With 5G just being rolled out across the globe, it may seem a little early to speak about the next-generation of mobile network technologies, or 6G. Yet, research has already begun, though it is still in its nascent stages today.
Already, the race is on to create the standards for what will offer 100 times faster speeds than 5G and a small fraction of the latency that 5G offers. It could be at least a decade away but a vision is being formed on what is possible in the future.
The first instances of 6G may indeed be different from what is envisioned today, said Roger Nichols, 5G and 6G program manager at Keysight Technologies, which delivers test software and equipment for telecom networks.
However, the innovation of the past few decades had been hard to imagine as well, he noted. “People my age were in their 20s before the idea of carrying a radio in your pocket to communicate with someone in another country moved from science fiction to reality.”
NOTE: Responses have been edited for brevity and style.
Q: Is it too early to start talking about 6G, when 5G is not yet fully rolled out?
A: Operators and vendors are not struggling to make 5G arrive. The rollout of this technology is consistent, if not faster than that of previous generations and it is paced by policy and economic factors. These investments are high and those making the investments have shareholders to please in the near-term.
It is not too soon to begin the 6G conversation, these evolutions take time. The technical hurdles are significant, and many innovators must contribute to overcome these hurdles and develop systems that will work consistently around the world.
People my age were in their 20s before the idea of carrying a radio in your pocket to communicate with someone in another country moved from science fiction to reality.
It may not seem that long, but the 10-20 years that it takes to work out a new generation means you start working before the previous generation is mature and sometimes before it is mainstream.
There is speculation that 5G was the last “monolithic” generational transition and that “6G” may look more like an evolution. This is not true because of the dramatic changes in system behavior described above. Like 5G, some of 6G will be an evolution, but some will be revolutionary and taken as a whole; we can expect a step-function in technical capability.
Two additional items I need to clarify: 5G and 6G interest developed a good decade apart from each other. There are no R&D projects on 6G now – only research. Same was true when I started working on 5G in 2014, no development, just research.
Q: We are still looking for applications for 5G today. What can 6G do?
A: It is a bit early to speculate on the “biggest application opportunity” since that is another word for “killer app” and such a word no longer has meaning. The use-models range from holographic communications that include information beyond sight and sound, to making digital twins far more sophisticated and thorough, to changing the way that we leverage data through machine learning (ML) and other forms of artificial intelligence (AI).
There is also an intent to leverage 6G for sophisticated emergency and disaster management, as well as huge scientific applications. Much of what is being described today looks like “5G, but better” and some of that will be the case.
We will see new use-cases evolve from the advancements of 5G that while being “introductory” in 5G, will become mainstream in 6G. These will have to do with systems that are more flexible, more efficient, and can intelligently handle orders of higher magnitudes.
Regarding latency, 6G will make two advancements on time-sensitive networking. First, the maximum latency KPI for critical applications for 6G as low as 100 microseconds. This will have a significant impact on new use-cases that are not just about getting information there quickly, but also leveraging for location-based services that are more precise.
Second, is the concept of a minimum latency requirement; in other words, precise timing of information transfer even if the latency is long. Some information can arrive too soon to be useful, so time-engineering in networks means knowing that a message may arrive before a certain amount of time passes, as well as the ability to plan exactly when the information will arrive.
Q: When do you think 6G will be available?
A: Most (researchers) are targeting a 2030 commercialisation timeframe. Like 5G, we could see commercial launches happen before the turn of the decade. This will be determined by that magical mix of technology, policy, and business model.
In any case, like all previous generations, commercialisation will start from small regional implementations to nation-wide networks with all the fits and starts of any new system.
Also, like 5G, the first manifestations of 6G will be a far cry from today’s vision – we are a good five years away from seeing 5G in full, based on vision-setting that was completed eight years ago.
Q: Will O-RAN continue to be a business opportunity under the 6G network?
A: Whether it is the O-RAN standard itself or something beyond that, the drivers for O-RAN are more pronounced in the 6G context.
The vision for 5G is to virtualise as much of the network as possible, to decrease costs and drive flexibility into the system. O-RAN is a natural fallout of that given that it is specific to virtualising the parts of the radio access network (RAN) for which such work is practical.
While virtualising power amplifiers and antennas is perhaps the realm of some future generation, the virtualisation of much of the baseband processing is happening now. This drives a disaggregation of the base station and from that is the need for common architectural boundaries and standardised interfaces between them.
This trend will accelerate to ensure 6G is flexible. One other fallout of O-RAN is the development of the RAN Intelligent Controller concept – a necessary control element that sits astride the RAN/Core boundary to ensure optimal use of the virtualised elements of the RAN. 6G is likely to drive the next generation of this architecture, with additional AI-based functionality and AI-based capability.
