For
the last few years, I've written and spoken extensively about 4G or 5G
cellular networks optimised for enterprises, whether that's for a
factory, a port, an electricity grid - or even just a medium-sized
office building. Recent trends confirm the acceleration of this model.
- CBRS in the US is growing rapidly, including for local and industrial/utility uses
- Localised
4G/5G spectrum is now available in UK, Germany, Netherlands, France,
Japan and elsewhere, with many new countries examining the options
- Many campus/dedicated network strategies by traditional mobile operators (MNOs)
- Assorted testbeds and trials sponsored by governments, groups like 5G ACIA etc.
- Growing intersections with Open RAN and neutral host models
An inflection point has now been reached.
Enterprise/local cellular is happening, finally
It's
been a long time coming. In fact, I've been following the broad concept
of enterprise cellular since about 2001, when I first met with a small
cell vendor, called ip.access. Around 2005-2009 there was a lot of
excitement about local 2G/3G networks, with the UK and Netherlands
releasing thin slices of suitable spectrum. A number of organisations
deployed networks, although it never hit the massmarket, for various
reasons.
Now, however, private 4G and 5G is becoming "real".
There's a critical mass of enterprises that are seriously interested, as
this intersects with ongoing trends around IoT deployment, workforce
automation, smart factory / city / building / etc concepts, and the
availability of localised spectrum and cloud-based elements like network
cores. It's still not easy, but the ingredients are much more
accessible and easier to "cook".
A binary choice of MNOs vs enterprise?
But throughout this whole story we've had an underlying narrative of a two-way choice:
- Enterprises
can obtain private / on-premise cellular networks from major MNOs as a
service, perhaps with dedicated coverage plus a "slice" of the main
macro network and core functions.
- Enterprises can build their
own cellular networks, in the same way they build Wi-Fi or wired
ethernet LANs today, or operate their wider private mobile radio (PMR)
system.
This is a "false binary". A fallacy that there's only two options. Black & white. Night & day.
In reality, there's a whole host of shades-of-grey - or perhaps a better analogy, multi-coloured dawns and sunsets.
Not just MNOs
There is a lengthening cast-list of other types of service provider
that can build, run and sell 4G and 5G networks to enterprises or
"verticals" (the quaint & rather parochial term that classical
telcos use to describe the other 97% of the economy).
An incomplete list of non-traditional MNOs targeting private mobile networks includes:
- Fixed
and cable operators, especially those which have traditionally had
large enterprise customer bases for broadband, VPNs, PBXs / UC, managed
Wi-Fi etc.
- MVNOs wanting to deploy some of their own radio
infrastructure to "offload" traffic from their usual host provider in
select locations.
- TowerCo's moving up the value chain into private or neutral networks (for instance, Cellnex and Digital Colony / Freshwave)
- IT
services firms affiliated to specific enterprises (for example, HubOne,
the IT subsidiary of the company running Paris's airports)
- Industrial
automation suppliers acting as "industrial mobile operators" on behalf
of customers (maybe a robot or crane supplier running/owning a local 5G
network for a manufacturer or port, as an integral part of their
systems)
- Utility companies running private 4G/5G and providing critical communications to other
utilities and sectors (for instance Southern Linc in the US), or
perhaps acting as a neutral host, such as a client in Asia that I've
advised.
- Dedicated MNOs for particular industries, such as oil & gas, often in specific regions
- Municipalities
and local authorities deploying networks for internal use, citizen
services or as public neutral-host networks for MNOs. The Liverpool 5G
testbed in the UK is a good example, while Sunderland's authority is
looking at becoming an NHN.
- Railway companies either for
neutral-host along tracks, or acting as FWA service providers in their
own right, to nearby homes and businesses.
- Specialist IoT connectivity providers, perhaps focusing on LPWAN connectivity, such as Puloli in the US.
- FWA / WISP networks shifting to 4G/5G and targetting enterprises (eg for agricultural IoT)
- Overseas
MNOs without national spectrum in a market, but which want to service
multinational enterprise clients' sites and offices. Verizon is looking
at private cellular in the UK, for instance - and it wouldn't surprise
me if Rakuten expands its footprint outside Japan.
- Property and construction companies, especially for major regeneration districts or whole new smart-city developments.
