Home5GFive thousand private 5G networks in China? BS! Talking definitions, storms in teacups

Five thousand private 5G networks in China? BS! Talking definitions, storms in teacups

Someone somewhere said something about 5,000 ‘private networks’ in China. (Forgive me for not having the reference to hand; I am trying to write quickly and cover ground.) The point, as always, seemed to be just that China is running away with it – that despite the excitement (in the supply market) in the west, China’s government-drive on private (cellular-based 5G) networks is working better, and way ahead. 

It was the same with NB-IoT; it is always the same, the message goes – that while the US and Europe kick the tyres, China burns rubber. But Enterprise IoT Insights put this 5,000-figure (cribbed from a local operator’s financial results, as we recall) to a China-based 5G vendor, doing decent 5G sales with Chinese enterprises, last week, and their response was emphatic. BS, they said (effectively, practically holding their nose). “There are 2,000-3,000 – and 95 percent run off a public core network.”

In other words, these are not all-edge private 5G networks, at all, but more like ‘hybrid private networks’, or ‘virtual private networks’. They are not, in 3GPP parlance, standalone non-public networks (NPNs), entirely separated from public network infrastructure –and supposed to drive hard-nosed Industry 4.0. These properly-private 5G networks, with everything under the control (nominally) of enterprises, are the yardstick for Industry 4.0, arguably.

This is what Dow Chemical is doing with Kyndryl and Nokia; this is what Schneider Electric is doing with NTT; this is what Airbus and Lufthansa Technik are doing, also. In some cases, said our China source, firms are using public RAN infrastructure, either with dedicated sub-let chunks of national spectrum or some early prototype-slicing. The sense, in the end, is China’s private 5G market is bigger and faster, but also more limited (and potentially limiting).

There was further commentary but, again, we are trying to make short work of this, more as a discussion point. (Plus, my notes are scrawled). But, in a separate conversation this week, with the local contingent of an Asia-based telco-cum-techco in the home of Industrie 4.0, the message was the same; their counterparts selling to factories in China tell how most business goes via public operators, and that little-or-nothing can be classified as all-edge NPN.

This is not what Industry 4.0 wants, they said; it is not the same as what they are offering.

We put the notion, as well, to a handful of well-placed market watchers; just that the China story with private 5G should be asterisked. Stefan Pongratz, vice president at Dell’Oro Group, said the term, ‘private networks’, should also be asterisked; that it is a marketing term, invariably co-opted by all sides for their own ends. “Three years into covering this private wireless market, I still spend a lot of time talking definitions,” he said.

There are private LoRaWAN networks, and private Wi-Fi networks, and private-everything networks, he noted, Yes, but in the context of industrial-grade 5G for Industry 4.0, which is how Enterprise IoT Insights prefers to judge (or care about) the market, some ‘networks’ are more private than others. Dell’Oro Group follows the 3GPP line about NPNs, that they may be standalone or stand-with, effectively; that the China count is correct, also.

These definitions might be further refined, he suggested – “to minimise… gaps across the analyst community”. At the same time, the only definition that really matters is about new 5G sales to enterprises. “As one major vendor said: ‘We don’t care what you call it; we just want to know how much growth there is on top of the existing mobile broadband (MBB / eMBB) market’. Which is part of the reason some prefer to just talk about enterprise versus carrier [network sales],” he replied.

On the reported China numbers, Pongratz said speculation they are somehow finessed is unfair. “While I don’t always trust the data coming from this direction, I don’t think they have been trying to be overly misleading this time around when it comes to the 3,000-plus 5G project definitions. Most operators and vendors actually prefer the term ‘5G2B’, versus private 5G – which might actually be a better term… to minimise room for interpretation.”

But 5,000 or 3,000-plus, or 3,000-max? Leo Gergs, analyst at ABI Research, reckons the (somewhere) reported China figure is wrong, also. “That number, of 5,000 private 5G networks, is too high; I would think 2,000-3,000 is much more reasonable. I would also agree that most cannot be classified as fully-isolated private 5G, with dedicated infrastructure on enterprise premises.”

He explained: “Just because Chinese operators would still need to be involved for spectrum access – and so they would typically want to utilise their existing public network resources. Our Chinese contacts often talk about ‘private lines’ for enterprises, which essentially means ‘virtual private networks’. The numbers in China are certainly skewed towards the operators; they are not really private networks, as they certainly use the public core.” 

On this, Pongratz agrees; although he thinks “manufacturing / industrial IoT is actually driving the lion’s share of the 3,000-plus 5G2B projects”, and that most are not sharing RAN infrastructure. He responded: “I don’t have the exact splits for core sharing, but dedicated small cells are being deployed in the majority of these projects. So from a small cell volume perspective, China is leading when it comes to on-site small cell deployments for vertical industries. 

“If one uses the definition that no core sharing is allowed with the carrier to fall into the private bucket, then, yes, I agree the portion that is allocated for ‘private only’ would be smaller. But for me, tracking private RAN, this approach would also be difficult to justify. Because our clients want to understand where the RAN growth is relative to the base case, with traditional MBB – new cap-ex versus old cap-ex is the priority.”

Gergs says the same, in effect; that the China numbers should be considered as a reflection of an operator-led market bent on popularising hybrid/virtual private 5G, but that it is enterprise-geared 5G, all the same, and that China’s momentum is good. “I still think China is doing relatively better on private 5G than Europe, the US, and other parts of Asia – just because of the influence of the government on enterprise decisions, and the perception in China that national operators are more viable digitisation partners.”

Last word to Pongratz, seeking to put this storm back into its teacup. “I don’t think everyone understands how small the private wireless market actually is right now, especially private 5G. These small cells are going for a few thousand dollars; they are not in the $10,000-plus range. And we are closer to 10 than to 100 cells per project. Huawei has 3,000 5G small cells in 200 coal mines. There is some activity in the US, Germany, France – but CBRS has been more focused on non-industrial verticals to date.”

Previous post
Semtech/Sierra, Telit/Thales set ground for new east-west IoT clash (Analyst Angle)
Next post
CommScope to enable 5G and CBRS at AT&T Center