Going forward, it will be more important than ever to produce pioneering and flexible human resources

Fumio Watanabe , Ph.D(Engineering)
Institute of Science Tokyo
Academy of Super Smart Society
Super Smart Society Promotion Consortium
Specially Appointed Specialist / Promotion Advisor

In the evolution of mobile communications, technological advancements drove social needs up to the 4G era, but 5G is presenting a different picture. We ask Dr. Fumio Watanabe, who has long been active at the forefront of the communications technology field and is also involved in student training, about his expectations for a Beyond 5G society, the roadmap for international standardization, and the direction Japan should take.

A shift in direction to meet the needs and wants of social issues and social conditions

Looking at the potential impacts of a Beyond 5G-enabled society, and world, what domains are considered particularly important or highly anticipated?

Watanabe: When we talk about “G-generations” such as “5G” and “Beyond 5G,” we sometimes refer to differences in technology such as transmission speed, low latency, and massive machine type communications, but here I would like to consider it from a different perspective.
Mobile communications, which originally handled person-to-person communication, evolved between 3G and 4G to connect people with servers. This is for uses such as connecting to Internet services. This expanded further to communication between machines and servers, or between machines themselves, giving rise to the “IoT (Internet of Things).” This can be considered a hallmark of 5G. And looking ahead to Beyond 5G, I believe that “AI-to-AI communication” will become the most important domain. In other words, a “Network for AI.”
Since mobile communications began, the goal of being able to communicate with anyone, anytime, anywhere has been achieved. However, I believe that future developments will not be a continuation of this trajectory. Until now, mobile communication technology has led the needs of society. It has progressed based on a provider-driven logic, but I feel that approach is now at a standstill. I believe we are at a stage that demands a qualitative shift to a direction that meets the needs and wants of today’s social challenges and conditions.

The Ministry of Internal Affairs and Communications has stated that Beyond 5G will be realized sometime in the 2030s. What are your thoughts on the current situation?

Watanabe: Up until 5G, advancements have come roughly every 10 years. Goal setting was mainly based on the evolution of wireless transmission system technology. The international standard for 3G was established in November 1999, at which time I was serving as vice-chairman of a working group at an international conference. Around 1990, the communication speed target for 3G was set at 2 Mbit/sec, but the prevailing view at the time was that such high speeds were overkill. However, in reality, the demand for high-speed communication grew rapidly with the spread of 3G. The experience has taught me that if you provide a powerful tool, the use cases will follow. The situation was similar with 4G. When we created the technology with the goal of increasing communication speed 10- to 100-fold, society transformed into one where everything is done on a smartphone. Up to this point, it can be said that development followed a provider-driven logic.
However, the situation is different with 5G. For 5G, too, technology has been developed with goals for transmission speed and low latency. As a result, while applications are emerging in fields such as the remote control of heavy machinery and remote surgery, it has yet to deliver what one would call a definitive 5G experience. Expanding mobile communication coverage requires enormous upfront investment, but the reality is that 5G-specific usage is recognized as being limited, and a clear path to monetizing such huge investment has not been found.
This suggests that when considering Beyond 5G, a provider-driven approach will lead to a dead end. I believe we will not succeed unless we shift our focus from what technology we want to realize in the 2030s to how we want to use technology to serve society, what new industries we want to create, and ultimately, how we want to achieve human happiness. We will need to reverse our traditional way of thinking by envisioning what society should be like in the 2030s and then considering what infrastructure will support it.
For example, it is crucial to create an overall vision, such as setting a goal to “realize a zero-accident traffic system by year XX,” and then planning to “achieve specific milestones by certain years” for that purpose.

Bolstering software

Regarding Beyond 5G, what is your assessment of Japan’s technological standing?

Watanabe: In technology fields such as optical devices and optical transmission, particularly at the research level, I believe Japan is world-class. Based on that, it is crucial to also address the areas where we are weak. Current network systems are predominantly software-based. In the automotive sector, cars were traditionally a collection of mechanical parts, with computers installed to add convenient and safe functions. However, the rise of battery-powered electric vehicles has reversed this dynamic. As they have become mechanically simpler, the concept has shifted to one of simply attaching a car body to a computer.
The same applies to communication networks; in the past, it was common for communication equipment to be different if the frequency was different, but that is no longer the case. Now, even if the wireless transmission method changes, a certain level of adaptation is possible simply by changing the software. What is important in software is not so much programming itself, but rather the operating system (OS) and architecture that support it. Architectural design is key, but I feel that Japan is weak in this area.

