"It will snow on Tuesday,” the driver told us on the trip in from Barcelona airport. And so it did.
For a few minutes, the MWC faithful abandoned their booths to record the fluttering flakes with – this being MWC, and also 2018 - their phones.
The week passed as predictably as the weather forecast. We had 5G demos, lookalike handsets, drones, AI and IoT.
Compared with previous years there were fewer cars and smart watches, but more robots.
But even in this ocean of predictability a few items stand out.
China Mobile certainly laid down some 5G markers. Its trials will be the biggest, its rollout will be one of the earliest and its commitment to standalone is a signal of intent. It's swinging for the fences.
But it’s a strategy not without risk. Of all people you’d think China Mobile, after its unfortunate standalone TD-SCMDA experience, would be alert to it.
The TD-SCDMA problem was mostly a handset issue, and as it happens the one potentially promising storyline of the week was the 5G handset.
Analyst Richard Windsor has noted mmwave has a propagation problem. It's especially severe in metal casings, but it's an issue with any material. No one has ever put such high spectrum radio in a small device before.
MediaTek told me one part of the problem was the performance specs aren't clear. Like, how much power should it draw and what performance level should they build to? They're waiting on 3GPP to clear those up.
But they also showed a more basic problem: you can block the signal with your hand. Even for multifunctional 5G, the human-carried device is a pretty big use case.
In a deployed network, the workaround is to default to the LTE network – though that’s not available if, say, you're going standalone 5G.
The engineering solution is ‘path diversity,’ or putting an antenna at either end of the device.
Yet that adds to the physical engineering problem.
Your 5G device will have to find room for 5G sub-6GHz, mmwave, LTE, UMTS and perhaps even GSM, not to mention WiFi, Bluetooth and NFC. How to make room?
Qualcomm claims to have it solved, which may be why their chips have been picked up by half a dozen OEMs.
But we won’t know until we can get hold of the devices performing in a field trial. So watch this space.
A different antenna that caught my eye was the Metnet worked up by UK firm CCS.
Not quite as sexy as a 5G phone, but it has the advantage of actually working.
It's based on a smart but demanding 30-year-old technology called S-TDMA (‘S’ meaning ‘spatial’) which is only now possible because of the amount of processing grunt required.
It can deliver 1.2Gbps in 28MHz spectrum and can also be deployed at 60GHz.
Its strength is it is self-organising. You can install it to backhaul data out of a dense urban environment and it can autonomously identify the optimal paths available.
If you have to add more cells for capacity, you still don’t have to rejig your spectrum planning among those cells – a huge opex saving. With the coming tsunami of small cell deployments, this looks like a winner.
Among the big vendors, AI was probably a bigger talking point than 5G, as it too is an actual thing.
Huawei’s intent-driven network is a typical example. It creates a digital profile of the user on the network to help optimise predict and optimise network resources.
Like a lot of technology coming out of China it’s both impressive and scary in equal measures. It's helpful for managing a network, but equally useful in tracking users.
Yang Jie (left) and his cubes
Yet even a big dog like Huawei can’t help getting drawn into the dumb things that its customers do.
For its own reasons, China Telecom held an audience hostage for two hours while it declared the opening of a Business Joint Innovation Centre.
Under questioning, China Telecom chairman Yang Jie acknowledged there was no centre, no staff, no investment, and that the ‘joint’ innovation is being carried out by China Telecom and Huawei in their own way.
It didn’t help that his people decided to call this ‘Rubik’s Cube’ innovation, apparently in the belief that customers love time-consuming, impossible-to-solve puzzles.
Not even Chinese journalists stayed to the end of that snoozefest.
I spied Yang Jie himself toying with his phone during one of his colleague’s monologues. You could've predicted that.