Your browser is not supported. Please update it.

24 May 2022

It’s time to stretch 5G capabilities to the limit, with edge compute to help

Special Report: Stretching 5G


In 2022, 5G is starting to enter the ‘interesting phase’. There are now many relatively large-scale, commercial 5G networks in operation and we are starting to learn some of the real world lessons about the technology’s performance, quality of service and relationship to other forms of connectivity. Some of these are highlighted in a new study by OpenSignal (see below).

But most of the networks are still based on 5G NR Non-Standalone, which uses the 4G packet core, and they are almost entirely focused on traditional user bases and use cases – basically ‘4G-plus’ mobile broadband. The interesting phase of 5G will come when its capabilities are harnessed in earnest for non-broadband use cases, especially in enterprise sectors.

As we have discussed many times, the 5G business case does not work for the vast majority of operators, if it remains focused on the same users and service models as 4G. Multi-billion dollar investments in boosting the speed of the mobile network to support enhanced consumer applications such as next generation gaming will not be justified by any sufficiently significant uptick in ARPU. This situation, ironically, is worst for operators with advanced 4G network.

There is a case for operators with limited 4G to leapfrog straight to 5G Standalone, which – as long as low-band spectrum is available – can improve the coverage/cost ration for rural and suburban connectivity including fixed wireless. This will also be a pattern in private and enterprise greenfield deployments.

But many operators have invested heavily in LTE-Advanced and there is plenty of life left in those networks before they reach their capacity limits, or their full return on investment. However, MNO dreams of deploying 5G very selectively at first, in areas of very high traffic demand, while keeping 4G as the workhorse, have largely been dashed by pressure from customers, regulators and shareholders to take a lead in 5G.

This makes it urgent to generate new revenue streams in high value enterprise and industrial environments in which there is rising demand for processes to be mobilized, as part of digital and operational transformations – with connectivity that delivers the same reliability and responsiveness as fiber. Supporting business-critical enterprise services can rarely be done with the basic eMBB-focused network.

Instead, investment must be made in implementing the new capabilities of 5G’s current releases and the upcoming 5G-Advanced waves of standards, which will support the specialized requirements of different B2B and B2B2C applications. The networks must support, for example, ultra-low latency, extreme device density (for IoT), AI to support analytics and traffic shaping, and so on.

This is not all about supporting the latest 3GPP specifications, but about planning, tuning and optimizing the network to implement all 5G’s potential capabilities as these evolve. In future, these will include ultra-precise positioning, deeper integration with non-terrestrial networks, new levels of reliability, and so on. Antenna technology is perhaps the most vital physical influence on performance and quality of experience, as BT has discussed this month (see below). The faster-than-expected adoption of Massive MIMO in 5G networks, and many other antenna innovations, show that the future network is not all about virtualization and changes in the baseband. The equipment on the tower is critical too.

If support for the varied demands of business-critical services were not enough of a headache for the network strategists and planners, they also have a new pressure – to make 5G as energy-efficient as possible. As we have covered in previous editions, energy efficiency in 5G is a very complex issue, and of course an increasingly critical one from cost and sustainability perspectives. Smart traffic management and AI support for automated base station switch-off are important ways to counteract the energy impact of densification and Massive MIMO (see below).

These are technologies that will function best with very low latency response, which means investing in capabilities at the edge of the network, even as far out as the cell site itself (which is how Dish is planning 5G). Investment in edge compute, by the telco or via a cloud partner, seems like just another cost in the mounting 5G budget, but is could have a profound impact on the commercial outcome.

Edge compute will help to leverage 5G capabilities to the maximum. Very few enterprise applications really need ultra-low latency – in fact, the operator’s own 5G RAN processing is one of the best examples – but getting the network close to the user will improve overall quality of experience and reliability, as well as saving energy.