- UC/UCaaS and related voice & communications-centric enterprise SPs, such as Tango Networks with CBRS
- Universities
creating campus networks for students, or other education/research
organisations servicing students, staff and visitors
- Major cloud
providers creating 4G / 5G networks for a variety of use-cases and
enterprise groups - Amazon and Google are both tightly involved (albeit
opaquely, beyond Google's SAS business), while Microsoft's acquisition
of Metaswitch points to cloud-delivered private 5G, albeit perhaps not
with spectrum and RAN managed itself.
- Tourism and hospitality
service providers providing connectivity solutions to hotels or resorts -
although that's probably taking a backseat given economic &
pandemic woes.
- Broadcasters, event-management and
content-production companies deploying private networks on behalf of
sports and entertainment venues, festivals
- Dozens more options -
I'm aware of numerous additional categories and more will inevitably
emerge in coming years. Ask me for details.
Conclusion: beyond the MNO/Enterprise binary fallacy
You
get the picture. The future of 4G / 5G isn't just going to split
between traditional "public mobile operators" (typically the GSMA
membership) vs. individual enterprises creating DIY networks. There will
be an entire new universe of SPs of many different types.
You can
call them "new telcos", "Specialist Wirelss SPs", "Alternative Mobile
Operators" or create assorted other categories. Many will be multi-site
operators. Some may be regional or national.
We will see MNOs set
up divisions that look like these new SP types, or perhaps acquire them.
Some vendors will become quasi-SPs for enterprise, too. This is a
hugely dynamic area, and trying to create fixed buckets and segments is a
fool's errand.
Understanding
this new and heterogeneous landscape is critical for enterprises,
policymakers, vendors and investors - as well as traditional MNOs. I've
been saying for years that "telecoms is too important to be left to the
telcos", and it appears to be becoming true at a rapid pace.
Many
in the mobile industry assert that 5G will transform industries. In many
cases it will.... but the first industry to get transformed is the
mobile industry itself.
This newsletter & my services
Thanks
for reading this article. If you haven't subscribed to my LinkedIn Newsletter updates, please
look for the "subscribe" button here. If it has resonated, please like this
post and share it with others, either on LinkedIn or on other
channels.
If you have a relevant interest in this and
related topics around the future of telecoms and technology, please
connect with me. (But no spammers and "lead generation" people, please).
I
do advisory projects, strategy workshops and brainstorms, or
real/virtual speaking engagements on the 5G, spectrum, private network
and broader "telecom futurism" space. Drop me a message about how I can
help you.
A key theme here is the fast-evolving model for #neutralhost mobile for small cells and network capacity in-fill in cities. An NH is a 3rd party wholesale provider which enables multiple tenant 4G/5G mobile providers - generally MNOs, but also potentially including private networks as well.
A few years ago when I was running NH workshops with Peter Curnow-Ford we identified this area of metro infill as one with potential, but limited actual deployments.
There are numerous challenges - MNOs ideally don't want separate deals with each city authority, while cities don't want multiple MNOs independently requesting 100s of sites with associated street clutter, road closures and soon. Authorities also want to both make money from access to assets such as lampposts, and to improve connectivity for citizens and businesses as fast as possible.
One option floated was for authorities to build out their own private 4G/5G networks, then allow MNOs to roam onto them, or use some sort of MOCN network-sharing arrangement. But MNOs each have different coverage / capacity holes, different spectrum bands, different customer groups - and also worry about security, ability to manage radio units, do carrier aggregation and so on. The idea of a single cell network in its own spectrum, with multiple MNO tenants is appealing, but sometimes unworkable. (It might work OK in villages or indoors, though).
What's happening is that another model is evolving. Local authorities like city councils are contracting with several infrastrucure specialists - companies like Cellnex UK , Freshwave, Ontix, BAI Communications and Shared Access to run (essentially) small-cell as a service offers. These act as intermediaries, allowing local authorities to create standard contracts, and for MNOs to have standardised processes for getting access at each site.
It reduces the frictions and costs of the paperwork - and also allows for infrastructure-sharing to evolve over time where it makes sense. Coupled with vRAN or open RAN it can put some of the electronics into central facilities, reducing street-side box numbers. And it means MNOs can get coverage in their preferred locations, with backhaul/fronthaul and power supplies simplified.
The competitive infraco/towerco angle, rather than exclusive area concessions, allows MNOs to choose the provider that is the best fit - and without needing different processes in each city.
It's not quite what I expected NH models to look like - and they may differ in the US or across Europe - but it seems to make good sense here in the UK.