Possible UK auction delay would be major blow for WiMAX

The UK 3G operators have no particular desire to see an early auction. If BT (or another bidder from outside their ranks) does win spectrum, they will have to face yet another competitor in a hugely pressurized market at an early stage. And they do not have an urgent need for 2.6GHz themselves yet--most are just rolling out HSPA and the 3G networks are not at capacity, and are performing pretty well in terms of most mobile data services. Since the five cellcos will almost certainly use LTE if they get 2.6GHz licenses, they will not be able to start build-out for at least another 18 months. In the meantime they will hope that economic uncertainty will prove a blessing in disguise, at least in terms of driving down spectrum costs, and that they will be able to assist that downwards pressure further with tactics like network or spectrum sharing deals.

For WiMAX vendors and other supporters, a delay in this flagship auction would be a major blow. To get any inroads into mature mobile markets, WiMAX needs to exploit its headstart over LTE, especially since the two technologies are very similar in what they actually deliver, and LTE is a more natural choice for 3G operators (commercially, if not technically). But the advantage of having systems ready to purchase, probably 18 months ahead of LTE, is pointless if potential customers cannot get their hands on suitable spectrum. If the operators have to wait for a year or more for auctions, the playing field with LTE will have levelled, even for non-3G carriers.

The UK has been a beacon of hope for WiMAX in the west European UMTS heartland. This is because, unusually, the incumbent telco, BT, has no mobile arm, which makes it likely to bid for a license. If it does and it wins, BT is likely to adopt WiMAX rather than LTE to take advantage of the headstart WiMAX would provide. BT has made it clear it will drop out of the race if prices go insanely high, as they did during the 3G auctions at the turn of the century, but it is still widely expected to acquire a license because wireless would enhance several of its businesses. While it would be a herculean task to launch a successful retail mobile brand against the five existing ones--especially since several of them are now BT customers on the backhaul and convergence front--BT could use an advanced mobile broadband network to support enterprise multimedia services; for fixed/mobile offerings to enhance its fixed broadband business; and as a wholesale network for innovative providers, media players, public sector initiatives, and any cellco that does not gain its own 2.6GHz licenses.

As BT rolls out its 21CN all-IP network, it will be impatient to know whether it can include broadband wireless in its service plans for 2009 onwards, and an auction delay would be a frustration. It would also make it more likely that, if it has to wait, BT would decide to go with the LTE mainstream and simplify aspects like roaming with other European carriers.

T-Mobile's argument revolves around the new willingness of many European regulators to allow GSM spectrum in 900MHz to be "re-farmed"--turned over to 3G or 4G technologies as GSM usage declines. This would increase the capacity available to operators for data services, without the need to purchase new spectrum; would presumably avoid any danger of the licenses having to be returned once GSM was turned off, and potentially falling into the hands of rivals; would make it more cost effective to provide 3G and 4G services to sparsely populated areas, because of the longer range and better indoor penetration of low frequencies; and would support a multi-frequency LTE roll-out, with high capacity hotzones in densely populated or high ARPU areas, combined with 3G or, in future, LTE offerings in 900MHz for rural areas. This multi-frequency approach is gaining increasing traction as multiband devices become commoditized, and the pressure to reduce network costs to support flat rate broadband becomes intense. 3G proved that it is commercially non-viable to build a high performance network in frequencies over 1GHz for an entire population, requiring huge numbers of cells, many of them covering very few active users. Not only would refarming allow desperate cellcos to fulfil their rural coverage obligations without a negative effect on the bottom line, but it would enable them to move more quickly towards 4G where demand is seen, since they already hold the licenses. The WiMAX vendors have been advanced in supporting multi-frequency deployments, and in making 802.16e available in different, often unofficial, bands. At the recent WiMAX World EMEA event in Munich, for instance, Telsima and its customer Mobilink Slovenia showed off a WiMAX deployment in 450MHz for rural and 3.5GHz for urban or high capacity areas. There is no reason to think the LTE community will not grasp the same opportunity and come up with 900MHz implementations for cellcos that are being allowed to refarm this spectrum--a trend that has taken off mainly in France and Scandinavia so far.

All this is clearly leading T-Mobile and others to question whether they need to pay large sums for 2.6GHz if they already have a store of spectrum that can be opened up for mobile broadband. The operator argues that it cannot accurately value the 2.6GHz licenses until it knows how much 900MHz it can use for 3G. Not that refarming will negate the need for 2.6GHz, since low frequencies are poor at supporting urban areas, because of interference and low capacity. But it would make some operators prefer to buy 2.6GHz only for high value urban zones, rather than nationwide - and so may put pressure on some regulators to offer regional licenses, which would then attract higher values for urban zones (but risk very low prices in rural, especially in countries that have large rural spaces and poor rural populations, which is generally not true of the UK).

The broader issue that is highlighted by the T-Mobile argument, however self-serving that may be, is that traditional auction rules for wide area mobile spectrum are becoming irrelevant to mobile broadband. Operators are calling for more of the spectrum to be allocated for TDD operation, because this is better at supporting data and packet services (and LTE will have a TDD flavor soon); or at least for the regulator to allow the market to decide on the FDD/TDD split, as Ofcom will. And they want more flexibility on regional licenses, so that the spectrum they purchase actually maps onto a roll-out plan that will focus initially on spot developments in high demand areas, with 900MHz or other bands filling in the gaps. In such a scenario, the national license becomes an unwieldy burden.

Ofcom said last month that it would rethink its plan to support refarming because it had received far higher interest than it had expected in the 900MHz frequencies, hence the delay in the existing cellcos being able to make firm plans for using these airwaves for 3G.