Sorry vendors, 5G ain’t the capex bonanza you’re hoping for: report

5G signal
Image credit: deepadesigns / Shutterstock.com

5G is here – more or less. The first standards are set, the first live public trials are underway in South Korea for the Winter Olympics, and vendors are already announcing a full line of 5G gear ready for cellcos to buy – either right now or later this year. That means 2018 is officially the year the 5G juggernaut kicks off, which in turn means a major capex windfall for vendors who have positioned themselves to capitalize on 5G early.

Right?

Not so fast, says Rethink Technology Research.

A new report from Rethink interviewed a sizable number of Tier-1/Tier-2 operators and vendors, analyzing the objectives, timelines and barriers cellcos see to implementing 5G strategies between 2017 and 2025.

Conclusion: vendors banking on a sudden slew of fat 5G contracts in the next few years are going to be gravely disappointed – during the early years of 5G, operators will be spending half of what they spent on 4G rollouts.

It’s been understood for awhile now that 5G won’t be the straightforward network upgrade that 3G and 4G were in the sense that it is a more heterogeneous affair, with heavy cell densification on the RAN side and a lot of network virtualization and cloudification in the core and the backend, as well as an emphasis on mobile edge computing. It’s also been understood that 5G is not a replacement of LTE, which will play a central role in the 5G ecosystem for at least the next decade.

The Rethink report confirms that 5G will follow a very different pattern to that of 3G and 4G in terms of architecture and regional patterns – operators will prioritize coexistence with 4G and architecture to prolong the life of existing investments. Translation: it won’t be a capex bonanza for the vendors:

There will be heavier reliance on outsourcing and on open platforms to reduce cost and transfer cost further than ever from capex into opex. Operators will spread their investment over a decade and will look to spend as little as half the capex on 5G rollout that they did on 4G.

Operators will embrace next-gen architectures virtualized RANs, hyper-densification, and massive MIMO antenna arrays, but most will deploy them in existing 4G networks, Rethink says:

One or two high profile operators plan to upgrade their Macro base stations first with 5G, because they see 5G as a medium term replacement for 4G. But most will use the new radios to densify selectively while keeping their 4G investment alive, as well as embracing Wi-Fi more aggressively.

Apart from maximizing 4G investments, another key reason most cellcos won’t plunge headlong into 5G is because, frankly, they still don’t see a solid business case for it – not to the point of justifying wide-scale deployment – and won’t spend that much on it until they do, the report says. The key motivations for 5G investment will be where it significantly reduces TCO of service delivery, cost-effectively fills coverage/capacity gaps, enables geographic/market expansion and (of course) enables new revenue streams that 4G can’t deliver (1ms-latency apps, network slicing for verticals, etc).

One other interesting observation from the Rethink report: one key difference of the 5G era is that it will be characterized by a far greater diversity of operators than ever, thanks to trends such as shared and dynamic spectrum for 5G; self-contained, virtualized local RAN and core; and flexible, on-demand wholesale platforms supported by network slicing:

For the first time, a new cellular standard will not be the preserve of just existing MNOs. We expect to see heavy industry pressure to accelerate work on 5G-Unlicensed and on multi-operator technologies which enable a wider range of service providers. The speed with which the industry responds will also help decide the role of Wi-Fi in 5G.

All the details are here.

Likes
Please feel free to share

Be the first to comment

What do you think?

This site uses Akismet to reduce spam. Learn how your comment data is processed.