The cost of mass-producing a physical object can never be reduced to zero, but software can just be copied, so its cost asymptotically approaches zero. I am concerned that it will be difficult for a hardware-first mindset to contend with those who think this way; it’s not even a competition.
Another critical element is AI. In the future, there is no doubt that wireless transmission, the backbone, fiber transmission, the core network, network operations and maintenance, and service delivery will all be AI-based. Fierce global competition is underway on how to utilize AI in networks—in other words, for “AI for Network.” The transition will be from hardware-based radio to SDR (Software Defined Radio), and then to AI Defined Radio. The same applies to networks and services. In AI development, a key factor is the volume of learning data that can be accumulated and processed, making it a field where financial strength is also critical. It is a test of endurance, so to speak, and countries and companies that lack the stamina will be left behind. While Japan is doing its best, I must admit we are in a difficult position.
The challenges are software and AI. Conversely, if Japan can become stronger in these fields, we can compete on the world stage and gain a strategic advantage.

Will international standardization also take on a different form?

What are your thoughts on the challenges for the international standardization of Beyond 5G?

Watanabe: Globally, the United States prefers “de facto” standards, which are not established through consensus. The philosophy being that if you capture the market, you establish the de facto standard. On the other hand, with its many countries, Europe often favors the “de jure” approach of creating a common good for everyone through a formal standardization body.
In the 2G era, the wireless systems and frequencies in the United States, Europe, and Japan were all different, meaning you couldn’t communicate when traveling abroad despite having a mobile device. Therefore, from 3G onward, it was internationally standardized so that it could be used anywhere in the world.
However, I believe Beyond 5G presents new challenges for international standardization.
First is the concept of “generation.” Technologists tend to view it as a “revolution,” a leap from one generation to the next. It is the mission of researchers and engineers to change society with revolutionary technology that is completely different from the past. Indeed, it can be said that everything up to 4G was a revolution. On the other hand, another name for 4G is “LTE: Long-term evolution.” The intention with 4G was to pursue gradual progress. However, with 5G, development and standardization were once again pursued with the mindset of sparking a revolution and a generational change. This brings us to the difficulty I mentioned earlier: despite the enormous capital investment required for this generational change, it has yet to bring about a major societal shift. I believe that we should be aiming for long-term evolution for Beyond 5G. Many functions of today’s wireless systems can be altered by rewriting the software, so I believe an evolutionary approach is fundamentally more appropriate. We will develop it while ensuring backward compatibility. Of course, I do think a revolution will be necessary at some point...
Another point is that, as I mentioned earlier, the keywords for Beyond 5G are “Network for AI” and “AI for Network.” However, much of AI remains a black box, and transparency cannot be ensured. Therefore, standardizing a network heavily integrated with AI is a difficult challenge. In some cases, standardization may not be possible, leading to a de facto situation where the creator of the superior product prevails.

What are your expectations or concerns regarding industry-academia-government collaborative research in Japan?

Watanabe: I certainly have high expectations for collaboration between industry, academia, and government. I believe international collaborative projects are particularly important.
Furthermore, to produce excellent results, I think the way requirements are described when soliciting project proposals is crucial. It may sound contradictory, but an approach is needed that clarifies the global direction we are aiming for without stifling the creativity of proposers.
Moreover, technology and institutional design must go hand in hand, and I believe collaboration is needed to ensure they develop in sync.

What do you think is key in promoting the research and development of Beyond 5G?

Watanabe: Ultimately, human activity is what drives both R&D and international standardization. Even as AI develops and efficiency increases, the human element will remain decisive. Therefore, I believe it is the role of society and educational institutions to consistently cultivate talent with a broad international perspective and a capacity for pioneering yet flexible thinking.
Recently, we held a HACKATHON (*an event where people from various backgrounds form teams to compete on results by intensively developing ideas and programs on a certain theme in a short period of time) using AI, with a mix of students and business professionals, and they inspired each other. I feel that until now, Japan has not done enough to create environments where students and business professionals can inspire one another. I believe Japan must also put more effort into fostering activities based on an open and inclusive mindset.