What infrastructure is required for Metaverse?
Dr. Thomas King, CTO of DE-CIX, discussed the challenges infrastructure providers will face and the steps necessary to establish the foundations for the future of the internet.
What will be needed to fuel the immersive, haptic internet? Dr. Thomas King from DE-CIX explored the difficulties infrastructure providers will encounter and the measures required to create the foundation for the internet of tomorrow.
How do we envision the future of the internet? From Star Trek’s Holo Deck to the Matrix, popular culture’s general take on this issue suggests that our digital twins will be “programmed” to mirror the “real” us. It promises an immersive internet comprised of digitally crafted 3D virtual spaces where individuals will engage in an environment where the distinction between reality and virtuality becomes blurred. While this has yet to be realized, progress is being made in many areas toward the internet of tomorrow.
To achieve a genuinely immersive, responsive internet experience, purpose-built end-user technology must be developed and adopted widely, and a digital infrastructure must be established to transport massive data volumes at the lowest possible latency. Over the next few decades, to gradually transform science fiction into reality, we will each need to be equipped with advanced devices, including video and audio equipment.
Our responsive internet experience will be centered around our digital twin’s interactions with the digital twins of others, real-world objects, as well as purely virtual objects and environments. This experience will be driven by virtual reality, but the technology will extend beyond mere headsets.
The responsive internet will necessitate wearable devices capable of transmitting sensations such as pressure, temperature, texture, aroma, and other sensory information. Data streams providing this information will be synchronized with audio and video streams representing three-dimensional spaces. In this context, even the slightest delay in one of the data streams will result in a disjointed experience, diminishing the authenticity and enjoyment of the interaction.
Programming the real world
The high-definition, 3D virtual representation of real-world spaces will rely on multiple video streams, providing a 360-degree virtual view. Wearable sensors and brain-computer interfaces (BCI), whether as embedded chips or surface sensors in headphones, will track our intentions and actions and translate them into motion in the virtual environment. This information will be used to control the video stream sent to each user. Tactile information about physically holding or moving objects, like shaking a partner’s hand or remotely controlling a vehicle, will also need to be integrated seamlessly into the data provided for each user or device.
All these data streams must be synchronized with one another to the millisecond using artificial intelligence to prevent the disorienting feeling of the world spinning and trying to catch up with you when you turn your head in virtual reality. Keep in mind that all these data streams need to be transmitted live, which requires substantial bandwidth and low latency. Cloud gaming and some virtual reality applications can provide an idea of the demand. Even today, a maximum delay of 20 milliseconds is considered acceptable for a good gaming experience.
The more interaction in the scenario, the lower the latency needs to be. For 360-degree 8K cloud VR, this will need to fall below 10 ms. From a bandwidth perspective, cloud gaming requires a solid 35-50 Mbps throughput. For a good virtual reality experience, such as interactive shopping, we can anticipate that 8K resolution data flow at 90 frames per second will require up to 1 Gigabit per second of bandwidth. Note that this is for a single user. If the kids are also online downstairs, you can calculate for yourself how fast the connection will need to be in a few years.
Cyber-physical continuity, symbolizing the seamless transition and interaction between the physical and digital world, is still under development. We saw the beginnings of this with the Nintendo Wii game console. But this entertaining bridge between the physical and virtual worlds is just the start. We already have the experiential foundations in current VR and AR technologies, and each new generation offers new possibilities. Exoskeletons surrounding the user’s body enable us to perform actions in harsh environments via telemanipulation.
Non-fungible tokens (NFTs) offer a mechanism to verify ownership of digital assets, including our digital identity, protecting our digital twin from abuse by third parties. Whether we will have a dedicated room in our home or office acting as a portal to our digitized lives, or whether we will carry it with us as an enhanced form of augmented reality (XR), both scenarios will likely coexist at different stages of technological evolution or for various use cases.
In conclusion, we will be able to shake hands and socialize with people in cyberspace, experience a new level of gaming, and hug our children and wish them good night. We will have to wait and see what else is possible. However, a word of caution: Even current applications are stretching the capacity of today’s hardware and internet access infrastructures, and handling new ones will become more complex.
Difficulty of distances – being as close to the user as possible
It’s natural for the luxury consumer tech front to spend big bucks on the forefront. However, what will make or break the aforementioned cyber future lies behind the scenes: a digital infrastructure that intelligently stores, analyzes, processes, exchanges, multiplexes, synchronizes, and transmits data streams to end-user devices. This needs to be done within a few milliseconds at most. Innovations in responsive virtual reality have extremely low latency needs.
