The next release of 5G standards, 3GPP Release 16, will help to deliver the platform’s key promise – to enable a wide range of enterprise and IoT services, expanding the cellular model well beyond voice and mobile broadband for the first time. ETSI, too, is broadening the remit of its standards work to look beyond operators, to other sectors which can leverage 5G to change their business.
However, the potential of 5G to help industrial organizations to transform their processes and services will be limited if everything is left to the MNOs. Most operators acknowledge that they are finding it hard to firm up a business case for many IoT and enterprise use cases. Those which are pushing further into the enterprise tend to be focusing on a few selected markets, such as automotive, where they feel confident about building an ecosystem and strong channels to market.
But that will leave many industries without strong operator attention, and the risk that 5G networks will not be built out in the locations they need, especially if those are remote, or deep indoors; or with the capabilities they need, if these go beyond high speed mobile broadband.
Network slicing may address the problems in time, but for the foreseeable future, the most logical solution is to create a robust environment to support enterprise specific service providers and private networks. Some of these can work as MVNOs, or in shared spectrum, but where demanding requirements are involved, such as tight integration with enterprise platforms, or critical availability, industries are keen to have greater control – which means their own spectrum, or at least spectrum owned by a neutral host provider focused on industrial needs.
As governments start to fear that the economic and industrial benefits they have envisaged from 5G will not materialize, they will become more open to making spectrum available for industrial, IoT, neutral host or non-broadband use. The UK missed the chance to do this in its first 3.5 GHz auction, but is consulting on it for the next allocation. Ireland, Portugal and Belgium made some 3.5 GHz spectrum available at a price that was viable for a new entrant, and so Dense Air – a subsidiary of Airspan planning to build neutral host networks for enterprise services – was able to acquire airwaves.
The most significant regulatory moves so far to support private, or semi-private networks are the FCC’s three-tiered CBRS scheme in 3.5 GHz, and Germany’s proposal to earmark some licensed spectrum for Industrie 4.0 purposes in its fortcoming 3.5 GHz sale. These relatively high frequency bands are well-suited to small cells and local, indoor networks that can support a self-contained RAN and core for an enterprise.
But the same operators which are hesitating over the IoT or indoor business case, and failing to address the requirements of industry, are lashing out furiously at plans to support a more varied service provider ecosystem. The German telcos have threatened legal action over the Industrie 4.0 plan, and at Mobile World Congress, Vodafone’s chief engineer offered a dire warning to regulators which try to set aside spectrum for industrial use.
Luke Ibbetson said the regulators would risk degrading services for the majority by limiting the amount of spectrum MNOs could use. He told a media briefing in Barcelona:
“There are some consultations in certain countries regarding ringfencing mobile spectrum for industrial use cases and for operators this is not the correct approach because you end up reducing the trunking efficiency by locking up spectrum in a way that doesn’t allow us to get the full benefits.”
He added: “I would advocate continued mobile spectrum use in a way where we can get the benefits of 5G and allow factory systems to use the same spectrum, but have it controlled in a seamless way by operator partners.”
That would be a reasonable argument if so many MNOs had failed repeatedly to address the most basic need of enterprise cellular, good indoor coverage, let alone anything more specialized like very low latency. That will be one of the capabilities enabled by Release 16, but it will only be implemented if operators see the business case to do so.
Ibbetson also warned about the risks of fragmentation and loss of scale, saying: “It is important to bring this back to using the same spectrum assets and not fragmenting by having industrial devices somehow requiring different frequency ranges to those components developed for smartphones. Otherwise we will lose the benefits of trying to scale the technology for those environments.”
And his CEO, Nick Read, echoed the sentiments, saying in a separate interview: “What we don’t support is if someone can get spectrum and build a little bit of network in cities and then ride off the rest of the industry on national roaming agreements. It undermines the investment we do.”
In the face of potentially seeing 5G spectrum divided up among more players, MNOs may start to work more seriously with enterprises, or at least lease their spectrum more affordably to private operators. Last week, Deutsche Telekom announced a deal with lighting manufacturer Osram to develop a 5G campus network for services such as robotics. It did not look like a coincidence that this came in the wake of the regulator’s plans for industrial spectrum allocations.
Meanwhile, the standards organizations are more amenable to the idea of diversification in the mobile services market. ETSI has been heavily focused on telcos, but its director general, Luis Jorge Romero, made it clear in an MWC interview that the enterprises needed to be considered too. Implicitly, he was acknowledging that MNOs cannot be entirely trusted to provide the best networks for every type of enterprise – their specific needs must be addressed right at the standards level too.
“They need connectivity, so they have to be brought into the equation for the standards,” said Romero. “They are the ones who know the issues and want to have a solution. They can translate the problem into what we understand.”
At ETSI, this means that there will increasingly be two types of working group for each new area of standardization. The high level ones will be to address telco needs, and then there will be vertical-specific applications groups which will drill down from those high level platforms to specific industry requirements. Romero gave the examples of the creation of an open data platform for marine asset tracking worldwide, or the participation of agricultural equipment maker John Deere in IoT working groups. He hopes this will lead to better mutual understanding between telcos and other industries, and future platforms that reflect that.
Over at the 3GPP, its SA 1 (service and system aspects) working group recently concluded an important aspect of the Release 16 standards work for vertical markets. This focused on 5G service requirements for what 3GPP calls cyberCAV (cyber-physical control applications in vertical domains), which underpins many IoT, robotics and other industrial control systems envisaged for 5G.
According to Michael Bahr from Siemens, rapporteur for the cyberCAV work, this work was concluded with the approval of Technical Specification (TS) 22.104, which lays the foundation for more detailed work in Release 16.
The cyberCAV efforts in SA1 are particularly focused on industrial automation and energy automation, relying on Release 16’s URLLC (ultra-reliable low latency communications) capabilities to share real time information between machines. TS 22.104 also contains service requirements, such as on-clock synchronization for time-sensitive networking in cyberCAV applications.
Broader cyberCAV requirements are included in another set of specs, TS 22.261 (‘Service requirements for next generation new services and markets’). This includes 5G requirements for security, network capability exposure, QoS monitoring and Ethernet transport. And it specifically includes the requirements of private 5G networks.
The second stage of work on cyberCAV has now started, and will define corresponding extensions of the 5G architecture, 5G features and functionalities.
Section 6.28 of TS 22.261 provides this definition of 5G cyberCAV:
- A vertical domain is a particular industry or group of enterprises in which similar products or services are developed, produced, and provided.
- Automation refers to the control of processes, devices, or systems in vertical domains by automatic means with sensors, transmitters, controllers, and actuators.
- Cyber-physical control applications are to be understood as applications that control physical processes. In automation, they follow certain activity patterns (open loop control, closed loop control, sequence control, and batch control).
- Communication services supporting cyber-physical control applications need to be ultra-reliable, dependable with a high communication service availability, and often require low or (in some cases) very low end-to-end latency.