Do we need another G?

Now that the mobile industry is well into the implementation of 5G networks, research is underway and early definitions have started to appear, attempting to scope and define 6G. At the heart of this is a big assumption: We need another Generational change.

So, what is a mobile Generation?¹ There doesn’t appear to be an agreed definition, the term being loosely used to mean ‘a change in the nature of the system, speed, technology, frequency, data capacity, latency etc. Each generation has standards, different capacities, new techniques and new features which differentiate it to the previous one’.

There is a key point missing from this definition, a Generation is a non-backwardly compatible change. To date, each Generation required a new system; dedicated spectrum, radio equipment, new UE’s and core functions. Operators had to make a step change in capital investment to roll out a new network before demand for the new capabilities emerged. Often with a race to be first.

This approach, used for 2G,3G,4G & 5G networks has served the industry well. However, we are still in the early days of 5G with the applications and services which will use the new capabilities still nascent. Time will tell.

Another assumption is that a new Generation comes along every 10 years. So, we are looking at 2030 for 6G. Much work has already gone into predicting the societal needs and hence technical capabilities that may be required in 2030 and beyond. A key reference is the white paper produced by 5GIC at the University of ^Surrey2. Much less work has been done on the constraints and technical solutions required to implement a new Generation. There are several predictions on the 2030 industry landscape which I believe pass the red face test.

I think it is safe to predict the demand for network capacity will continue its rapid growth and that the expectation for ever decreasing price will continue. Operators will have deployed all available sub 6GHz spectrum to deliver wide area services both in macro and small cells. I’d also predict that mmWave and higher frequencies will continue to only be economically viable in specialist use cases. An operator’s appetite to invest in a new network before real demand emerges may be very limited in 2030.

The way networks are built is going through a revolution driven by virtualisation, disaggregation, open architecture and the application of AI. By 2030, networks will look like a set of distributed compute resources. It would seem odd that the mobile industry sticks to a generational change development model when the infrastructure is at the forefront of the IT industry. It is also a safe prediction that coverage, capacity and energy use will remain key drivers. As wireless networks continue to become national critical infrastructure resilience, reliability and security will also become more prominent.

So, in 2030 radio resources will be fully utilised & service continuity will be paramount. Given this, how does an operator even deploy another non-backwardly compatible Generational change?

In 5G standards, implementation into a non-green field network did have some consideration. Examples are 5G non-standalone³ which provided a mechanism to migrate from a 4G core network and DSS⁴ did provided a method to allocate subcarriers to 5G. 3GPP claim forward and backward compatibility, this has been very limited⁵.

My view is that we should question whether a new non-backwardly compatible generation is required. It may be some time before a compelling reason emerges. Maybe the mobile industry could take some learning from the way that Wi-Fi standards have approached the problem. For example, to date there has been no need to re-farm spectrum to introduce a new Wi-Fi standard.

If a new Generation is required, then equal effort into research, development and standardisation should go into both in 6G capabilities and how it will be implemented.

Eugene Lewis Fordsworthe later changed his view, recognising that sometimes good comes from assumptions. He said that sometimes when there is not enough information a person must make assumptions to make progress. 6G is useful shorthand to define a North Star goal and drive research and development. I think we have enough information to support the idea that we should avoid a non-backwardly compatible G, but if it a compelling need does arise, research, development and standardisation should focus on how it can be economically implemented.

P.S. What ever happened to LTE, Long Term Evolution?

⁽¹⁾ One of the few definitions of a Mobile Generation I’ve found https://ijmter.com/papers/volume-2/issue-10/evolution-of-mobile-generation-technology-1g-to-5g-and-review-of-5g.pdf

⁽²⁾ University of Surrey 6G Wireless: A New Strategic Vision https://www.surrey.ac.uk/sites/default/files/2020-11/6g-wireless-a-new-strategic-vision-paper.pdf

⁽³⁾ 5G- SA architecture https://en.wikipedia.org/wiki/5G_NR

⁽⁴⁾ Dynamic Spectrum Sharing https://www.3gpp.org/dss

⁽⁵⁾ 3GPP’s generational approach https://www.3gpp.org/about-3gpp