For some usage scenarios, it is necessary to go down to 1 millisecond or even half a millisecond. The pure physics of the speed of light means that latency-based interactions of around one millisecond will only be possible with users and objects in the immediate vicinity (figures will include around 80km, the most direct route and lowest latency interconnection, processing time, and round trip time, assuming 5G and fiber). Thus, the use cases here will be limited to localized tasks such as remotely controlling robots for clean-up work, rescue operations in hazardous environments, or remotely performing highly complex special surgeries at a central hospital in the local area. Interpersonal interactions, entertainment, gaming, and online shopping can tolerate latency of 5 to 15 milliseconds. This makes a distance of about 400 – 1,200 km to the data center acceptable.
So how feasible is this? On a 5G-equipped campus, we can already achieve high bandwidth and low latency of up to 1 millisecond. But a standalone 5G island cannot compare to the size of the digital infrastructure that connects enterprise networks and end-user networks to the clouds, data centers, internet backbone, and the rest of the world. Indeed, as other infrastructure components are further developed to match the performance of next-generation wireless networks, responsive internet will likely disrupt 5G and beyond 5G networks, as well as Wi-Fi 6 and 7.
For this, the digital infrastructure will need to get much, much closer to the user. This is edge computing on the next level (container data centers in every neighborhood and computing boxes in basements), whether it’s 5G or FTTH/B, it will require high-bandwidth last-mile ports and concentrated access to connected infrastructure. This is the extreme that speed can reach.
The internet of the future will require tech neutrality
We still have a long way to go to achieve this level of digital infrastructure. Utilizing both existing and new networking technologies, a densely interconnected infrastructure network will be necessary to support immersive internet use cases. We will require many more data centers than we currently have, as well as fiber and 5G networks serving end-users everywhere. Not just in megacities, but all small cities around the world will need their own scalable interconnect infrastructures to achieve such low-latency, high-bandwidth environments.
At the same time, there is a need to preserve the openness and neutrality of the internet and provide flexibility for users. We’ve already seen the consequences of lacking interoperability and portability in systems. Consider the concerns companies have about being reliant on specific providers when it comes to cloud services. The rapid evolution of the Internet was rooted in its open nature, which laid the foundation for interoperability and standardization. Therefore, we will need various levels of standardization to ensure interoperability, and we will need a diverse and technologically neutral infrastructure environment to maintain openness.
The strength of the immersive responsive internet lies in allowing systems and perceptions to interoperate and assets to flow seamlessly across different domains within the primary infrastructure that constitutes the internet of digital twins. As is often discussed today, there will not be a single “metaverse” in our lives, nor will there be a single cloud. Instead, we will see multiple universes that must have common protocols and be based on shared understandings of data management in order to interact with each other. To make this a reality, the spaces between these universes will need to be interconnected by an infrastructure network as seamless as the spaces within these universes.
The connection mesh to carry the tactile internet
In the early stages of the future internet, all network technologies working together – potentially even 3G mobile and copper cables in some regions – will be needed to achieve the required coverage density. Over time, there will be no choice but to upgrade legacy technologies to those with much higher bandwidth and lower latency to support increasingly advanced applications. Internet service providers will need network caching of content and media ready to serve end users, much like streaming providers do today. The need for data centers much closer to the user will increase to host AI applications and digital twins, as well as to cache larger virtual environments.
To meet the bandwidth demands of new immersive applications, in the long term, we will need high-performance interconnect platforms that keep traffic local and offer 800GE and Terabit Ethernet (TE) connections, not just 400 Gigabit Ethernet (GE) connections every 50-80 km. The next-generation Internet Exchange Points, which will facilitate data exchange at the lowest latency, will be fully automatic, secured with the latest encryption technology, and highly flexible. These interconnection platforms will need to be data center and carrier agnostic to bring together a critical mass of digital infrastructure players to work together to connect different evolving universes. Connecting platforms will draw networks and data centers closer together.
In this way, we can weave together the connective fabric of the future. Intelligent interconnection of devices, data streams, clouds, and data centers at low latency and high bandwidth will become possible. However, if we take 5-15 milliseconds as the standard for general-purpose applications of the immersive internet, there will be limitations to interactions at longer distances as well. The best possible RTT for data to travel halfway around the world is over 130 milliseconds, not to mention the processing time, let alone the latency when communicating with humans on the Moon or Mars. In the coming decades, there will be significant delays in intercontinental or interplanetary communication, and different applications will be needed. But perhaps by then,
