4 March 2022

Attending this year’s Mobile World Congress (MWC) in Barcelona – the international and influential show now building back to pre-pandemic form – Aetha Consulting’s Andrew Wright and Jonathan Wall report from the connected tech event on the topics demanding industry attention.

Maximising the potential of 5G, the latest in AI developments and defining the Metaverse set the scene for MWC 2022. While many of the themes dominating this year’s Congress were already informing the agenda at the pre-pandemic 2019 show, some areas are clearly progressing – and are certainly worth watching:

• The realisation of 5G networks (cloud-based core and Open RAN)
• Private 5G for enterprise
• The rise of the Metaverse
• AI everywhere.

5G networks: cloud-based core

Perhaps the most striking trend in networks is the move of the 5G core and elements of the RAN to the cloud. This seems likely to be the default approach for enterprise 5G networks, and is also being increasingly adopted by mobile operator public networks.

The highest-profile cloud-based 5G core is Dish Networks in the United States, hosted by AWS. While Dish’s beta launch has suffered several delays as it attempts to integrate the various components of its network, there is compelling logic to hosting the core in the cloud – it makes efficient use of cloud computing resources and backbone connectivity between the data centres is provided as part of the cloud service.

AWS CEO Adam Selipsky’s keynote speech reminded us of last year’s announcement that Swisscom would use AWS for its cloud-based 5G core, and that Telenor’s Swedish brand Vimla would also base its 5G core in AWS’s cloud.

A cloud-based core together with Open RAN radios is likely to be the standard approach for enterprise private 5G. If one of the early adopters of carrier cloud-based 5G cores manages to make it work effectively, it may become a widely adopted approach for carriers too.

5G networks: Open RAN

Open RAN continues to be a significant theme. The key idea persists that open interfaces between network components and use of edge cloud and virtualised network functions should drive a more efficient and competitive RAN layer, through allowing components to be sourced from a variety of vendors. The Aetha Consulting team has previously written in more detail about these trends – read the insight article here.

While traditional equipment vendors such as Ericsson and Nokia are O-RAN Alliance members, the main impetus behind Open RAN has come from (predominantly American) technology and cloud players. Mavenir’s message for the industry at MWC, for example, is “one network, any cloud, all software” – pushing expertise in virtualisation to compete with traditional equipment vendors. Again, however, this is not a new development.

New and interesting contributions mainly came from smaller innovators demonstrating how open architectures can be used to maximise network performance. Innovations on the Aetha team’s radar include:

• Using Delay Doppler-based channel estimation and prediction to allow inter-cell coordination on a scale of future milliseconds to optimise cell edge performance
• Optimisation of base station design and feeding to maximise spectral efficiency – something that could generate benefits for operators even in advance of full Open RAN adoption
• Using disaggregated xHaul (both fronthauling from radio units – including small cells – to the base station and backhauling from the base station to the core) to enable a wider range of transport solutions. While not strictly Open RAN, the use of open architectures here could mean network operators do not necessarily have to run fibre to 5G cell sites, using high-capacity microwave instead, or even to the radio unit – though we expect fibre to remain the major technology for both applications.

However, it remains clear that we are several years away from full Open RAN adoption, and there is some scepticism within the industry as to whether it will deliver the promised cost efficiencies from enabling multiple vendors. BT in particular voiced concerns that without additional underlying component providers, there could be limited scope for reduction in prices. It is easy to foresee a situation where the current Ericsson/Nokia/Huawei grouping is displaced by a similarly small number of cloud and software providers without competition in RAN equipment supply significantly increasing.

Private 5G for enterprises

At MWC Barcelona 2019, there was already significant interest in private networks for ‘Industry 4.0’ (aka the Fourth Industrial Revolution) but many of the examples were 4G networks. Times have changed dramatically: almost all discussion is now around private 5G.

In several countries, spectrum has been made available for private 5G to enable enterprises easy access. For example, in Germany the 3.7–3.8GHz band is reserved for licences which enterprises can apply for on a local area basis.

To complement this availability of spectrum, AWS has productised their cloud-based 5G core, making it easier for enterprises to get a 5G network up and running using Open RAN radios and AWS’s cloud-based 5G core. Not to be outdone, Microsoft Azure has announced a collaboration with Qualcomm to also provide a cloud-based core that customers can use with Open RAN radios.

This year’s MWC included numerous Open RAN products aimed at the enterprise user and there were several demonstrations of the high reliability achieved using Cooperative Multi-Point (CoMP) techniques, with devices attached to several base stations. In addition to 5G connectivity, there was a focus on positioning accuracy that could be achieved in private 5G networks – with <30cm positioning accuracy being demonstrated.

The advantages of wireless data connectivity in the smart factory, and other industrial installations, are clear. There were numerous live examples on display of smart factories, fulfilment centres, farms, railways and other industrial installations using private 5G. However, these solutions do not necessarily need to be 5G-based – there were also demonstrations of similar installations using WiFi 6, presumably at a lower price point and without the need to buy a cloud-based 5G core.

The rise of the Metaverse

An emerging theme from this year’s event was the Metaverse, a generally buzzy topic in the wider technology world. That said, the main players in the mobile space are still working out exactly how the Metaverse will develop, with an array of terminology deployed by attendees. Concepts such as Virtual Reality (VR), Augmented Reality (AR), Mixed Reality (MR) and Extended Reality (XR) reflect innovators’ differing ambitions for the Metaverse and consensus on the key applications is yet to be reached.

However, MWC did bring some examples of initial technical solutions for Metaverse-related capabilities, with Qualcomm showcasing a range of ideas:

• Mobile location sensing, using RF from multiple cells to assess the placement and shape of objects in a target area. Combining this with existing imagery creates a digital twin which could be used to insert future Metaverse overlays into the existing area.
• Enhanced gaming on the move and AR wearables, facilitated by using 5G and on-device processing to reduce end-to-end latency to ~50ms. In the current proof-of-concept, the wearable is connected to a handset via Wi-Fi and then via 5G to the cell site. The key latency requirement here is “M2R2P” – motion to rendering to photon.

Perhaps the most eye-catching exhibition was SK Telecom’s Metaverse “rollercoaster” – a robotic arm controlling the motion of users while a synchronised VR headset gave the movements an exciting context. Certainly, the length of queues for this exhibit suggested an exciting future for such applications.

The Metaverse might be heading our way, but defining 5G’s role in the Metaverse is at an early stage and we can look forward to ‘AR’ gradually evolving into the mobile component of the Metaverse.

AI everywhere

AI is emerging across all parts of the service delivery chain. Network optimisation is an area where AI has been promoted for some time, and in addition network monitoring and fault reporting now incorporate AI. Even the device radio modem now includes AI. Qualcomm’s latest modem features an AI processor for mmWave beam control, apparently providing significant improvements in range and throughput.

AI is everywhere – not just in the ubiquitous robot dogs patrolling so many of the exhibitors’ stands. With its promise of generating value for business, AI is here to stay.