Analysis So, the UK's 4G auction is over, but questions remain: who bought what exactly, why did they pay so little, and, most importantly, when can we expect some 4G goodness?
The hammer has fallen on Blighty's airwaves, but the exact frequency allocations haven't yet been decided: winning bids were for blocks of spectrum within specific bands, but the exact slots now have to be worked out between the winners over the next few days. Once that's done the networks can start deploying some infrastructure, or switching on the kit they've already deployed.
We do know which block O2 owner Telefonica bought: 811-821MHz paired with 852-862MHz, as that pair comes with an obligation for the owner to roll out 2Mbps mobile internet connectivity to 98 per cent of the UK population by the end of 2017. That means at least 95 per cent of all the people in the United Kingdom's constituent countries, rather than, say, everyone in England and a handful in Northern Ireland.
That doesn't mean the Highlands will get 4G: instead they'll most likely get the 3G that Telefonica already deploys in its holdings at 900MHz. This tech, which is more than capable of the aforementioned minimum speed, can be switched on at existing sites, and should be enough to change the lives of many out in the countryside.
It's strange that Telefonica didn't buy more high-frequency bands, as Vodafone did. Such bands aren't so good for sparsely populated rural areas, but are ideal for tightly packed cities: high-frequency signals don't propagate very far but that doesn't matter because each phone mast can only handle a certain number of users in its cell area anyway, and this capacity is more likely to be a problem than the distance to the nearest mast.
Vodafone certainly thinks those higher bands will be important, and spent more than anyone else (almost £800m) securing a huge chunk at 2.6GHz along with some unpaired spectrum in the same band and the same amount at 800MHz as Telefonica - only without the coverage requirement.
Three didn't bother with the high end either, but that's because come October it takes possession of 20MHz at 1800MHz that it bought from EE a few months ago, with a neighbouring 10MHz being transferred two years later. That gives Three enough spectrum to continue competing, just, but it's unlikely to launch 4G LTE before that first block changes hands.
Nor will Three be making a big deal of 4G deployments, not least thanks to a corporate brand name that risks dating it. Three will be using words like "super fast" and "mobile broadband" rather than splitting hairs about which wireless variant offers slightly better speeds. This is probably for the best, given the way the term "LTE Advanced" is being thrown around these days.
Overall there's not much that's changed following the auction. Telefonica's decision to play it cool has left it behind slightly, but other than that, spectrum ownership remains with the same players in roughly the same proportions:
|Mobile operator||Total Bandwidth Today||Total After Auction||Under 1GHz (for better penetration)|
The most interesting winning bidder is BT, through its subsidiary Niche Spectrum Ventures. We had expected BT to pitch for some low-power bands in which to run campus-wide networks, and that's probably what the two failed bidders (MLL and HKT) had in mind. Now BT has ended up with enough bandwidth to run a national network should it so wish.
For comparison, Three provides something akin to national coverage in 30MHz of spectrum at 2.1GHz - and once the auction shakes down, BT will hold 50MHz slightly up the dial at 2.6GHz. But BT has said publicly it won't build a national network, so what does the former Brit monopoly have in mind?
We only have a brief statement to work from, but apparently the spectrum will enable BT to "provide its business and consumer customers with an enhanced range of mobile broadband services" and, critically, "building in its existing strength in Wi-Fi".
That strength in Wi-Fi is, in part at least, achieved by putting a publicly accessible network hotspot into every T Broadband router issued by the telco. BT took a lot of stick for that, and some customers opted out, but it's left BT with 4.5 million wireless internet points across the UK from the mountains of Scotland to the Isles of Scilly, and it takes no great leap of imagination to see how those Wi-Fi sites might become small 4G LTE masts - femtocells in other words - using licensed frequencies.
BT could broadcast at much greater power in the licensed band, although it's unlikely to ramp up much in a domestic setting, but such a network wouldn't suffer interference endemic to the 2.4GHz band where Wi-Fi lives, and the handoff procedures of LTE would make for seamless roaming too.
A low-power licence may have been enough for that, but it wouldn't have allowed BT to cover airports or similar locations where BT already has decent backhaul to the network backbone in place, and wouldn't want to pay roaming rates.
So BT could offer in-home LTE, backhauled over broadband whenever one was near a BT Broadband customer or OpenZone spot, and roaming to one of the operators the rest of the time. With the right pricing that could be a very competitive offering, and something the network operators should be concerned about, which is good as there's nothing else in these auction results that would concern them otherwise. ®