alexskopje - Fotolia

IOWN explained: Everything you need to know

In this essential guide, Computer Weekly investigates the Innovative Optical and Wireless Network project, and why the technology it’s aiming to produce could form the future basis of networking

As political events over the past 10 years or so have taught us, there are very few certainties in life, yet to death and taxes we can add increased data consumption.

In addition to what would have been the “normal” evolution of the usage of connected devices as fixed and wireless communications infrastructures proliferate around the world, artificial intelligence (AI) and large language model (LLM) use cases are set to drive an almost exponentially rising demand for data. There will also be a commensurate rise in energy consumption due to the vast amounts of compute power required by future applications.

The current pace of development of “traditional” networking technology looks unlikely to be able to address such rising data demand and the related rise in energy consumption. With that in mind, IT giant NTT has launched the Innovative Optical and Wireless Network (IOWN) initiative, which has the core mission of “turning science fiction into science fact”, showing how speed-of-light technologies will improve daily life, work and society at large. The overarching emphasis is to address a world that is moving quickly to optics, a world that is moving to replacing “electrons with protons”.

From a technological basis, IOWN comprises three major technical fields: an all-photonics network (APN), digital twin computing and a cognitive foundation. The networks and information processing infrastructure including terminals are intended to provide high-speed, high-capacity communication using technology focused on optics, as well as large computational resources.

Technologically, IOWN comprises three major domains: an APN, digital twin computing and a cognitive foundation. The APN’s photonics and optics-based technologies are designed to achieve three performance targets: ultra-high capacity with data processing of 125 times greater than networks today by volume; ultra-low latency offering near-instant transmissions with end-to-end latency reduced by over 200 times; and ultra-low power consumption with a goal of 100 times more efficiency than current transmissions, reducing carbon emissions by 45%.

The digital twin computing element is an extension of conventional digital twins that uses photonics-based computational capabilities to perform calculations on virtual models of objects and humans to accurately make predictions about the future. The Cognitive Foundation forms an optimised, autonomous control of tech resources, including cloud, edge, networks and terminals based on AI and machine learning.

In a 2023 report looking at the prospective emerging candidate technologies that would underpin IOWN, NTT highlighted photonic-ready, distributed datacentres and IOWN for AI and LLMs. It also showed the initiative’s goal of 100 times more efficient transmissions. Additionally, it mapped out the future deployment of APN technologies, whose low-latency capabilities will enable a shift from centralised datacentres to distributed datacentre “stations”.

The datacentre work is regarded as necessary through the rising data demand and the related rise in energy consumption, expected to exceed 3,000TWh worldwide by 2030. In the domain of AI and LLMs, NTT noted that these areas require massive amounts of compute power, calculating the energy required to train an LLM once is equivalent to that needed to power a nuclear plant for an hour. IOWN’s ultra-high-capacity goals look to address these issues.

Additionally, NTT announced its own proprietary LLM, Tsuzumi, an ultra-lightweight model to reduce learning and inference costs.

Which companies are driving IOWN?

In its mission, NTT is being supported by the likes of Ericsson, Nokia, Sony, Ciena, Intel, Nvidia, Microsoft, Orange, Telefónica, Google and other leading technology firms to create a next-generation global communications infrastructure capable of enabling high-speed, high-capacity internet services utilising photonics-based technologies.

NTT has also committed to promoting the realisation of a new communications platform consisting of the APN, which includes silicon photonics, edge computing and wireless distributed computing, through the development of new technologies, frameworks, technical specifications and designs. Testing of NTT’s APN in datacentres began in Tokyo in 2023, in partnership with Amazon Web Services (AWS).

To realise the benefits of IOWN, NTT said it was necessary to craft a “concrete” technology roadmap and promote technological development. Once drawn, the company said that would accelerate the incorporation of data-centric and disaggregated computing technology into IOWN. It was also deemed essential to gain buy-in for its plans from the wider technology ecosystem.

To this end, NTT and other key partners have established the IOWN Global Forum (IOWN GF). In explaining its mission, IOWN GF is keen to emphasise that it’s not just “any” technology groups and technology projects forum. It stresses that among the reasons why it’s different from other groups of its kind is that fundamentally it is real deployment and use case-oriented, and goes direct to users with the clear aim of bringing optical networking – and the domain of computing – to the masses.

In April 2024, the forum convened in Vancouver for its first-ever public meeting to reveal not just the progress made in the project, and the likely technological advances, but also how the ecosystem is growing and what IOWN means to the key stakeholders.

Among the insights offered into the mission and work of IOWN GF, the sessions – attended by around 200 delegates from over 140 companies – showed the fundamental vision of the comms world in 2030; how to prepare for a more secure and sustainable world, powered by light; how open source collaboration would be key to driving the future of communications infrastructures; and the possibilities of how IOWN technologies will likely emerge and what they could achieve in this timeframe.

Gonzalo Camarillo, head of implementation components at Ericsson and IOWN GF marketing steering committee chair, said a key message the forum needs to drive to the industry is that IOWN is “not selling smoke” to a communications ecosystem that is well used to alliances and trade associations with regular advances in technology. Instead, he insists that IOWN GF will show how to create value from the technologies that are being developed and take advantage of the sheer diversity of the membership when looking at the three pillars of mission, vision and focus.

In addition to the entities that develop and deploy the technology, the consortium also has people doing applied research.

What are the likely initial use cases for IOWN?

IOWN said its focus will be applying the three pillars of operation with those that are going to be using technologies developed under the IOWN umbrella.

Right now, the focus of attention is, not surprisingly, centred on growing AI use cases for IOWN-based networking. These include AI-based diagnostics, low-latency remote surgeries and the further development of bio-digital twins. Focus use cases with AI will also involve graphics processing units and how to handle the expected massive data growth.

As regards current industry verticals, IOWN is in contact with companies from the banking and finance industries; broadcasting, TV and entertainment; the telecoms industry; and automotive.

In an example given by Camarillo, a bank could be using an IOWN network for their communications that a car manufacturer could also use. The next-generation infrastructure is supporting its associated connected multi-disciplinary manufacturing ecosystem. Such companies will before too long, he said, be looking at real use cases as they put together requirements specifications.

“They will say, ‘Look, I need to connect these two datacentres’, or, ‘I need to perform these financial transactions and these are the issues I’m facing, and this is kind of what I would like to see’,” he said. “With those use cases, we can analyse them or take out and extract the requirements. And then we can start working on the technology that would address those requirements. Once we have the requirements, [with] blueprints and architectures and all that, we can move onto proof of concepts to prove that the technologies we are developing make sense. And after that, we move into early deployment.”

Where are the key deployment areas for IOWN products and services?

While IOWN is a truly global project, the current leading players in the ecosystem are mostly Asian. Looking at what it’s trying to achieve in the face of the ever-increasing data explosion, there is particular prominence in Japan where AI adoption is seeing data growing. IOWN strongly believes this presents challenges that cannot be shirked, as AI is seen as vital for the future prospects of the country’s economy as a whole.

Indeed, it adds that data flowing through the internet is growing exponentially in Tokyo mobile networks that are handling double the amount they did before Covid. There is a key macroeconomic element about Japan and AI. It’s seen as a solution to skills crises, meaning AI is driving the massive consumption of data.

When are products and services expected to be rolled out?

IOWN has an official vision for 2030, but the deployment timeline starts next in 2025 when early deployments are expected. While likely to be very complementary, the development roadmap of IOWN and that of the forthcoming 6G ecosystem are really coincidental, even though both promise to be as transformational and create smarter societies. 

Read more on Telecoms networks and broadband communications