Distributed ICT Infrastructure to Foster a Constructive Explosion of Data

Major telecommunications providers have experienced a number of large-scale network failures over the past several years. These outages disrupted public life by rendering voice communications, and even electronic payment and ticketing systems, unusable. The effects reached a broad range of economic sectors from banking and finance to transport, parcel delivery, and medical care. Communications infrastructure, like that for energy, is a foundation upon which other infrastructure depends, so when it ceases to work, disastrous consequences can ripple throughout society.

And it’s not just communications. Japanese enterprises also have seen their business activities paralyzed by disruptions to the services of GAFAM and other Big Tech platforms based overseas. Thus when we postulate that failures and outages will occur, the resilience of these systems becomes an issue that must be addressed if the maintenance of functions critical to society are going to depend on a limited number of specific vendors.

Looking back on historical precedent, the building of the internet was significant because it provided a way to establish communications between any two points while ensuring resilience: Servers were distributed geographically and redundant transmission routes were maintained between them to ensure a robust network that was highly resistant to disruption even in the midst of crises.

Then during the internet’s subsequent commercialization phase, Big Tech began its ascent as the focus of value shifted to gathering and analyzing data along with the rise of adtech*. Already with a hegemonic position over services and operating systems, Big Tech companies stepped up their involvement in cloud services and other similarly crucial aspects of ICT infrastructure. The digital industry’s concentration in the hands of a few substantial players erased the early internet’s inherent advantages—like its decentralized nature and the spirit of collaboration among its users—resulting in weakening of the net’s overall resilience^†.

This overconcentration of not only communications infrastructure^‡, which is subject of tight regulation to ensure reliability and security, but also the lightly regulated services layer as well, resulted in progressive weakening of resilience, a critical risk for digital society. And this risk will be even greater when technologies like autonomous cars and transport, remote surgery, and remote monitoring and control of vital infrastructure take hold.

All this points to the urgency of making ICT infrastructure more resilient to ensure our digital society’s viability.

Another critical issue standing shoulder-to-shoulder with resilience is a lack of autonomy. What I mean by autonomy here is the will and the ability to innovate on one’s own when applying digital tech to resolve issues—rather than simply looking to existing services for solutions or waiting for others to come up with them.

Since ceding the ICT leadership it had back in the 1990s, Japan has had to settle for a situation that could be termed digital defeat, a defeat that can be attributed to a stagnation of efficiency because businesses dragged their feet in pursuing DX*. The exodus of consumers to digital services continues to speed up^†, yet companies have failed to innovate their work and address their issues. This points to a lack of autonomy in making use of digital technologies.

In looking back to the 1990s, autonomy was one reason for the rapid adoption of web technologies. People were excited to share what they had to say or the systems they’d developed with the world. Users vied with one another to exercise the autonomy this decentralized and completely open platform gave them. But as time passed, Big Tech emerged and in the process reduced the room for users to exercise their autonomy as before.

Many creators today might seem to enjoy a high degree of autonomy in expressing themselves via social media. However, even YouTubers have to work within the business framework determined by Google. Big Tech has complete control over whether creators flourish or perish, including the rules they have to abide by, how and how much they can profit, and what happens to and with their data.

Similar events are taking place in just about all domains of digital society, and in many instances consumers and businesses find themselves forced to relinquish their autonomy in exchange for convenience. Thus the greater dependence on Big Tech, the less innovation will occur from creative workarounds and constructive competition. And if this situation continues, Japan’s digital defeat could well become a long-term affliction.

The digital society should be prefaced on a decentralized model of growth with three essential elements: comfort, fairness, and growth.

Anyone and everyone should be able to enjoy the fruits of digitalization with their security and autonomy guaranteed (Figure 1). Assuming that comfort and growth will come naturally with digital transformation, the greatest benefits can be ensured by focusing on the elements that contribute to fairness like accessibility, autonomy, and security. Things like Web3* and the government’s Vision for a Digital Garden City Nation^† are signs of changes founded on similar values. The key to achieving these lies in restoring resilience and autonomy among internet users, possible through decentralized growth.

