5G Taipei Summit: 5 big takeaways for the enterprise

Автор: | 11.06.2018

With the publication of the 5G NR standard, the sense of urgency to begin deployment was the driving message at the fifth 5G Taipei Summit, held in early June as part of Computex 2018. While the name implies the summit focuses exclusively on Taiwan, this year’s summit was held in conjunction with European Innovation Week, as a collaboration between the European Commission and Taiwanese government agencies—with the issues presented applicable across the world.

Here are TechRepublic’s five big takeaways from the 2018 5G Taipei Summit:

1. Wider rollouts are more complex in Europe

As noted by Dr. Colin Willcock, the Infrastructure Association chair of the 5G Infrastructure Public Private Partnership, the regulatory and technical hurdles of deploying in the EU are higher than in Asia, owing to governmental differences between the (presently) 28 member states.

Compared to Taiwan—as well as other Asian countries, such as Japan and South Korea—early rollouts in Europe are more likely to be limited to specialized needs, such as providing connectivity in the 2020 UEFA European Football Championship, as well as priority vertical sectors, such as healthcare and connected factories.

2. ULRRC makes 5G vital for industrial IoT applications

Ultra-reliable and low-latency communications (ULRRC) are effectively prerequisites for wireless communications in industrial internet of things (IoT) use cases, as consistent networking greatly simplifies the use of robotics and telepresence systems.

While traditional wireless networks generally deliver data on a best effort basis—with latency spikes occurring with network congestion—the ability for 5G to deliver ULRRC, with consistent and predictable delays, makes discrete manufacturing and warehouse operations safer. Latency is further reduced with the presence of edge cloud servers.

SEE: Enterprise IoT research: Uses, strategy, and security (Tech Pro Research)

3. Network slicing enables 5G to provide customized service for independent use cases

While more broadly applicable than ULRRC for industrial IoT applications, the use of network slicing—effectively network virtualizing, the ability to create multiple virtual networks with differing capacities for differing needs—will ease the deployment of new types of services. For applications such as autonomous cars, which require a low latency though not necessarily a high throughput, the network resources to achieve this can be partitioned off from the rest of the network. This will allow 5G to be used as the one wireless communication standard, supplanting situations where multiple heterogeneous (and often mutually incompatible) networks were deployed before.

4. 5G and machine learning can be paired to improve indoor positioning

Rajeev Agrawal, Nokia’s Algorithm Innovation Lead, described a method at the summit to create a more accurate indoor positioning system without the use of GPS signals. Current technology allows for a 5-15 meter accuracy. The proposed model uses WinProp to generate detailed RF fingerprinting of a building, which is subsequently used with a 3D map to train a neural network. With existing LTE networks in a test case at a shopping mall, used with a Feed-Forward neural network, the mean positioning error was measured at 3.4 meters, while this was was reduced to 1 meter when used with a Recurrent neural network. (Both measurements on the 1st floor.)

5. Millimeter wave deployments are possible, though complex

The 5G standards divide frequencies into two groups, FR1 (450 MHz — 6 GHz) and FR2 (24 GHz — 52 GHz). Most early deployments will be in the FR1 space. Research is ongoing into using FR2 frequencies, though will practically require massive MIMO antennas for use.

Millimeter wave networks are also highly susceptible to atmospheric interference—effects such as rain fade are problematic for outdoor use, though even nearby foliage can disrupt a signal. Additionally, Millimeter wave networks are also susceptible to building effect signal degradation.

Also see


Image: iStockphoto/jamesteohart


Добавить комментарий