet” – was conventional thinking at the time. Seems crazy in retrospect, right? Of course, the iPhone changed all of that, taking the limited value of “smart phones” all the way to the handheld computer that we now call a phone. It was audacious enough to suggest that consumers would want the entire internet when they moved. If you need to refresh your memory on all this, wait for your phone to revert to 3G when you’re traveling and try using the internet. That’s what many people thought “mobile internet” was going to be.
Clearly, wireless needed to conform to the existing internet. Driven by the bold launch of the iPhone, and despite resistance from incumbents, it did. The same is true today – wireless networks need to conform to cloud architectures. And they will.
It’s this simple: cloud architectures offer insurmountable benefits in economics, speed of deployment, and agility. Perhaps most important of all, a cloud architecture is the only way for today’s operators to penetrate the enterprise market. The enterprise opportunity was the justification for bidding massive amounts on 5G spectrum, as operators explained plans to move beyond consumer “all you can eat” plans to the high margin services needed by enterprises. In theory, network slices would provide guaranteed service levels so that enterprises could use wireless networks to drive industrial automation, remote medicine, augmented and virtual reality, etc. Unfortunately, today’s cellular 5G networks fall short on this goal. Operators are now proposing “private 5G” for the enterprise, and private 5G is a cloud architecture. Yes, “private 5G” is cloud.
History doesn’t always repeat itself, but it often rhymes. Operators have paid large fortunes for spectrum only to find the technology lacking. In time, broader cellular 5G networks (perhaps renamed as 6G) will evolve to cloud. There is no other way to establish the economics to support the continued competition as consumers demand more access for the same or less money. And there is no other way to unlock the massive enterprise market.
Will 6G establish a new waveform?
Compared to 5G and its predecessors, 6G will be a new telecoms waveform (or “language”) that carries information differently between the device and the network. This is because the low latency, high mobility use cases (such as driverless cars) and critical applications (think real-time healthcare technology) aren’t efficient enough on existing OFDM waveforms. Instead, a new waveform will be required to enable the high-quality connection for devices on the move. One candidate emerging is Orthogonal Time Frequency Space (OTFS), which suffers less distortion and can enable wireless networks to live up to their potential in ways that 5G has as yet failed to do.
In practice, it is likely that multiple waveforms will be adopted to facilitate the roll out of 6G. OTFS is a flexible waveform, and it can be dimensioned differently in multiple flavors to address the unique needs of various applications: some are mostly mobility-focused, some are mostly latency-focused. OTFS can adapt to both existing and future applications. This type of approach is the only way to bridge the gap between the rapid cycle of cloud innovation and the prolonged cycles of wireless standards. It’s time to break the notion that wireless standards happen every 10 years. Embrace software. Embrace cloud.
Software and cloud vendors join the supply chain
As wireless networks evolve to cloud, new players will enter the supply chain. Today’s required skills in building and operating massive wireless networks will merge with the skills needed to run massive cloud networks and serve the needs of the enterprise.
This is an extremely exciting transition. We will soon have massive global wireless networks running on merchant silicon, with software stacks that evolve more quickly as new services are tested and brought to market. The debate that rages today about whether the compute resources live under the tower or move to the edge datacenter will reach a proper compromise. Some additional compute resources will be deployed locally at the tower due to ultra-low latency applications that can’t be managed from the cloud. Most of the exploding compute resources, however, will grow in a new data center architecture – where they belong. The intelligence that exists today in the “base station” will migrate to the data center, and a new “cell-less” architecture will emerge that can be controlled and scheduled primarily from the cloud.
This has large structural and economic implications. The early decades of wireless (1980s and 90s) were dominated by profitable network suppliers, while handset leadership rotated between players like Motorola and Nokia. Since then, the network suppliers have struggled as digital caused pricing to collapse, leaving the industry with consolidation as its only strategy. Meanwhile, profit pools migrated to newer handsets (smartphones) and the few silicon providers that could invest in the massive challenges of multi-mode integration.
6G cloud will shift this picture again. Networks will naturally provide multi-mode access, relieving the device burden in cases where costs are critical and needs are simple, such as the IoT. Operators will easily and affordably support modes almost in perpetuity since they can be delivered in software containers in the cloud depending on how markets evolve. This includes 2G, 3G and 4G. Imagine operators being able to decommission older networks and allowing the tail of those markets to ride on new networks (5G or 6G, call it what you want) until those devices disappear. The underlying spectrum could be repurposed for the deployment of far more efficient networks that can naturally host any generation. Only a new cloud platform will host these modes naturally.
Despite this shake-up, it won’t all be bad news for the current industry stalwarts. They’ll just need to be prepared to take a risk and adapt. Cloud and 6G are unavoidably interlinked: The G is Dead, long live the cloud. The companies that want to take advantage of this must act fast and evolve. For cloud players this means stepping into a new area of technology and connectivity, and for telcos this means competing in a whole new sphere, against hungry, emerging innovators.
Join Ray Dolan, 50 other expert speakers and over 200 senior attendees at 6GSymposium next week, taking place live in Washington DC. Or, if you can’t be there, 6GWorld will be publishing footage from the event throughout October.
