The award of 74 spectrum licences for 5G campus networks by Germany’s federal network agency BNetza heralds a likely spate of similar developments in other countries as interest grows across the manufacturing sector in Industrie 4.0 applications under the IIoT (Industrial IoT) banner. But this also highlights a major challenge for MNOs as they face a fight to take a significant slice of the burgeoning private 5G networking market, with the risk of losing out on a major revenue stream.
After all, not only private enterprises but also a growing number of utilities and other public sector bodies are already building their own private 5G networks, where that is already allowed, for example in Japan, the UK and USA. Germany is a particular hotbed of activity, reflecting the eminence of manufacturing there, and by opening 5G frequencies to all comers, BNetza is helping spread the technology beyond MNOs in both the private and public sectors.
This has left MNOs relying on network slicing to offer private network services within their own infrastructures, but this requires a 5G Standalone network with its own 5G core. Meanwhile, some of the top operators have been setting up dedicated private 5G networks. Examples include VodafoneUK working with Ford Motor at the carmaker’s UK electric vehicle manufacturing plant, with the aim of accelerating production of electric batteries. In Germany, Vodafone has been working with Lufthansa Technik to build a 5G network at an aircraft hangar at Hamburg Airport.
For many such projects though, MNOs have been left out, with none at all involved, for example, in a private 5G project for Deutsche Bahn, Germany’s national railway service, built with Nokia equipment to support automated trains. MNOs do face a challenge convincing enterprises that they should be entrusted with private networks.
Salvation may come from provision of virtual campus networks (VCNs) at sites with more complex and dynamic private 5G requirements, where demands are continually changing over both time and space. There are various industries and situations where networks are not conveniently confined to dedicated plants, because operations span large diffuse sites peppered with volatile clusters of activity that are ever-changing.
These pose challenges for management, monitoring and maintenance, while often being under competitive pressure to increase efficiency and speed of deployment. Such operations include gas exploration sites, mining sites, oil rigs, universities and sports stadia, possibly in decreasing order of dynamism. Many such operations with distributed infrastructure constantly evolving lack the expertise to set up and run their own private networks and so should be candidates for 5G VCNs created out of existing public networks employing network slicing for the required flexibility and agility.
While factories have relied to date largely on wired links with their inherent limitations for processes involving moving parts, many of these remote distributed sites, such as oil rigs and gas platforms, have depended instead on satellite communications, which is expensive, complex to manage and lacking in performance. Where backhaul can be extended to such sites, either over fixed or wireless links, 5G VCNs offer a more robust and low latency option for private networks, particularly to serve growing numbers of unattended mobile units for exploration as well as monitoring.
In some cases the VCNs could be supported by edge computing to localize processing and reduce latency, while still enabling at least outbound connectivity to public networks without jeopardizing security. But exactly where the edge is deployed will depend on the use case, with some better served by a somewhat centralized approach.
This might include temporary installations such as mega events or festivals, which during non-pandemic times attract huge numbers of visitors for a short time.
Construction sites also have temporary communication requirements over a slightly longer period. Both these could be served by VCNs within public 5G networks where the multi-access edge computing is fixed at a more central data center serving these transient use cases.
The point is all these use cases can exploit the robustness of emerging 5G networks and present opportunities for MNOs to seek the revenues they desperately need to pay back their investments in 5G spectrum and infrastructure. Indeed, although enterprises may be wary about security and privacy, network slicing will prove increasingly attractive, especially if they are convinced that this does also provide the same logical isolation as a private network operating in unlicensed spectrum.
It is certainly the case that by around 2024, network slicing via public networks will offer cost-effective options for 5G based VCNs. This would exploit the 5G service-based architecture (SBA), which enables MNOs to partition their public networks and create in principle any number of effectively private 5G (or for that matter LTE) sub-networks or network instances.
At the same time, only larger enterprises will have the internal IT resources and expertise to construct and manage the infrastructure for a private 5G network. This will include a 5G core, 5G RAN and some transport using either fiber or microwave facilities and/or edge compute capabilities. Quite clearly smaller enterprises could be allured by the prospect of a ringfenced corner of a public network circumscribed by network slicing and underwritten by a SLA (service level agreement), without the aggravation of acquiring spectrum and then managing the network.
MNOs will also have the prospect of setting up private networks in unlicensed spectrum themselves as part of their service. On this front, an important development is the 5G NR in unlicensed spectrum (5G NR-U), which was finalized in June 2020 for inclusion in the latest 3GPP Release 16.
Qualcomm described NR-U as a major chapter not just in the 5G story but the cellular connectivity trajectory as a whole, by bringing the force of emerging wireless technology to unlicensed spectrum. This has certainly galvanized the WiFi community (see separate story). It is the first global cellular standard supporting both licence-assisted and standalone use of unlicensed spectrum, enabling access to 400 MHz downlinks and 100 MHz uplinks for individual devices. It does this by building on the licence-assisted access (LAA) method for accessing unlicensed spectrum that was introduced in 4G as LTE-LAA, the addition being support for 5G NR.
Under LAA, a carrier operating in licensed spectrum acts as an anchor that stays connected, while local unlicensed carriers are added or dropped. Release 16 adds the 5G NR support, allowing use of shared spectrum as an anchor for unlicensed spectrum with NR-U. This includes the 3.5 GHz CBRS band in the US, but also higher bands, noting that the FCC recently released 1200 MHz in the 6 GHz band for WiFi and 5G NR-U.
This finally brings us back to those 74 German spectrum licences because it highlights the tensions between MNOs and enterprises that may become more widespread. A lot of the private networking activity in Germany has been driven by the country’s automotive giants, which have made no secret of their desire to lock MNOs out of their internal 5G infrastructures.
This story has already been widely aired but one of the gripes from operators was that allocation of this private spectrum artificially created a spectrum shortage that pushed up prices at the auctions. Our main point here is that while it does look as if the MNOs will struggle to gain much traction for private networking in the automotive sector and among other large manufacturers, they have better prospects lower down the ladder among medium sized industrial enterprises.