Ofcom says trade away to defragment midband 5G spectrum

Ofcom has declined to take action to defragment the UK's 5G midband spectrum. Instead it wants the operators to negotiate trades amongst themselves. Will it work?

As it auctions off a new tranche of spectrum, Ofcom is proposing that UK operators trade spectrum amongst themselves to defragment the 3.4-3.8GHz band. But it will not take regulatory steps – including revoking and reissuing licenses – to ensure that it happens.

Ofcom’s auction statement said: “In the circumstances, we do not consider that it is appropriate or proportionate for us to intervene to guarantee spectrum contiguity for all operators.”

So what’s the problem?


Look at this graph. It shows the awards already made at 3.4-3.6. Now Ofcom is offering 3.6-3.8 spectrum in the right half of the picture. You can see the holdings of Vodafone, O2 and EE in the 3.4-3.6 band are separated from the 3.6-3.8 band by a big red block belonging to Three (via its Uk Broadband acquisition).

If things stay like this, then most operators are faced with a position of having non-contiguous blocks of spectrum within the overall 3.4-3.8 band. Only Three (H3G and UKB) has a more consolidated position. 

That is why they would prefer to unshuffle the deck so within those 400MHz, they can all be allocated contiguous blocks. Contiguous blocks help because you may need fewer active antennas on the masts – as you are not serving different bands, and you can also maximise traffic capacity more easily. All the operators except Three, which holds the most spectrum in this band, asked Ofcom to use this auction to carry out a fuller reassignment and defragmentation of the wider 3.4-3.8 GHz band. Vodafone and O2 argued that it was Ofcom’s legal duty to ensure that operators have 80MHz of contiguous spectrum. Ofcom refuted this.

This situation is why some operators initially asked for Ofcom to auction the whole block together, rather than in two stages. But Ofcom said that it wanted to get ahead with things, so that rollouts could get started. That’s because it will be 2022 at least before spectrum in 3.6-3.8 is nationally available (although it will be cleared in some areas.) 3.4-3.6, however, was immediately available.

Ofcom acknowledges that “As a result of the current allocation, absent shifts of existing frequency assignments, mobile network operators’ holdings in the 3.4-3.8 GHz band after the auction will be fragmented.” But it is not proposing to defragment holdings across the band. It said that to do so would require too intrusive a regulatory action – i.e. making significant changes to existing licenses.

Initially, the regulator said that while contiguity would in theory be better, proximity might be enough. It said that the largest spectrum holders should be granted spectrum in the middle of the band, with those only buying smaller slots sorted to the edges. It still wants to limit those with less than 20MHz to bands at the top and bottom of the allocation. 

Additionally, Ofcom proposed a trading mechanism to enable operators to unshuffle the deck so they can sort their suits together.

Now Ofcom wants to see this spectrum trading go ahead, and is designing in the ability for operators to sort of their trades within the auction process. It has decided there will be a four week negotiation phase between operators during the auction.

One potential stumbling block that some have raised is that Three will have no incentive to trade. It sits with the most spectrum, mostly contiguous – why should it move aside for the others?

Ofcom counters that Three may well trade some of this band in order to get access to sub 1GHz spectrum, where it is much weaker. It also said that Three “wrote to us confirming that its willingness to trade has not decreased as a result of our decision to vary its 3.6 GHz licence and set up a process last year where it attempted to arrange trades of its 3.4-3.6 GHz spectrum before the auction. We consider that these actions provide support for our view that it has incentives to trade.”

So we will enter a new auction process for this set of bids – with operators making an initial grab, and then starting to trade amongst themselves to achieve their desired outcome. Will some buy smaller slots just to frustrate others, or to give themselves leverage for other goals? Will they decide that in the end, they will make do with separated slots? We will find out.