Two days ago I published a post predicting that 6G will become reality around 2035, meaning that people will really start using it. Of course I am expecting labs demo and trials well before that, like it happened in the previous Gs and it is happening now with 5G.
Yet, among the several comments I received, one was stating with certainty that 7G (yes, 7, not 6) will be available in 2032. Unfortunately the comment did not articulate on what bases that prediction was made. Anyhow, it made me think. Could we reasonably expect a 7G to occur some 13 years from now?
If I look back I can see an amazing linearity in evolution (time-wise) for Gs: we basically have 10 years separation from one G to the next, both if you take as reference frame the first trials, the first deployment, the mass market uptake and even its demise. Based on this one should say that as we will have real mass market adoption of 5G in the first part of the next decade (I placed 2025 as the point where most everybody would be able to own a 5G device and get access to a 5G network -and I consider this as an ambitious target) than 6 G should come after 10 years, in the 2035 time frame and 7G 10 years further down the lane, i.e. 2045.
On the other hand, some people are making forecast based on the law of accelerated returns, like Ray Kurzweil, stating that evolution is accelerating and what used to take 10 years will be taking 7 years, and then just 5 and so on. I guess that based on this accelerate return hypotheses you get a shift from 5G to 7G in 12/13 years and that might be the reason behind the comment (I hope the one who made it will come forward with some substantiation, always nice to hear different viewpoints).
Personally, I do not think that the law of accelerated returns works in infrastructures. This is an area where evolution is steered by economics, with technology being an enabler. Besides, the economic drive, provided by the demand side, tends to become weaker and the overall infrastructure is exceeding customers/users needs. More on this in a few moments.
There is another reasons why I do not think we will be seeing an acceleration towards new Gs. If you think about the basics of wireless you see that the real technology driver that has enabled the progress from one G to the next has to be found in the evolution (increase) of processing capability flanked by the increase in battery density (capacity). This latter, however, cannot go on forever (although we still have plenty of room to decrease the need for power of electronic components -the Landauer limit is some 100 years away-) because as you increase power use you have to manage the power dissipation and you won’t like to have a red hot brick in your hand…
The processing capacity increase made possible to use higher and higher frequencies, with 6G we might jump into the THz space, but now the processing capacity increase is levelling out when we look at a single chip. In mobile devices you don’t want to have many chips, they won’t fit in the sleek cases we have come to love, and they would increase power consumption. We will keep seeing for a few more decades processing capacity increase but at a lower pace (the areas of GPUs was an exception to these rules but that was the result , mostly, of parallel processing).
So on the technology side there will be evolution and it might just be that the law of accelerated returns may compensate for the slowing down in technology evolution in the processing area, being able to maintain the pace we have seen so far. I doubt it could accelerate that pace.
Back now to economic considerations. As we move up in frequency (something enabled by the increased processing capacity) we are confronted with propagation issues, and we are forced to either increase the wireless power (something that is not allowed by regulators and that would decrease battery time) or make cells smaller. The latter is what needs to happen. The problem with smaller cells is that you need to increase the investment at the edges. The number of antennas required by 5G, assuming you want to deliver the 5G capacity potential and cover the same area, is ten times more than 4G: that means a huge infrastructure investment (in antennas, optical fibre drops and space rent).
You move into the THz and you have to scale up the density of the infrastructure with skyrocketing cost.
Actually, I do not think that a 6G pervasive infrastructure can be economically sustainable, based on today’s business models and approach to infrastructures ownership.
What is most likely to happen (it is already part, technically speaking, of the 5G architecture) is to have the wireless edge created by a myriad of players, each one investing on his own, with no need to recap the investment in a direct way (i.e. by getting money back). What I can imagine is that I will buy a new device, a car, a wearable smartphone… whatever, and that device will become an edge network node. The big infrastructures, pipes, will grow at their own pace, sufficient to manage any increase in traffic, and the costly edges will grow asynchronously through disseminated and dispersed investment. In this sense there might be some devices that will start using higher frequencies in the tens of THz 20 years from now but I would not consider them as a real shift to a new G, to 7G, as I did not consider 5G some wireless point to point experiments in the years as 5G.
An additional point to take into consideration is the time it takes to define and agree a new infrastructure standard. It is likely that in the future we will need to change some aspects of the standardisation process but that, also, will take time. Industries are looking for standards to enable new business in an effective way but at the same time they are slow in progressing to protect existing investment.