This article originally appeared on IEEE Spectrum. by Michael Koziol.
If you listen to the hype about 5G, with its promises of self-driving vehicles and immersive virtual reality, it doesn’t take long to realize how much data the coming generation of wireless will require. But have engineers been so preoccupied with delivering low-latency networks to feed data-hungry applications that they’ve forgotten about the rest of our vast, tangled telecommunications network?
That concern has sparked some researchers to start thinking about where all that data will go after it travels from your phone to the nearest cell tower.
The International Telecommunication Union, an agency of the United Nations that coordinates telecom infrastructure between countries, recently launched a new focus group to, in part, address an emerging imbalance in our wireless communications. The group, Network 2030—more accurately, the ITU-T Focus Group Technologies for Network 2030 (FG NET-2030)—will explore ways to close the growing gap between the fixed and mobile components of future communications networks.
The fixed side and the mobile side are both parts of the unified system that sends information to all of our wireless devices. Even so, Richard Li, the chief scientist of future networks at Huawei and the chairman of the ITU Network 2030 group, sees enough of a distinction to consider them as two separate components. And that distinction is where he sees problems beginning to emerge.
Think of the mobile side as the antennas and radio waves that directly deliver data to our devices. This is the side that has gotten a lot of attention in recent years with the advent of 5G, along with beamforming, massive MIMO, and millimeter waves. The fixed side is everything else—including antennas to beam data wirelessly between two fixed points, and also the cables, fibers, and switches that handle the vast majority of our long-distance communications.
“In the last few years, in 5G, people are working on the radio side. Right now, when people start deploying, it’s the mobile side,” Li says. “But the fixed network side is still 4G. They do not match.” Li is concerned that, with the emphasis on bringing 5G to fruition to deliver gigabit speeds to personal devices, the greater infrastructure has been neglected.
The upshot is that while the larger amounts of data heralded by 5G will zip through edge infrastructure without delay, older, less advanced infrastructure could very well throttle that same data over longer distances.
Li says that Network 2030 isn’t going to play catch-up to 5G. Instead, as the name implies, the group will look beyond that generation and think about what comes next.
That said, he’s prepared to take a broad view of what future generations of communication technology will bring. When asked if the group would think in terms of possible 6G technology developments, Li says he was asked the same question after a recent speech in Canada. “6G is not defined yet,” was his answer both times. “I’m going to act like a wily fox and let other people define it.”
In fact, he doesn’t want the group to think in terms of what 6G might be, but instead what it might require of the network’s backbone. “Fixed networks that will be able to support 6G networks: That’s the key,” he says.
A big part of the problem is that much of how data moves through our networks has been designed to be efficient for the mobile side. In the process, it has become inefficient and prone to clogging on the fixed side.
Take, for example, two people sitting on a couch, both streaming a soccer game to their phones. Li explains that, currently, each device will individually use a GPRS (General Packet Radio Service) Tunneling Protocol to communicate with the network. In essence, GTP establishes a connection, called a tunnel, between each phone and the nearest router. Those tunnels allow each phone to communicate with the router individually, but can create a lot of redundancy when the router sends data further into the infrastructure.
If two people are both sitting on a couch and streaming the same soccer game to their phones, for example, the router currently requests and receives two copies of the same streaming data. “It’s a huge waste, because it just adds tunnel after tunnel on the fixed side,” says Li.
There’s no guarantee that today’s fixed networks can meet the guarantees of 5G, Li says. “It’s now best effort,” he says. The problem is that 5G promises low latency but has little to say on the topic of throughput. Low latency means little if there are so many data packets moving through the network that there are continuous delays.
For example, people feel dizzy using VR goggles if the movement delay when they look at something new is 20 milliseconds or more. 5G promises low enough latency to handle image capturing, framing, transmitting, displaying—everything needed for VR experiences in that time frame—which Li estimates leaves only 5 to 7 ms to transport the data through the network in both directions.
But it won’t matter how quickly the network can usually deliver the data if it’s always swamped by millions of VR users. Most fixed networks simply don’t have a high-enough throughput to push all that data without requiring it to wait at switches and routers like cars caught at red lights during rush hour.
And as some researchers have already started to suggest, 6G will bring applications with even higher throughput requirements. Li says autonomous vehicles, massive machine-type communications, tactile Internet, and holographic communications are all on the table for the coming years. Despite the promises of the mobile side and those building it, the fixed side just can’t withstand the coming surge.
“Right now, we are in a phase of Internet allocation expansion,” Li says. “Transport, health care, manufacturing: Low latency is important in all these cases.” For Li, the only sensible option is to think about these future networks now, to prevent a debilitating data deluge in the years to come.