neXt Curve attended Interdigital’s first 6G Summit that was a two-day exploration of the technologies and the possibilities that will come after 5G. This event was a great opportunity to compare notes with some of the industry thought leaders and inventors who are ideating through the future of communications with the technology horizon research we did for Ofcom in 2019. Will 6G be a revolution on top of an expected revolution or just 5G cans kicked down the road?
5G is barely out of the gates yet we are already talking about 6G. Is it premature to be talking about the next generation network beyond the next generation network? Verdict, not really. It takes time for the science and technologies that will shape the future of communications to develop and mature to the point where they can be commercialized a decade from now when we can expect the 3GPP (or will it be the 5GPP by then?) to specify the first release of the sixth-generation network technology.
From October 20 to 21 of this year, Interdigital assembled a venerable roster of technologists, academics, regulators, and technology vendors to share their hypotheses on 6G. The presentations were surprisingly thoughtful and very technical. It was refreshing to hear grounded discussions on the problems that 6G might solve by some of the leading minds in the industry. We attended the following sessions:
- Beyond 5G and 6G Initiatives in the United States
- Research, Innovation and Public-Private Industry Collaborations
- The Future of Communications Beyond 5G
- 6G Vision
- Shaping Up 6G: Drivers, Use-Cases and KPI Requirements
- Understanding the 6G Tech Landscape
- The Role of AL/ML in 6G Wireless Systems
- Platforms for Experimental Research: Are we ready for 6G?
You can find the full 6G Symposium agenda here.
After the two days of engaging sessions, it became very clear that we are mired in the 5G challenges of today which may be clouding our vision for what might come after. There was plenty of the convenient-but-impractical use cases presented in the more visionary discussions. This gave the impression that we are in very early days in our thinking about 6G and why it will matter for consumers, enterprises and operators.
Here are our key takeaways from our coverage of Interdigital’s 6G Summit 2020:
Takeaway 1: A lot of 5G cans getting kicked down the road
Maybe it is too early to hypothesize what new applications 6G will enable. By and large, they have yet to be imagined. The early hypotheses for killer 6G applications presented by numerous panelists and keynotes at the 6G Symposium seemed similar to or were exactly the 5G killer applications such as VR, AR, drones and autonomous vehicles that have shaped our expectations of the next generation mobile wireless technology over the past five years.
This is not surprising given that the industry does not have a clear view of what 5G will bring about beyond 3GPP Release 18 other than incremental improvements to URLLC (Ultra-Reliable Low-Latency Communications) and mMTC (Massive Machine Type Communications) to make the technology more suitable for industrial applications. If anything, this revelation highlights what may be a sizable disconnect between what the end market is expecting or is being told 5G will enable and what the scientists and the engineer think can be reasonably delivered over the next ten years.
Many of the prognostications of the challenge areas that will shape 6G technology priorities were well-articulated in Mazin Gilbert’s (AT&T’s VP of Network Analytics & Automation) presentation of key research challenges. They include:
- Energy efficiency
- Spectrum efficiency and sharing
- Adaptive and self-learning networks
- Traffic optimization
- Quality of experience
- Cost of deployment and operation
- Planning of massive network densification
- Network slicing
In terms of new capabilities and technologies, there were frequent mentions of the idea of a “future RAT (Radio Access Technology)” that might be based on new waveform beyond OFDM. There were also notions of technologies that enable new network topologies as well as multi-access integration and continuity of experience. Nokia Bell Labs’ Head of Access Research Labs, Peter Vetter, did a good job of grouping a wide range of scattershot technologies and ideas into what he called the six areas for “6G experimental platforms”:
- AI/ML Air-Interface
- New Spectrum Technologies
- Network as a sensor
- RAN-Core Convergence & Specialization
- Extreme Connectivity
- Security and Trust
While these are pretty broad and general categories, they seem to be good starting points for 6G ideation and technology scanning. We have to start somewhere, right?
Takeaway 2: How do we make autonomous easy and dirt cheap?
There is no argument. Network infrastructure will become more intelligent. That is GSMA’s moniker for 5G, the intelligent network. What comes after the intelligent network? The consensus among the experts at the 6G Symposium seems to be the autonomous network.
In these early discussions about the 6G autonomous network, the core area of focus for the application of AI and ML was the RAN or the concept of an “intelligent air interface” which leverages compute on the network to make better use of spectrum. The idea is driven off of the fact that computation growth is far outpacing bandwidth, and higher frequency bandwidth is not as economically viable to work with.
The opportunity with 6G is to realize an autonomous and adaptive air interface that is able to squeeze out more quality of experience more ubiquitously using more practical and accessible but underutilized spectrum. The hope is that the added intelligence of the RAN will enable new spectrum and resource sharing models that will avail novel approaches to drive spectral efficiencies and scale coverage.
