Telco core upgrades, mmWave spectrum to unlock 5G for enterprises

Both Telstra and Optus last year began offering 5G mobile coverage in select locations, with Australia’s third MNO, Vodafone Hutchison Australia, revealing in December that it had partnered with Nokia to upgrade its network from 4G to the new cellular standard.

Telstra in May 2019 began selling its first 5G-capable mobile device. Optus in January 2019 revealed details of a 5G wireless home broadband service and then later on launched support for 5G mobile services.

From the perspective of Australia’s telcos, early 5G adoption will play a role in capacity offload for their 4G networks. For users within the still-limited 5G footprints, it will enable super-fast downloads.

However, for enterprises many of the mooted use cases for 5G, such as edge compute, supporting IoT deployments, and as a potential alternative to Wi-Fi, will be delivered courtesy high-frequency millimetre wave (mmWave) spectrum and the move to 5G cores by MNOs.

Telstra’s network engineering chief, Channa Seneviratne, said that the telco saw the auction of spectrum in the 26GHz band, which will take place later this year, as a key step to unlocking 5G enterprise use cases in Australia.

“We would really hope for 10x the amount of spectrum we have today, but it depends on the competition limits set by [the Australian Communications and Media Authority] and the government,” the Telstra executive said in an interview conducted on the sidelines of Telstra’s 2019 Vantage enterprise conference

In October 2019, the federal government issued a direction to the ACMA to prepare spectrum in the 26GHz band to be auctioned for use with 5G. (In 2018 the ACMA auctioned off spectrum in the 3.6GHz band, which is being used to deliver the first generation of 5G services in Australia.)

The second key step will be Telstra’s shift to a 5G core, Seneviratne said. Currently both Telstra and Optus are employing the 3GPP-developed Non-Standalone 5G NR (New Radio) standard (based on 3GPP Release 15). At the moment, for example, 5G devices still rely on telcos’ LTE networks for some functions (call origination and termination, for example), and they are still employing a 4G core.

“Much of 5G’s potential performance requires the installation of ‘stand-alone’ infrastructure, ideally compliant with R16, which is expected to be finalized around June 2020,” noted an October 2019 market trends report from Gartner.

The 5G standard outlines a number of new features that will become available as telcos upgrade their cores. “So for example, one thing that we want to demonstrate as early as we can is a network slice,” Seneviratne said.

“A network slice is where we can say, either at an enterprise level or a traffic class level, ‘Here’s what you can do in terms of dialling up the bandwidth you want, the availability you want, and also the latency you want.’”

That’s “something I’m driving my team mad about,” the Telstra executive said: “How do we stitch the slice up through the 5G network, through the IP core network, all the way — truly end-to-end slicing. That is when we’re really going to start seeing enterprise use cases really coming to life.”

Another feature the Telstra exec sees a lot of potential in is the Network Exposure Function (NEF), which will facilitate secure interaction between enterprise applications and a 5G network.

“That’s another way you can expose network services up to a top-level application, including analytics and a whole bunch of things,” Seneviratne said. “That, combined with the slicing, is where you not only create the slice but you can connect it up to different apps.”

5G will eventually be delivered over a combination of frequencies: Low-band (sub-1GHz), mid-band (1-6GHz) and mmWave (26GHz in Australia — although Telstra and Optus have raised questions about the future of the 28GHz band).

In a 5G briefing held late in 2019, Telstra’s networks principal, mobile innovation rollout, Paul Milford, said that company expected mmWave to be the key spectrum for enterprise and industrial applications, while mid- and low-band spectrum will be used to deliver enterprise and consumer mobility services.

mmWave will not only enable extremely fast data speeds, but the upcoming 26GHz auction will also mean that the amount of spectrum that telcos can use for 5G is greatly expanded: Up to 2.4GHz will be on offer in the ACMA-run auction.

“The 5G core network is not going to go into production until early 2021,” Seneviratne said. “So the 5G core network in production and millimetre wave spectrum coming available are actually happening at the same time in terms of timing.”

Telstra has already exploring the use of 5G for enterprise applications. In February 2019, Telstra and the Commonwealth Bank of Australia released details of their collaboration to assess the use of 5G edge computing at CBA branches.

“We’re excited to be working with Telstra and Ericsson to test and learn on 5G and edge computing,” the bank’s general manager digital and retail operations and technology, Pete Steel, said at the time.

“These technologies have significant potential to enhance the availability, stability and performance of our network infrastructure and we hope they can help us provide quicker and better digital experiences for our customers.”

An early focus of the partnership has been the potential of 5G to slash the infrastructure footprint of a bank branch, as well as moving CBA applications to the edge.

Seneviratne told Computerworld that Telstra was working hard, including through its collaboration with CBA, to develop a 5G edge compute framework that can be replicated across a range of industry verticals.

“In edge computing, higher-bandwidth and latency-sensitive applications benefit from placing compute as close as possible to the wireless client device, especially when bandwidth is constrained from the edge back into the core,” Gartner noted in its October 2019 analysis Market Trends: Will the Advent of 5G Make Enterprise Wi-Fi Connectivity Less Relevant?

“5G can accelerate the deployments of truly distributed computing architectures where information is exchanged and processed in near real time across multiple nodes that are geographically distributed.”

Credit: Rohan Pearce, Computerworld