MITRE's Tech Futures Podcast

Our 6G Future

June 09, 2022 MITRE Season 1 Episode 3
MITRE's Tech Futures Podcast
Our 6G Future
Show Notes Transcript

As the worldwide rollout of 5th Generation cellular networks picks up pace, a team of researchers is already looking towards 6G.

By Brinley Macnamara

Guests: Dr. Curtis Watson, Dr. Kelvin Woods, and Jim Houchens

Brinley Macnamara (host) (00:00):

Welcome to the Tech Futures podcast. I'm your host, Brinley Macnamara. I'm also a networks engineer here at MITRE. And in today's episode, I'm going to tell you about how a group of MITRE engineers are looking beyond 5G to pave the way to our future, hyper-intelligent 6G internet. In this episode, I hope to inform you about how artificial intelligence will enable this future 6G technology and why getting ahead on the development of 6G is more geopolitically important than ever before. But before I begin, I wanted to say a huge thank you to Dr. Kris Rosfjord, the Tech Futures Innovation Area leader. This episode would not have happened without her support. Now, without further ado, I bring you the Tech Futures podcast episode number three.

Dr. Curtis Watson (00:59):

One thing that I don't think will completely exist in 5G that could possibly exist in 6G is immersive, augmented and virtual Reality.

Brinley Macnamara (host) (01:09):

That's Dr. Curtis Watson talking. He's a principal communications engineer in MITRE Labs, and was one of the co-principal investigators on a research project looking into how artificial intelligence will spur the development of 6G cellular networks.

Dr. Curtis Watson (01:23):

The augmented part probably will happen in 5G with personal advertisements as you walk down the street or what not as devices realize that you're a frequent shopper here and there. But, the one thing that I keep on thinking about is, especially right now in COVID times where we're all experiencing remote working, there's been several times where I wished for this virtual reality environment where I could walk into this room, be there with my teammates, and we're actually having a whiteboard session. That would be just like how it was in the office or some other kind of team gathering. And so being able to project and interact with all those virtual representations is a lot of data, high data rates would have to be achieved so that it doesn't seem very herky jerky and realistic feel.

Dr. Curtis Watson (02:10):

So, the thinking is possibly, is there digital twin AI models of every individual that can understand people's traits, quirks, behaviors, so that it feels like you're actually working with that virtual version of Kelvin. And then at that point, all you have to do is load in those models and send actuator functions instead of actually broadcast the whole data transfer. So, like any future predictions, I'm most likely wrong, but that's one thing that I think 6G could provide that 5G probably won't be able to.

Brinley Macnamara (host) (02:49):

6G stands for sixth generation. And, like it sounds, it will be the sixth generation of standards for cellular networks. Like each past generation of cellular technology, 6G will bring with it the immense potential to unlock applications like the immersive, augmented and virtual reality experiences that Dr. Watson is describing that we've yet to see outside of sci-fi movies. Now you might be wondering why are we even talking about 6G when the rollout of 5G has only just begun? It's a question that I certainly had in my mind as I was making this podcast, especially given that we sort of need to see 5G in action to better understand its limitations in order to then make key design decisions about following generations of cellular technology.

Brinley Macnamara (host) (03:34):

That said there are a few core characteristics of 6G that everyone seems to be in agreement upon. The first is a 100 fold increase in data rates. Peak data rates in 5G networks will be a gigabit per second, whereas scientists would like to see 100 gigabit per second speeds in 6G networks and have proposed higher bandwidth methods for transmitting data that will facilitate these blazing fast speeds.

Brinley Macnamara (host) (03:59):

By the way, if you're not familiar with networking jargon, you can think of bandwidth as a number of lanes on a highway and data rates is how fast cars will move on said highway. Hence, with more bandwidth, more data can travel at any given time which translates to faster data rates. The second core characteristic of 6G will be a 100 fold decrease in latency. And you can think of latencies as your driving time on the highway that is a function of the distance you are traveling as well as other factors, such as traffic congestion. Target latencies and 5G networks will hover around a millisecond, whereas latencies in 6G networks could be as low as 10 microseconds. That's about one one-hundredth of a millisecond.

Brinley Macnamara (host) (04:42):

The third is a much higher density of connected devices, which you can think of as the number of cars driving on a highway at any given time. And the fourth and final is the pervasive deployment of artificial intelligence in 6G networks. So all that to say, no matter how wide the 6G highway gets, there will still be practical limits on how fast data can travel across that highway. And this is where pervasive AI will come in as wider highways alone won't be sufficient in handling the gargantuan traffic flows that we expect to see in our future 6G networks. I asked Dr. Kelvin Woods, a chief engineer at MITRE, another investigator looking into 6G, what pervasive AI in 6G networks will look like.

Dr. Kelvin Woods (05:27):

When you think about the definition of pervasive, like spread widely throughout an area or a group, so the thought there is that AI and ML are going to be pervasive throughout all the different components within in 6G. So in 5G you're seeing the startings of some components that have particular AI components, they may not necessarily talk to other components, but in 6G, we're looking for this to kind be woven into the whole infrastructure.