The panel on the session for “The Role of AL/ML in 6G Wireless Systems” outlined a number of key areas of research and development that will be vital for fostering intelligence across the network in the 5G and 6G eras:
- Federated and dynamic learning
- Predictive and proactive networks
- Best effort AI
- Multi-variate optimization of energy, spectrum usage, security, etc.
- Anomaly detection
- Closed loop control system architectures
The era of 5G will be an era in which operators will be evolving their infrastructure up the AI maturity curve. They will have to deal with a mix of old and new technologies that will challenge their ability to broadly apply AI and ML applications across their networks. While data will be prized, its quality will be diverse and its availability spotty. In other words, it will be a big problem and a challenge to get data up to snuff. The data problem could very well be one of the top issues that 6G technologies may need to address before intelligent networks can evolve into autonomous networks.
Once we get close to realizing autonomous network they will generate, learn from and act on what we can only anticipate will be titanic volumes of data; most of it highly perishable. What if the metadata of the network scales faster than the volumes of data that is communicated over it? How might new AI and ML technologies and application architectures address this massive data problem as we enter the 6G era? How do we practically and economically scale the autonomous network? Perhaps these important question will be addressed at the next 6G Symposium.
Takeaway 3: 6G is still afraid of the security elephant
Why is it that security is always at the bottom of any list of priorities in most if not all technology conversations? Is it because it’s so massive a hole to fill that an honest and substantial discussion about it would stymie innovation and invention? Yes, we want uber-resilient, self-aware networks that can operate on unlicensed and licensed spectrum with full autonomy. But what if that network is not secure and not trustworthy?
6G will likely expand upon security concerns and risks that 5G is already introducing as it seeks mindshare and market share in the broad universes of consumer and industrial Internet of Things. This begs the question, has the industry considered that security and trust could be the most important issues of the 5G era and most certainly the 6G era? How do we scale security as quickly as we scale communications and the applications, people and the things they connect?
At best, the topics of security and trust in the 6G era are sizable question marks. They will probably remain cloudy topics until security crises illuminate the nature and scale of the problems we will face with 5G and beyond.
Takeaway 4: Is expanding capacity with terahertz spectrum the next frontier?
5G is about mmWave, so it would seem that the next generation network beyond the next generation network will be about terahertz spectrum, right? Access to this spectrum that ranges from 300 GHz to 2 THz could theoretically deliver 100 times the throughput of 5G. Yet, early 6G research is focused on the D-Band (110 to 170 GHz) sub-terahertz band which resides in the middle of mmWave spectrum range.
Despite all the hype about mmWave that accompanied the run up to the first commercial 5G network deployment two years ago, it seems that the marquee spectrum of 5G will keep the industry busy through the next decade or more as operators and enterprises grapple with the challenges of making it useful. Sure, operators around the world have trialed mmWave and achieved impressive downlink speeds in ideal conditions, but cracking the coverage code is no trivial endeavor today. Our impression is the experts think that this problem will vex the industry well into the foreseeable future.
From a wireless network perspective, terahertz spectrum poses a number of technical challenges that will require huge advances in silicon and favorable economics to make its use practical and meaningful. Terahertz spectrum is even less coverage-friendly than mmWave and is a more difficult spectrum to use for terrestrial applications. Panelists at the 6G Symposium cited a number of daunting challenge areas including the processing power required, sustainability, and testing and measurement for terahertz communications.
As the industry begins to figure out mmWave and addresses the many opportunities to improve spectrum policies across the globe, 6G could very well be about volumetric spectrum efficiency (spectral and spatial optimization) and the technologies that enable it in and below the mmWave range. With an intelligent and autonomous network infrastructure, improved spectrum use of a 6G network could make 5G cans that have been kicked down the road real possibilities.
Implications for Business and Technology Leaders
For most business and technology leaders, 6G concepts are still too squishy to worry about for the moment. 5G still has much further to go before we start to worry about 6G in any meaningful way. In fact, in these early days of 6G ideation, there really isn’t much beyond elevating current 5G concepts to the plaid level. Uber-reliable, incredibly low-latency communications and ridiculously massive MIMO may be extreme but they are also incremental.
As neXt Curve cited in our Ofcom tech horizon research, communications will become about applications that deliver rich cyber physical experiences whether they are through immersive communications, novel interactive media, or a highly tailored business service with privacy secured end to end. It is likely that 6G will be about technologies that address the changing nature of communications and all the challenges that will come along with it, in particular security and trust which the industry continues to treat as inhibitors to achieving the industrial aspirations of 5G and beyond.
Overall, Interdigital’s 6G Symposium was a great start to the 6G conversation. It is apparent that there is still quite a bit of soul searching that needs to happen before innovators and inventors looking at 6G determine a meaningful direction. neXt Curve is looking forward to attending next year’s 6G Symposium to see how 6G thinking gels and surfaces candidate technologies that will define the next, next-generation mobile network.