The road to 6G

An overview

Uwe Herzog

5G networks are currently being deployed around the globe. Commercial marketing towards consumers mainly emphasizes unprecedented bitrates, and vertical industries look into the potential that 5G could offer to their type of business. While there is more in 5G than that, and even though more will still be included with the 5G enhancements that are currently being standardised with 3GPP Release 17, thinking about what to address in a further generation of mobile communications has started in many regions of the world.

When I talk to people that are not in IT and I mention that I am in mobile telecommunications research, I often see them surprised: to them everything seems to be there, and maybe most noticeable to consumers are the improvements in handsets that are pushed into the market at short intervals. They do not see the huge innovation and investments that are needed in the mobile networks behind the handsets. The question of what will be in 6G as yet another generation is not answered yet. It needs careful consideration given the huge efforts and investment that can be expected from its introduction.

What 6G will add to 5G

Thinking about what 6G will be, I remember a presentation by a well-known colleague from TU Dresden, I think it was at EuCNC 2019 in Valencia, where he said – probably with a wink of the eye – that 6G will be there to fix the issues in 5G. While there is certainly some truth in that, further innovations will be part of 6G. Many initiatives have been launched in Europe, the US, China, Japan and other regions, marking the start of a 6G vision phase.

Several white papers were already published that look into promising technological advances and innovations that are potential candidates for adoption in 6G. Which of these will finally make it into the standard is difficult to forecast, but one can certainly expect that with 6G we will see the usual increments, e.g. in bitrates and latency. Given the trend to machine-centric communications, 6G will also need to be able to handle a much larger number of communication endpoints, i.e. humans and things, and to process the increasing amount of data, in absolute terms but also per area. Various technologies can contribute to that, but further expanding mm-wave technology and exploiting it, which is happening only slowly in 5G at the moment, might be an essential element to that end. From the network design view, efforts will continue to make 6G more dynamic and flexible, and, e.g., cloudification and disaggregation are ingredients for that. In addition, open APIs – think, e.g., of Open RAN – will likely play a role. A technology that is expected to receive much attention for 6G is Artificial Intelligence (AI). With further improvements of the technology, an increasing amount of available data and improved training methods, AI chipsets, and more, it could become a commodity that it is currently not yet, and can enable improvements in many aspects and elements of a 6G network, e.g., to make network management and operation more efficient and more secure.

Cost and energy consumption

A strong and overarching design principle for 6G could be to keep cost under control and to reduce energy consumption. Cost has been a recurring aspect in particular for telecom operators in a low-margin business environment, and lower cost is also important for services to be affordable for end users. Energy consumption and the ecological footprint in general have received increasing attention over the past years in which ICT has however continued to increase its share in the global carbon emissions. In view of the ongoing global warming discussions and the increasing sense of urgency that can be witnessed very strongly here in Europe and elsewhere, this topic might still get a relevance for telecom network operations that is not expected today.


One theme that I see at risk of not receiving proper attention is network and service security. The theft of 50 million user data at T-Mobile US, which got known in August 2021, has underlined once more how critical it is to have proper security measures in place. The increasing number of hacking attacks that can be observed make it clear that this is not a problem that will soon fade away. Implementing proper security has a cost, but like an insurance, it does not have an immediately visible benefit nor can it create corresponding additional revenue.

From a pure cost aspect, one might argue that it is cheaper to fix an issue rather than implementing the required security mechanisms, but the potential damage might go beyond lost revenue. Possibly, strict requirements on trust and security from vertical sector users can cause a shift here. However, it will also remain a challenge for security research to come up with innovative solutions that are cost-efficient yet effective, offering a high degree of automation in line with ETSI Zero Touch Management or through the use of AI / Machine Learning technologies, that enable a low threshold for the introduction and operation of security solutions.