Brinley Macnamara (host) (06:01):

As part of the fabric of 6G, AI models will be deployed and trained across every imaginable device. In the cloud-native AI paradigm of 4G and 5G the vast majority of data captured at the edge is lost due to the bandwidth constraints of sending said data to the cloud. In the edge-native AI paradigm of 6G the vast sums of data collected and analyzed at the edge will provide both the raw material, i.e. the unbiased data that the field of AI desperately needs to generate better predictive models, while at the same time enabling new applications like autonomous driving and immersive, augmented and virtual reality. And this close coupling of AI and 6G will mean that countries who take the lead on the development and deployment of 6G will likely take the lead in the AI race as well. Not surprisingly China, the world's leader in 5G has already started to prioritize 6G.

Dr. Kelvin Woods (06:56):

So previously in the earlier generations, we were major players there, so we want to maintain our seat at the table. Now's the time where we should be starting to dig in and look at how all these components are operating, why we're designing them this way and try to have some seat at the table there because other parts of the world are already taking grave notice of this, and they're starting, they're digging in, they're already digging in. They're already starting to move to this point where they are presenting papers and building and trying to look at things particularly in terms of the standards bodies and how we can shape and have that impact there. So we also have to kind of respond and try to dig in there.

Brinley Macnamara (host) (07:45):

I asked Jim Houchens, the director of MITRE's 5G and mobility platforms about the risks of the U.S. falling behind on 5G and 6G development.

Jim Houchens (07:54):

There's a couple of threats. So one is, there's a belief that because the U.S. rolled out LTE so quickly and got ahead of the world in 4G that, that created the environment that allowed our innovators to flourish and gave us things such as Uber and Lyft and the whole gig economy from the phone. It gave us the movement from the PC to the phone for Facebook and many of these other internet based companies, and really is credited with driving trillions of dollars in GDP for the U.S. Other countries have looked at that and they see that as the path to emulate. So if we're slow rolling out 5G to enough of country, then our innovators and our enterprises aren't as driven to come up with the new use cases and the new kind of applications for 5G until a critical mass can start using these applications. So, one thing right there is we start falling behind in our own innovation. And maybe we won't have the next Facebook or the next Uber or other app like that. It will be from another country that has rolled out 5G first.

Jim Houchens (09:35):

Another area, particularly with China, is the control of upstream. And this value... and this applies to, let's say, other countries, right, as they decide whose network to purchase in their country, China comes in and they'll sell the network as part of a whole package that they give a country. And the fear is that they can start using that network to shape the traffic. So let's say if instead of Amazon, the traffic starts getting shaped to Alibaba and other such Chinese industries as part of their global strategy for driving value for their country, for their companies. And so if China begins to be the 5G and 6G vendor to the world, it's not just about losing your market share of your radio vendors and your people making the core software, but it could really have these upstream effects and impact large segments of our digital economy.

Brinley Macnamara (host) (11:07):

As I chatted with Jim over Teams, I got the sense that he is a visionary. So I was also interested to hear his thoughts on what our 6G future might look like for end users.

Jim Houchens (11:17):

One of the concepts I heard of and I loved it was this idea of mental prosthetics and the use of sensors that you keep with you that record your surroundings, whatever you want. Right? The default is that it records everything around you. And imagine if this ties into artificial intelligence, that's at the edge, almost your own personal cloud that follows you around that not only records these things so you can go back and look at them, but it does sense-making. It knows who you were around. It knows what your conversations were. It can make sense of those things. It could give you reminders that you told someone you were going to see them this week, or you were going to get them something, without you having to type into your to-do list, right?

Jim Houchens (12:21):

Now, it's scary too, right? This is kind of the idea that everything you do might get recorded and other people may have recordings of you in all sorts of situations. It's scary, and there's all sorts of things that need to be worked out to make that happen. But to have that, you just need a network that goes well beyond what 5G is doing, and that natively applies AI and machine learning in 6G, both to optimize the 6G network, but also the network services being provided to AI and ML such as is high performance computing at the edge, right? Distributed all over the place. So as you move around, you have access to that high performance computing, wherever you are.

Brinley Macnamara (host) (13:15):

As we push the theoretical limits of the speed at which information can be transmitted across our vast and heterogeneous cellular networks, we will inevitably run into challenges. While recent experiments in research labs have shown the feasibility of 6G speeds, the deployment of networks that can support 100+ gigabit per second connections over the air and latencies on the order of microseconds into the so-called wild, will be much more challenging. This is because features of our world such as natural barriers, buildings, exponentially growing numbers of connected devices, and even trade agreements will create challenges that are significantly less predictable than the electromagnetic limits on data rates that we can control for in our labs. And while the physical laws that govern data rates cannot be overcome, the pervasive deployment of artificial intelligence in our future cellular networks will provide us with the revolutionary new optimizations that will allow us to achieve 6G speeds and beyond.

Brinley Macnamara (host) (14:19):

In other words, with the help of AI, Jim's vision of personal private clouds fed by 24/7 sensing of our daily activities might not be so far off after all, but this means that there is a lot riding on us setting the tone for 6G. As Jim later noted, the most important thing on the line is whether our country's democratic ideals will be upheld in the internet of the future. We don't have to wait to see the fall out of China getting ahead of us on 5G. From their deployment of AI enabled surveillance systems on their 5G networks, to their launch of the first 5G satellite that will soon bring a Chinese owned 5G network to everyone on earth, China's leadership in 5G means it will be imperative for us to act quickly to enable the U.S. government to lead the world in 6G because, well, the future of the internet depends on it.

Brinley Macnamara (host) (15:23):

The music in this podcast is brought to you by Ooyy, From Now On, Nylonia, Guustavv, and Trevor Kowalski. That's it for today's episode. Thank you so much for listening.