First the good news. Network slicing remains a topic that is undoubtedly engaging both operators and verticals. This year’s Network Slicing World Summit – the second – which took place in Berlin in late October, was collocated with the Broadband World Forum 2018, which gave it the opportunity to expand its depth of content, scope and relevance, which it did.
The papers and presentations were wide-ranging, taking in research, the operator challenge, AI, standards development, service enablers, ecosystem creation, security, use cases, regulation, the business case, business drivers, the potential role of 4G and a lot more.
Real Wireless, whose role in network slicing techno-economics research has long been established through projects like MoNArch and NORMA, chaired two sessions: one on operators providing what the verticals expect and need and another on security.
The former was certainly interesting. Mark Roddy, of the Cork Institute of Technology, talked about the theoretical benefits of slicing. The other presenter was Xueli An, Vice-Chair Use Case and Requirements of Working Group 1 of the 5G Alliance for Connected Industries and Automation (5G ACIA). Xueli discussed the features that would be needed in networks for manufacturing or robotics as well as automated systems that could benefit from network connectivity – a good sign of progress on an industry eco-system coalescing around key features to influence standards.
At the security session a cybersecurity expert focused, as you might expect, on what he felt was the inherently risky side of network slicing. However, many MNO, Wi-Fi and enterprise attendees seemed to feel that security is a value proposition of telco carrier-grade equipment that is already dealt with by the standards – part of the DNA, you might say. In fact, they argue that, potentially, slicing is a way to isolate different services and to allow for different types of security provisioning on each slice. That being the case, cybersecurity concerns should not retard the adoption of slicing.
The Real Wireless view is if the wireless industry vision for vertical adoption is to be realised, this area needs to be exposed. It isn’t enough to claim that mobile networks have traditionally been well engineered in this regard and therefore “there’s nothing to see here”. In a converged IT and communications networking domain that is embedded into the critical infrastructure this area must receive due scrutiny. It certainly will from Real Wireless. We can already call on leading wireless security expertise, but this is an area we expect to see grow in relevance and we plan to enhance our ability to address it accordingly.
Right now, however, some of the discussions in this and other areas of network slicing are both complex and theoretical – and that seemed to be the overarching theme of the Network Slicing World Summit.
For example, a first-generation realisation of a RAN static slicing arrangement running on a public network is a reality in the ESN (emergency services network) service on the EE public network. Fixed network slicing propositions exist. However, the concept that was mostly being discussed in the network slicing conference – and also in RAN World, incidentally – was dynamic end-to-end slicing, which offers no shortage of technical and commercial challenges related to more dynamic resource management and convergence of the fixed and wireless domains. It’s undoubtedly part of the work programmes for the R&D groups of the larger MNOs and OEMs, and there’s certainly an aspiration to be able to define products and services for the market, but there’s some time to go before any of this becomes a reality.
Is that bad news? No. Network slicing may be a desirable option, but everyone involved knows it’s not an easy one. The expectation, after all, is that it will be one of the defining innovations in the 5G-era that enables a competitive advantage in the service layer. That won’t happen overnight.
Through MoNArch and NORMA, we at Real Wireless can say with some confidence that we understand important concepts like the defining of slicing templates, what to consider in the dimensioning of a slicing-capable network, and how to define the operational requirements of a slice.
But we can also say – and much of the content at the Network Slicing World Summit bore this out – that network slicing remains a topic of intense industry research, with academics also finding many aspects to investigate. Releases 15 and 16 of 3GPP have set a direction for standardised product development and features, but at-scale commercial propositions are yet to emerge.
In other words there is definitely a will to make network slicing happen – and we do expect it to happen. The only question is when.