Scrambling for spectrum: What to do when we run out

Snatch it off the military or build more base stations?

ITU Telecom World The lack of radio spectrum is probably the most debated subject amongst the delegates at the ITU's annual talking shop, with the military, emergency services and broadcast TV all in the firing line as punters want more data.

Mobile data consumption is rising, and the popular airwaves are getting full, but strategies for addressing that vary widely. Snatching spectrum off the military is a popular option, but then the military isn't here at ITU Telecom World to defend itself. Emergency services are also being fingered as wasteful consumers of spectrum and broadcast TV is on the chopping block, but consensus is that we need more base stations, at lower power, to properly eradicate the wasteful white spaces.

When it comes to stripping the armed forces of their bandwidth, the ITU has form. The 2.1GHz and 2.6GHz bands were allocated to the military in most countires but are now recognised around the world as telephony bands (3G and LTE respectively). In the UK, Ofcom is already working to bill the Ministry of Defence for its allocated radio spectrum, to encourage it to hand over what's spare, but ITU delegates were also promoting a more dynamic approach.

The military doesn't use most of its bands, most of the time, as demonstrated by the plethora of secondary users already lurking in the knowledge that they can be kicked out by the army at any time. The rules limit what those secondary users can do, to avoid them interfering with military operations, while the military has free reign to stomp all over them at will, but a more cooperative approach could release frequencies without preventing military use.

At its most basic that means the army calling up when it fancies doing some manoeuvres, requiring secondary users to switch off. A more advanced approach might involve an online database with which devices would be required to register, and regularly check back, to ensure the MoD didn't want the bands.

The same thing could be applied to the emergency services, but without the necessity for a database if the technology can be standardised. LTE includes priority at a packet level, so a responding police car can override other phone users to stream video of the fleeing miscreant, without the need for dedicated infrastructure or frequencies.

The problem is those other phone users, who may object to having their calls cut off. America has been trying to build a network along these lines for years, but companies have been remarkably reluctant to invest in a network which will have to be sold to customers on the basis that they are second-class users.

There is more bandwidth available in the television broadcast channels, but we'd have to give up broadcast television to get at that. Switzerland's regulator told the assembled crowd that it plans to do just that, thanks to Switzerland's pervasive cable penetration, and Ofcom has made noises about it being an eventual aim. Even the chair of the DVB Consortium accepted that it would happen eventually, though pointed out that broadcast is still the most efficient way to send video to more than one person at a time.

But all those bands are useless without cheap equipment to take advantage of them, and that needs international harmonisation. RIM's VP Mark Pecen was on hand to remind everyone that a handset capable of using the 41 different bands already recognised for LTE employments would be physically huge, and power consumptive, while one that just used the 2.6GHz band (which is recognised almost globally) need be no bigger than today's devices.

Limiting ourselves to internationally harmonised existing bands means squeezing more data through, and that means smaller cells and a lot more of them. A single cell will have a specific capacity. Reduce the power and put four cells into the same space and you have four times the capacity without using any more spectrum. The Wi-Fi standard is a lot less efficient than 3G, but Wi-Fi networks carry a lot more traffic because they are very small and consequently reuse frequencies much more, network operators deploying small cells could cope with the predicted rises in data consumption, if they were allowed to.

So said Alcatel-Lucent, coincidentally makers of the LightRadio cloud-based architecture which reduces the cost of tiny cells, assuming one can get every cell connected to a fiber-optic backhaul for low-latency command and control. Alcatel-Lucent reckons regulation is what's preventing the deployment of high-density networks, which could go a long way to resolving the spectrum shortage.

None of this is going to solve the problem immediately, but fortunately lack of radio bandwidth is not an immediate problem. Today most people are more concerned with coverage than connection speed, but the ITU has to think ahead and consider how increasing amounts of data are going to fit into static quantities of radio spectrum over the next few decades and beyond. ®