EE says that O2's 5G lite claims are 'misleading'

EE has been quick to respond to O2's claims that pre-2020 5G will lack critical features

Although he admits that the 5G network EE is launching in 2019 won't use the ‘final version' of the technology, Howard Jones - head of network communications at the operator - thinks that O2's recent claims about "5G lite" do not tell the full story.

Some background: Two weeks ago, BT (owned by EE) CEO Gavin Patterson announced that the telco would launch a 5G network in 2019. Then, last Friday, an O2 spokesperson told 5G.co.uk that any network launched that early would not include the full capabilities of the next mobile standard.

O2 justifies its claims by saying that the necessary standards enabling all of the 5G goodness, like the latency and speed to support fully autonomous vehicles, won't be published until late 2019 or 2020.

EE is, unsurprisingly, a bit miffed that its competitor is (also unsurprisingly) trying to downplay its announcement. The company got in touch with us to clear up the confusion.

‘5G lite' is misleading - Howard Jones, EE

Jones said that, while EE's 5G network will not support everything, it will still be a significant upgrade to speed and latency that consumers experience today - and further benefits will come in time:

"In reality, no mobile standard-based technology ever launches complete - it never has done. If you look at 4G from when we launched it in 2012 to now, it's completely different...in terms of things like coverage, capacity and improvements to voice. Voice calls were carried over 2G and 3G at launch… We introduced [4G voice calls] in 2016. It's completely normal for any technology to evolve from launch."

O2's argument is based on the fact that 3GPP Release 15, expected in June, will not cover 5G's full capabilities. However, Jones argues that this is not the case.

"Release 15 does include a full 5G core: it's Option 7 of 3GPP Release 15. In reality what they [O2] are saying is not ‘Release 15' or ‘Release 16'; it's standalone vs non-standalone. Do you run a 5G radio network through your 4G core, or a full 5G core?"

EE is going for the former approach at launch, known as a non-standalone network (Release 15 Option 3). Jones expects most operators through 2019 and 2020 to take the same route, but move towards a full 5G core in time.

"I expect full capabilities by the mid-2020s. [5G] will keep evolving as new capabilities come in… ‘5G lite' is quite misleading."

Jones was keen to highlight the differences between what EE thinks will be the main consumer reason for adoption - mobile internet, which the faster speeds and lower latencies will benefit at launch - and industrial uses. These include IoT at scale, driverless cars and similar application that will require an evolution of 5G.

"It was latency that drove the customer experience when moving from 100ms 3G to 30ms 4G. But 20-30ms isn't good enough for multiplayer mobile gaming, or VR and AR. It's the difference between a phenomenal experience and one that makes players feel sick."

EE's network will have a latency of between 20ms and 30ms at launch, while future 5G could go as low as 0-5ms. Jones expects a drop to sub-10ms in 2021/22, and "almost zero latency" in the mid-2020s. At this point the necessary small cell infrastructure will have been built (he hopes), and companies will have started to use technologies like Massive MIMO and millimetre wave (mmW).

It is likely that most operators will use a mixture of technologies (non-standalone) to get their 5G signals to end-users in the network's early days; and who knows how long that will last? We're now six years on from the launch of 4G, and the standard is still not prevalent in rural areas.

Will 5G approach ubiquity any faster? It's doubtful - the infrastructure alone is a much more significant investment than 4G was. The technology is likely to be limited to specific areas (cities) and applications for many years to come.