Challenges to Wireless Networks as the Metaverse Becomes Reality

Interactions between the real and virtual worlds are getting closer. But what kind of wireless network will the metaverse need? What challenges will mobile operators face with wireless network operations when the metaverse becomes a reality?

The metaverse redefines network capabilities

The design of products and services for the metaverse alongside the user experience they provide will depend on network capabilities and constraints. Bandwidth, latency, and reliability are three key indicators of network capabilities.

Key indicator 1: Bandwidth

When we look at bandwidth, we must understand the amount of data that the metaverse will require. Serving as a glimpse into this new world, Microsoft Flight Simulator is the most realistic and comprehensive flight simulator ever. It includes 2 trillion individually rendered trees, 1.5 billion buildings, and almost all the roads, mountains, cities, and airports around the world. All these life-like objects are based on high-quality scans of real objects, requiring more than 2.5 petabytes (2,500,000 gigabytes) of data to make this possible.

Local storage of such a massive amount of data is clearly unfeasible, as no consumer device (and very little enterprise equipment) has sufficient capacity. Providers of metaverse-related products and services can reduce their dependence on transmission bandwidth through pre-downloading and transmitting variable parameters and the adaptive adjustment of video coding schemes. However, interactions in a large real-time, shared, and persistent virtual environment will inevitably generate massive cloud data streams and require bandwidth of more than 10 Gbit/s.

Figure 1: Microsoft Flight Simulator

Key indicator 2: Latency

Requirements for latency may be even more stringent than for bandwidth. Latency requirements range from 400 ms for long videos to about 200 ms for short videos and less than 100 ms for real-time gaming. The more interactive an application is, the more stringent the requirements on latency.

According to the global network platform Subspace, each extra 10 milliseconds of latency decreases the time gamers spend playing each week by an average of 6%. Immersive AAA multiplayer online games probably have the closest latency requirements to the metaverse: Gamers will be frustrated by 50-ms latency, and even non-gamers will find 110-ms latency unsatisfying.

Real data better illustrates this issue: The straight-line distance from Beijing to New York is about 11,000 kilometers. Even if optical fiber is used end to end, the one-way latency would be about 50 ms. As higher latency is expected if copper or coaxial cable is used in some parts of the line, shortening latency in backbone transmission is a huge challenge. Latency in the wireless network also needs to be minimized, as this is the last mile for network access. 5G is already a great improvement over 4G, but we still expect millisecond-level latency in 5.5G or 6G.

Key indicator 3: Reliability

The feasibility of virtual labor and education depends directly on service reliability. The above issues may seem alarmist to people who "live online", as current video applications can already run smoothly in 1080p or 4K resolution. However, this will not be true for the metaverse or for gaming.

Non-live streaming video services like Netflix receive video source files hours or even months before they are available to viewers. This allows them to conduct extensive analysis to determine what information can be discarded when they compress a file to reduce its size. For example, Netflix's algorithms can "look" at a scene with a blue sky and then reduce 500 kinds of blue to 200, 50, or fewer if the viewer's access speed drops. Streaming media analysis can even help determine which parts of a story viewers can tolerate with a higher compression ratio and which can be done through multi-channel encoding. Netflix also uses idle bandwidth to send videos to users' devices before they need it. This ensures an uncompromised end user experience when there is a sudden drop in access quality or increase in latency.

Netflix also pre-downloads content to local nodes, so when viewers watch a new episode of a TV series, the files to be loaded are probably stored just a few blocks away. This is not feasible for video or data created in real time, which is why it is more difficult to efficiently transfer 1 GB of live video than an on-demand video of the same size.

Therefore, assuming that its network requirements are as high as or higher than AAA online games, the metaverse will clearly have higher service reliability requirements. If a device cannot instantly receive all the information needed to run the metaverse, user experience cannot be guaranteed no matter how powerful local computing is.

Is a simplified 6G network coming?

Based on our vision of the metaverse and available reference applications, wireless networks need to provide over 10 Gbit/s bandwidth, millisecond-level latency, and reliability not lower than that required by AAA online games. What does this mean for operators' network deployment and operations?

Figure 2: 10 Gbit/s experience achieved with sub-100 GHz bands

Higher rates naturally require more frequency bands. Using 5G technology, we can achieve a rate of 10 Gbit/s, but we need to use all available bands, including 6 GHz and millimeter wave.

This means that operators need to operate three times more frequency bands in the 5G era than in the 4G era. Macro-micro coordination and inter-site coordination are necessary to make the most of these bands and avoid high penetration loss at high frequencies. Coleago Consulting estimates that there must be 30 to 100 times more micro sites than macro sites to deliver a true 5G experience. A network of such scale may take 10 to 20 years to build, and the difficulty of optimizing this network and the cost of operating it will significantly increase. Therefore, providing a high-quality, low-cost, and easy-to-operate wireless network will be critical for mobile operators to facilitate the metaverse.

Current 6G standards working groups all aim to solve the above issue with the vision of enabling digital twin and metaverse-related applications. The IMT-2030 (6G) Promotion Group believes that 6G network architecture should be designed to be intelligent and simple. Through innovative network architecture design, networks need to transform from centralized to distributed, from heavy-duty and incremental to simple and integrated, from add-on to intrinsic, and from terrestrial access to ubiquitous access. These transformations will enable new services and scenarios for digital twins.

For network governance, a distributed management mechanism based on intelligent autonomy and digital-twin technology can enable the self-organization and self-evolution of a 6G network.

Taking a similar approach, China Mobile puts simplicity at the core of 6G radio access network (RAN) design. China Mobile believes that costs and energy consumption are bottlenecks in wireless network development. These problems have already emerged during 5G network construction and will become increasingly prominent in the 6G era. For 6G, flexible and dynamic networks should be considered, and enabling technologies supporting plug-and-play should be explored. Simplified 6G networks will be characterized by on-demand deployment, green and all-scenario 3D networks, and a high level of network autonomy.

The metaverse opens a door of interactions between the physical and virtual worlds, while posing great challenges to current network technologies, design philosophy, and operation approaches. We still have a long way to go, but our vision of the metaverse will lead us to a place where the real and virtual worlds meet.