Centralization and decentralization each have their respective advantages and disadvantages. Generally speaking, centralized models are conducive to efficiencies like economies of scale but unconducive to diversity, resilience, and autonomy. Society seeks for an optimal balance of the two.

ICT infrastructure of late has veered toward centralization, resulting in less resilience and room for autonomy; meanwhile, much valuable data remains scattered and unutilized throughout many industries. We propose decentralized growth to address these issues and establish an optimal balance of centralization with decentralization and coordination.

Decentralized growth needs to happen in all the layers that make up ICT infrastructure: telecommunications, data, and services. However, each layer needs to be decentralized in a different way.

At the data layer, the establishment of decentralized governance to ensure data integrity will be essential for breaking free of the structural dependency on Big Tech. At the services layer, it is crucial to ensure that local communities and individual users are able to exercise autonomy; this entails a wide range of issues for discussion, among them education, behavior change, and technological innovation*.

At the communications layer, Japan’s infrastructure underpinning data utilization is among the best in the world, exemplified by fiber broadband availability (FTTH household coverage) exceeding 99 percent. But in the ICT domain, today’s winners are not necessarily tomorrow’s*. Thus to maintain its edge, Japan urgently needs to enhance resilience at the national level and sovereignty over data at the local and user levels by decentralizing data and ensuring sufficient infrastructure for accommodating a coming explosion of data traffic.

We developed a data-explosion model to analyze quantitatively how ICT infrastructure should work in accommodating the of Beyond 5G era and data explosion. We selected over 100 typical use-case scenarios and estimated usage rates and data volumes generated in each.

In 2040, about when adoption of 6G ought to be just about complete, potential demand for infrastructure capacity created by data traffic is estimated to be 309 times the 2020 level, the equivalent of about 33 percent average annual growth. Meanwhile, in our scenario foreseeing infrastructure-supply bottlenecks and a failure to cultivate promising use cases, data traffic would generate demand 38 times that of 2020, the equivalent of 20 percent average annual growth.

For Japan to become a world leader in the utilization of data, it needs to promote the wholesome growth of use cases by moving early to provide ICT infrastructure capable of accommodating a 300-fold increase over today’s data traffic.

To increase by several hundred times the bandwidth of wireless networks in particular, it is critical that Japan secure investment in networks, including base stations capable of handling high volume data traffic, in addition to technical development to utilize high-frequency bandwidth more effectively. Abroad, underinvestment in the US, UK, and other developed countries has resulted in stagnation of broadband deployment. For Japan to secure investment leeway while keeping communications costs low for consumers, one idea would be to build a framework for having a broad swath of data-user business sectors share the investment burden as well as looking into policies for effectively deploying private-sector 5G and 6G networks on the local level.

But volume is not the only thing about this data that will be different. By 2040 most data being exchanged over networks will have been generated in the metaverse and by applications like self-driving motor vehicles. To secure high instantaneousness, thus low latency, and enhance resilience in disaster-struck localities, the exchange, storage, and use of all local data will have to take place in within the localities where it is generated rather than having any of it “shipped in” from Big Tech cloud centers concentrated in places like Tokyo and Osaka.

An additional, huge constraint is power consumption. Estimates drawing on past trends suggest that Japan’s ICT domain will in 20 years achieve a 38-fold increase in energy efficiency^†; using this figure, a simple calculation says that power consumption will increase about eight times its current level assuming that data traffic explodes by 300-fold. This means that calls are likely to grow stronger for dynamically moving CPU-intensive data processing, such as that for machining learning and data mining, between locations with renewable energy surpluses. Focus will also turn to the development of technologies—like photonics-electronics convergence technology—for dramatically improving energy efficiency^‡.

Society will take a turn toward local generation and local consumption of data if demands continue for low latency, resilience, and energy consumption in the use of data, and further trends toward ensuring clarity of local data sovereignty with regard to medical and educational records and other critical data (Figure 2).