Since 2018, I’ve been working in the area of putting 5G into practice. Back then, early versions of the 3GPP Release 15 specification had become available and equipment vendors started shipping first hardware and software. The standards were not developed overnight, however, but were preceded by visions, ideas, studies, and trials by at least half a decade. In other words, 5G was envisioned while the 4G LTE rollout in the real world was just in its early stages at best. So going from one wireless network generation to the next is very much an overlapping process. And as far as 6G is concerned, this overlapping process is starting to gain speed again.
Jaeku over at Sharetechnote has started collecting resources on 6G already several years ago and from his timeline, one can see that the first phase of creating a vision, collecting ideas, and evaluating what might be technically feasible in 10 years down the road has already started in 2018. That’s just when we started working with 5G equipment in the real world and half a year before 5G was even launched in Europe.
Today at the beginning of 2021, 6G is still in the vision/concept phase, there are no studies, no technology trials, and consequently, no 3GPP Technical Reports yet. After reading a few papers, I can particularly recommend this Samsung whitepaper and this somewhat more technical paper as a good starting point.
And here are my own first thoughts on the topic:
To create a vision for 6G, I think it is necessary to reflect for a moment where we are with 4G, and 5G today and how networks can evolve with these technologies in the next few years. This potential evolution can’t be part of the vision for 6G, it’s already in the specifications today. Rather, 6G must be the next leap beyond what is standardized and imagined for today and the near future. And that’s quite a bit:
5G – The ‘More of The Same’ Part
With 4G LTE, we have a wireless technology today that makes great use of spectrum below 3 GHz. With 5G, spectrum between 3 and 6 GHz has been added. This has brought about a significant capacity increase in traditional cellular network architectures, as there is enough space for channel bandwidths of up to 100 MHz in this part of the spectrum. Speeds of well over a Gbit/s over a distance of a few hundred meters have not only become possible in theory, but also in live networks. In addition, active antenna systems, beamforming, and multi-user MIMO have brought about a further increase in the overall channel capacity compared to the passive 2×2 and 4×4 antennas used in LTE. While signals above 3 GHz bands are theoretically even more range limited than even the highest LTE band used today (2.6 GHz, band 7), active antennas compensate for this to a high degree. From my own experience, good 5G band n78 coverage at 3.5 GHz can be achieved in cities based on today’s cell-site density.
5G – The Pioneering Aspects
In addition to the ‘more of the same’ functionality described in the previous paragraph, 5G is pioneering the use of the mmWave technology between 26 and 43 GHz, and channel bandwidths of up to 400 MHz. Datarates that can be achieved in practice are even higher than in the 100 MHz channels below 6 GHz. The challenge with mmWave spectrum today is that the range is significantly more limited compared to sub-6 GHz channels, and indoor coverage from outdoor cells is difficult at best. That probably makes use of mmWave technology more expensive in most outdoor scenarios as long as capacity requirements can be fulfilled by 5G macro cells below 6 GHz. Satisfying the increasing in-house bandwidth demands in places like train stations, airports, stadiums, and exhibition venues might just be the territory where mmWave deployments can shine and make a massive difference. It’s not an exaggeration to say that gigabits of network capacity in hyper-dense places can be achieved with current technologies, and, from my point of view, the evolution of Wi-Fi, 4G and 5G will be able to cope with the demand of today’s applications for many years to come. Here’s an example: The 36c3 Congress at the end of 2019 in Leipzig with 17.000 participants, who knew how to use ‘more bandwidth’, generated a peak demand of 16.3 Gbit/s in the downlink from cabled and wireless devices (for details see here). In other words: You don’t need 6G for GBit/s speeds.
The other significant new idea in 5G over previous cellular network technologies is the use of cellular network technologies in campus networks, for example in factories for machine to machine communication. Concepts such as edge computing, edge gateways, network slicing, and a flexible 5G air interface have been standardized to create the right mix of power consumption, speed, latency, data locality and priority for many different kinds of applications and locations with particular connectivity needs. And on a nationwide scale, NB-IoT and LTE CAT-M1 exist today in live networks to connect huge numbers of embedded devices. This makes 5G an ideal revolutionary vehicle in this domain. It is still early days in this domain, and it is difficult to predict from the current status quo where we will be with these important topics in five years from now.
So there we go, mmWave and the ideas behind campus networks are the revolutionary 5G technologies that are at our doorstep in practice. Consequently, they can’t be part of a new vision for 6G. Thus, we have closed the circle again on the ‘overlapping process’ part in the day and age of 5G, where a new vision for the future is required while the previous vision is only in the early phases of being realized and hence can’t be interpolated from.
So what will 6G be?
What is clear at this point is that the next network generation has to go far beyond what 5G can do today in the near future, but not so far to make it science fiction for a faraway future. Instead, visions for 6G have to describe possibilities based on technology that is likely to materialize within the next 10-15 years. And that’s what the whitepapers I linked to above aim to do.
As there is no point in discussing Gbit/s speeds anymore, visions for 6G have consequently started to explore how to move to the Tbit/s speed range. As for mmWave in 5G, there are two major challenges. The first is that such high speeds require a new frequency range to have enough bandwidth. That means that ways have to be found to transmit ultra broad signals over the THz frequency range. There’s no technology that can do this today, but the whitepapers point out potential ways to get there. This leads straight to the question if channels in the THz band are just usable for very short range point-to-point data transmissions, or also for a more centralized cellular approach where data has to travel over larger distances. mmWave is already challenging here and transmission channels in the THz region will push this challenge to a whole new level.
Based on the assumption that THz bands can be made usable for TBit/s data streams, the next question that 6G visions have to answer is what to actually do with that amount of data. Especially the Samsung whitepaper linked to above points out that Virtual Reality and Holographic real-time imaging would easily absorb such data rates, particularly if processing has to be ‘outsourced’ from the end device to high-performance computing clusters at the edge of the network. Yes, that perhaps sounds like science fiction. However, go back to the year 2000 (if you are old enough to remember) and ask yourself the question if you could imagine the kind of mobile devices we have today and cellular networks that could support data rates of Gbit/s for a monthly subscription of a few tens of euros. When I compare my answer to this question to what is proposed as a realistic vision for 6G in those whitepapers, it does not sound far-fetched at all. As long as we remember it’s a vision, not detailed technology and implementation description at this point.
So those are my initial thoughts about 6G and I’d be happy to read about your ideas in the comment section below. And I’m very excited to dig more deeply into the topic going forward no to learn what other people envision and how that matches with my ideas of what will become possible in the next decade or two.