Even though 6G networks aren’t expected to be operational until 2030, planning of use cases, network architecture vision and the technology is underway now.
Worldwide standards bodies such as the International Telecommunications Union (ITU) and the 3GPP are busy with the behind the scenes work of imagining what could be and then developing the standard for the wireless technology that will make 6G real.
While there is nothing concrete about 6G yet, what’s exciting for me is the direction of these early discussions. More network disaggregation and cloud-native technologies to build and run scalable Open vRAN, in modern and dynamic environments such as public, private, and hybrid clouds, to pick up on work we’ve done at Rakuten Symphony. We are also devoting resources to 6G advanced research in Japan, and to Open vRAN innovation at our new lab in India, leading industrialized automation of cloud native mobile networks.
I also care deeply about the security and cloud-native aspects of networks, and I am excited about the potential for much stronger network security. I had the chance to present a keynote at the 1st International Conference on 6G Networking event in Paris in July on how to make 6G cloud-native and secure by design. In fact, many are calling 6G the first “AI-native” network with deeply embedded artificial intelligence / machine learning capabilities that will help to manage complexity and support semantic communication capabilities by design, but 6G wireless will also be critical to better network security and to run inside containers.
6G is expected to inherit all cloud-native architectural and operational principles, including K8s (Kubernetes) lifecycle management, agility, resilience, and observability. In Open vRAN, the disaggregated architecture and containerized approach to run the virtual centralized unit (vCU) and distributed unit (vDU) on commercial off-the-shelf (COTS) hardware – designed to be easily installed and to interoperate with existing system components – brings many benefits to carriers, particularly in terms of the total cost of ownership (TOC), automation and innovation.
What will be new and different about 6G depends on who you talk to and what standards body they are following. The Working Party (WP) for the overall radio system aspects of International Mobile Telecommunications (IMT) systems is the ITU-R WP5D, but the 3GPP is also discussing 5G Advanced/6G as a part of its release process with 6G requirements to be defined as part of Releases 19 and 20 starting in 2024.
Analyst firm STL Partners has put together one of the most comprehensive updates on the status of 6G, titled “6G: Hype Versus Reality,” where analysts list some of the current thinking from various vendors and standards bodies about what will be a part of the 6G standard. This excerpt from the executive summary of the report touches on six categories of current planning:
Finally, there is a discussion on tailoring 6G for use of/deployment by other industries beyond traditional telcos ("verticals"), such as enhancements for sectors including rail, broadcast, agriculture and utilities, among others, which may require specific features for coverage, sector specific protocols or legacy interoperability.
One of the logical additions to the standard that is also under discussion is better support for network disaggregation, containerization and a deeper integration of the network with Open vRAN to maximize the benefits this framework has on the network.
"With a tighter integration of Open vRAN and 6G, the benefits will include improved interoperability, and innovation."
This is impactful because it will bring significant benefits to mobile network operators. In using Rakuten Symphony’s Symworld cloud-native platform to build out Rakuten Mobile’s Japan-wide network, we were able to build the network quickly and at the same time eliminate inefficiencies and introduce new levels of automation that have had a significant impact on network operations and the launch of new features and services. With a tighter integration of Open vRAN and 6G, the benefits will include improved automation, total cost of ownership and innovation.
Even the largest proprietary RAN vendors anticipate that 6G will open the floodgates even further for Open vRAN technology. The main standards body for Open vRAN is the O-RAN Alliance and they have formed a next-generation research group (nGRG) to research the role that an open and intelligent vRAN will play in 6G.
6G networks will need to be AI-native to handle increased complexity, especially around the network air interface, deployment timelines and self-organizing network (SON) requirements to manage signal overlap between cells. AI can also help manage the complexity of placing network functions on the right servers as well as helping to manage 6G services such as quality of service, mobility management and spectrum-sharing among others. AI can also play a role in smart applications that leverage situational information and content awareness and use this information to customize privacy capability. These apps will help meet the diverse, complex, and novel privacy challenges that come with 6G. Beyond that, 6G Wireless is expected to incorporate outer semantic channels, starting precisely from the initial Shannon and Weaver’s categorization, which was inspired by Nikola Tesla — who stated, in 1926: ‘‘When wireless is perfectly applied, the whole Earth will be converted into a huge brain.”
Being AI-native offers a lot of potential for the comprehensive security that is being discussed for 6G. As a baseline, 6G is expected to take the best of 5G security including zero-trust architectures, distributed ledger technology, quantum-resistant cryptography, adversarial machine learning and cyber resiliency through continuous detection and appropriate response to adverse events and the ability to withstand attacks, autonomously evolve and adapt to threats.
But artificial intelligence should play a big role in integrating security at the very core of the network enabling the deployment of a defense-in-depth strategy, augmented by a zero-trust model, that will result in security standards built into every level of the network, where not a single asset is trusted implicitly and continuous access control, authentication and identification are used inside the network. This is a dramatic improvement from the security-enhanced strategy of current networks.
One application for AI-based security is the use of a pair of deep neural AI networks that are trained with differential privacy, a new system that can share information about a dataset but not include any information about the individuals whose data is in that set. This privacy framework makes it much less likely that the network could reveal personally identifiable information about network users.
AI also brings federated learning as a possibility for quickly building global AI-based security models. With federated learning, model training is done using distributed “individual learner” machines that use local or regional data to train a model before sending that model to an aggregator to contribute to the building of a global model. This streamlines the training of security models when large amounts of data are generated locally.
How can MNOs prepare for 6G even with so much of it undecided? By continuing to deploy cloud-native strategies, MNOs will build in the flexibility, service agility and security that will benefit their network operations today and lay the foundation for the next generation. IaaS, PaaS and AIaaS, borrowed from cloud services, could very well coexist as they would cover diversified AI service requirements from very different sectors. AI services that run on this advanced infrastructure will bring many advantages: from global AI to local AI, from offline AI to real-time.