5G is the single most repeated term at Mobile World Congress 2018, as indeed it was in 2017. According to the big vendors – Ericsson, Huawei, Nokia and Samsung – 5G is now a reality with all citing numerous memorandums of understanding (MoUs) and a handful of commercial contracts with deployments beginning in Q3 2018. But, and it’s a big but, the latter are for the simpler use cases of enhanced mobile broadband and fixed wireless access.
This is not a criticism; these deployments in the US and Northeast Asia are being driven by very real needs for more capacity and offer operators revenue growth potential.
Capacity, or rather capacity constraints in 4G, are providing the business case for early 5G deployment, re-enforced by a political battle for the global leadership position between the US and China.
However, it also means that the 5G vision, the new use cases and their associated new revenue streams are long term propositions – more like 5-10 years away. And that is a problem as demand is growing from vertical businesses.
Representatives from vertical industries including Alibaba and Rakuten from e-commerce, Bosch for manufacturing, Philips from medicine and Toyota from automotive, all laid out their visions for what they want to do with 5G. The exhibition halls are full of demonstrations – some on live 5G networks, most not – that suggest even more potential use cases, from connected diggers and port logistics, to robots and self-drive cars.
The representatives from these verticals are truly excited by the potential efficiency, quality and performance gains that 5G offers their business processes. They have ideas aplenty for how to build transformational business propositions on top of 5G based on a high speed, ultra-low latency, low power, secure, reliable and flexible network that can be infinitely and dynamically sliced up, all backed by enforceable SLAs, controlled by them and delivered as a cost substantially lower than current networks. But the truth is, CSPs aren’t really close to being ready to deliver on a commercial basis.
The first three of these requirements we can tick off. 5G new radio (NR) delivers them all. Furthermore, there is potential for the cost disruption as Ericsson claims that 5G spectral, radio and core efficiency drives costs to the operator down to 10% of the equivalent cost per byte for 4G. This is the 5G that will be deployed towards the end of 2018 and the achievement should not be underestimated. However, it is not enough to release the potential for the new revenue streams associated with the massive machine and mission critical use cases.
Ericsson puts a revenue number on these new revenue streams for CSPs at anything from $209 billion to $619 billion depending on the role CSPs take. Whether you believe these numbers or not, it is fair to say that there is strong revenue potential in these areas and it will be demand-driven.
Andreas Mueller, senior expert and project manager at Bosch is representative of the type of customer CSPs can expect. He claims that “Industry 4.0 is killer app for 5G” and while the concept of a ‘killer app’ is one that should have died with 3G, it does illustrate just how much he and other industry vertical players want to convince the mobile world and CSPs particularly, of their market value.
Mind the expectation gap
However, the problem is that there is a considerable expectation gap developing. Mueller and his counterparts are far from illiterate when it comes to the technical elements of 5G. He has a strong understanding of his need for network slicing – and not just a network slice per vertical but many, many more. He claims, “We need use case slices or it [5G] is useless.” Furthermore, he wants sub-slices that he, not the MNO control and the ability to ramp up and tear down slices in less than 10 minutes. He also wants services that works across network and country borders and that means interconnection and security over the interconnection, which isn’t currently part of the 3GPP standards although it is being worked on. He also wants that international connection to be low latency.
For example, he wants to be able to remotely operate production machines in China from headquarters in Germany. A simple request but one that is beset with technical issues and complexity as the remote-control device includes an emergency stop button. That means local edge intelligent control as well as central control and it means complicated and dynamic orchestration.
Neither does Mueller’s wish list stop there. He wants end-to-end QoS, backed by accurate and granular SLA monitoring in real time and the potential for third party SLA monitoring. In fact the list of requirements just goes on and on.
There is no doubt that 5G has potential to create much-needed new revenue streams for CSPs but to unlock them, they will have to do much more than deploy and release. They need to create entirely new network topologies, put orchestration at the heart of their network development and build a new generation of IT to deliver the network management and operational and business support services will turn technology into services and revenue. 5G has started but it has a long way to go.