Finding edge applications on 5G

5G is considered a key part of delivering edge computing applications. But the biggest edge computing application may be 5G networks.

[[280106]]

Discussions about edge computing inevitably involve 5G, the next generation of mobile data networks, which will be deployed in the next few years. 5G will provide higher data rates, and it is expected that edge computing applications will require 5G.

After years of preparation, 5G networks began to appear in April and May 2019. SK Telecom and others are bringing 5G networks to South Korea, Verizon and AT&T have launched in the United States, and EE is launching 5G in the United Kingdom. Networks in other countries and regions are quickly following.

Early testers say the technology is living up to expectations. TechRadar has seen speeds of more than 1Gbps in the US and around 500Mbps in the US, which is significantly faster than 4G networks.

Like previous generations, 5G upgrades the core networks that carriers operate and the radio protocols they use to communicate with end users and devices. This new radio technology, helpfully called 5G NR (for "New Radio"), can deliver speeds of up to 20Gbps, but that speed will be largely a theoretical one—much like the 2.4Gbps top speed that you don't get from today's LTE networks.

5GNR operates in two frequency bands, one below 6GHz (typically 3.5GHz) and the other in the millimeter wave band between 30-300GHz. The lower frequency band is most like the most widely deployed existing networks and offers modest improvements over 4G. The millimeter wave band option enables more exotic deployments and may provide faster speeds at the expense of shorter distances, which will require the deployment of more base stations.

Again, like previous generations of mobile communications technology, this technology will gradually use the "non-independent" mode to share the existing core network.

The media is asking about 5G phones, hotspots, routers, and other tools that humans use. The digital infrastructure industry will want to know: What can 5G do for edge applications like the Internet of Things and self-driving cars?

For IoT, important will be devices that support machine-to-machine communications for IoT sensors, with 75 billion devices expected to be connected by 2025. These standards support low-power wide-area networks (LPWA), including options for NB-IoT (narrowband IoT) and LTE-M (LTE machine-type communications).

But these are a bit of an oddity in the 5G hype, as they were already defined for the previous 4G standard and intentionally use lower speeds than the main benefit to the end user.

The truth is, the IoT may have billions of sensors, but most of them don’t need high bandwidth. They’ll submit temperature readings or other data, and most need long battery life and good coverage. 5G can provide the first, but it’ll be a while before it gets close to the second. For a long time, 4G will be found to be perfectly adequate for the IoT.

Some hope that self-driving cars could be the "killer app" for 5G. Duncan Ellis of Wave2Wave, an IoT company, said it would be too expensive for them to do enough processing to be fully autonomous. Instead, they will need sub-millisecond responses from external pattern recognition to avoid obstacles such as pedestrians.

However, even 5G has the potential for latency and signal loss, so there is a strong push in the autonomous vehicle effort to try to deploy more of what is needed into vehicles, effectively removing the need for 5G.

Paradoxically, the applications that will really put pressure on 5G networks may be those human users that we talked about earlier. They are pumping in the same data as before, but they are processing it faster than ever before, which may change things.

"We can't continue to use centralized data center infrastructure because of constraints such as energy utilization and latency," Ellis said. A user taking 1Gbps of data would effectively utilize the entire backhaul in a cell today.

Fiber optic cables running to cell towers could alleviate that bottleneck, but that approach would be very expensive. For now, networks will have to get better to get data closer to users. “That’s a huge challenge for those networks that are being built because you don’t know where the data is going at all times,” Ellis said.

Finally, the largest edge applications on 5G networks may be the 5G network itself and its users.