With the fifth generation of mobile internet taking its early steps, a significant number of policymakers, operators, and agencies are concerned about how to deploy 5G and to turn this technology into a useful and profitable resource. Would now be the best time to start discussing the next step towards the future of wireless connectivity?
Data suggests just that. During the Covid-19 pandemic, much user behaviour has changed. According to AT&T, Wi-Fi calling increased by 100% from April to October, while 60% of the traffic moved to video, and mobile voice grew by 40%. The company carries over 400 TB of data per day.
“We are living through unprecedented times,” stated Mazin Gilbert, Vice President of Network Analytics and Automation at AT&T, during the 6G Symposium, an event organised by InterDigital and Northeastern University. “People’s way of living, playing, and being mobile completely changed.”
That is one of the reasons why, in his view, 6G – or whatever the next technology is called – is inevitable. “This is the perfect time for governments, industry, academia to get together and solve some tough, sophisticated, advanced problems we are facing as a society. The future is absolutely exciting and interesting,” he added.
The future Gilbert is talking about is also highly connected, to a degree that exceeds what humans have experienced so far.
There will be around 500 billion connected machines in the world by 2030 according to Sunghyun Choi, Head of Samsung’s Advanced Communications Research Centre. That would be almost 60 times Earth’s population at the time when, according to him, 6G is expected to be a reality.
“At that time, possibly the main users of 6G could be machines along with people. And the fact is that […] machines capabilities can far exceed humans’ limitations. So, 6G should be able to meet the needs of these machines,” Choi explained during the symposium.
So, what does exactly a new network need in order to match those needs? According to experts, there will be at least four areas in which technology needs to advance if 6G is to be realised.
According to Choi, 6G should achieve a much-enhanced performance that allows machines and consumers to experience almost flawless connectivity. Some of the fundamental ones are:
- 1 Tbps (15 times today’s cap)
- User experience data rate 1 Gbps (10 times today’s cap)
- Air latency of 100 microseconds (a tenth of today’s cap)
No Guarantees, No Deal
Without precisely knowing what the network can deliver, and when, working with the potential technologies of the future – autonomous cars or drones, for instance – could be a challenge. That is why Gilbert stressed the need for stringent service-level agreements (SLAs).
“The world we are heading requires SLAs that are guaranteed. To be able to have an autonomous vehicle, without needing a driver driving that vehicle, you expect a guarantee for ultra-low latency, for liability,” the expert said. “Most networks today are built on best-efforts, not guarantees or SLAs. And that is the promise we need to go moving to 6G.”
Machine learning and AI
The amount of data generated by almost nine billion people using over 500 billion connected machines will be massive. That is why specialists agree that machine learning and artificial intelligence will play a crucial role in 6G.
Experiments in a variety of fields are already being undertaken, and it is becoming more and more clear how AI and ML can help easing the creation of 6G, although it is still a work in progress.
John Kaewell, Senior Principal at InterDigital, exemplified that during the 6G Symposium. Sharing the outcomes of a study published by Ahmed Alkhateeb, a professor at Arizona State University, Kaewell explained that the author was able to combine ML and MIMO applications.
“The author demonstrated that they could use a sub-6GHz pilot that would set from a user equipment in a field to select or predict the best set of MIMO waves for millimetre Waves transmission from the base station back to there,” he said.
Another area where AI will greatly impact 6G is spectrum sharing. According to FCC chairman Ajit Pai, these technologies are also crucial enablers for operators to deliver a better service.
“Monitoring how people use the network at different times, different places, different intensities, all that is generating data. Instead of flying blind, if you will, network operators are very interested in incorporating AI so that they can optimise the network. Ultimately, it is all about how do we serve customers better. Instead of saying ‘ok, here are the three options we think are the best for you’ you are learning from the data you have got,” Pai said during the conference.
Pai also said AI and ML will probably assist the FCC in the future, as the operations and regulation of networks will become more decentralised as a result.
Opening and disaggregating
One of the main aspects of future networks is bringing cloud processing closer to the edge, that is, to the end-user. Instead of traveling a long distance before being processed, the data would be dealt with close by, which decreases latency. The animation below, presented in the 6G Symposium by Larry Peterson, CTO of ONF, explains the difference between 4G’s process and the disaggregated model that is being adopted in 5G and will be paramount for 6G:
“This disaggregation is going to take us to an important place, I think,” Peterson said about the several outcomes that come along with it, like the proliferation of an open-source mindset and the cloud moving closer to the user.
“This will have a transformative effect on how the technology evolves and changes over time, it’s going to be a much, much faster pace. That will have a fundamental impact on what becomes 6G.”
Journalist since eight years old, when I would read the newspaper out loud and pretend it was a radio show. Based in São Paulo, I have worked for Brazilian websites as reporter and editor before joining 6GWorld