Original URL: http://www.theregister.co.uk/2008/02/21/google_balloons/
Google to build 80,000 foot radio tower?
The truth hertz
Google's quest to deploy wireless networks without all that mucking about with steel towers has led its roving eye to settle on Space Data Corp. The company, which provides wireless connectivity via hydrogen-filled balloons floating around the southern US, has caught Google's interest according to reports in the Wall Street Journal.
But the search giant isn't alone in this mission. Even with the existing 2G, 3G and Tetra technologies the UK has over 52,000 cell sites, connected by more then 40,000 licensed microwave connections. According to AntennaSearch the USA has nearly two million antennas (some of which will be on the same tower), so anyone looking to deploy a new nationwide network is going to find locating their sites problematic to say the least.
However, with the 700-MHz auction trundling on in the USA and parts of 900MHz becoming available in the UK - not to mention the rest of the digital dividend - anyone buying up a significant chunk of spectrum is going to have to put up an awful lot of new towers, or find some other way of reaching out to their customers.
Anatomy of a cell site
Cell sites typically consist of six directional panels, each covering 60 degrees (sometimes only three, each spanning 120 degrees), and a circular dish for microwave back haul. At the bottom of the tower is a small shack, or box, with half a dozen rack-mounted transceivers and baseband controllers. Adding 3G to a site means replicating that, and towers used by more than one operator also tend to have a complete replication of the kit.
In the UK, just under a third (16,870) of sites host multiple technologies, or companies, with over a third (22,602) of the remainder being 2G only, and just under a fifth (9,644) being dedicated to 3G (the rest are Tetra). It would be interesting to see how that would change given Vodafone and Orange's recent move to share 3000 sites - but we're unlikely to find out, as last year the operators stopped voluntarily providing information. Ofcom remains "optimistic" that they'll eventually come round and update the info, but until then the information we have access to become more out of date every day.
Sharing sites is complicated by the need to talk to each other when you need access - the same thing that prevents bolting cell phone receivers onto electricity pylons - in addition to the understandable reluctance of cell-phone engineers to go climbing around pylons (interference can also be a problem, but one that can be mitigated).
Floating a New Idea
One of the problems with cell phone base stations is they don't want to be particularly high up, especially in busy areas. Greater elevation increases cell size, which is great in the countryside; but since capacity is on a per-cell basis, in most cases the cells don't want to be more than 20 metres up, which can make siting them difficult.
Squeezing GSM transmitters into a pregnant lamp post is one option - though most upright rods you see at the side of the road are actually sewer chimneys. Pretending they are trees is another, but the required network equipment generally gives the game away. Once we talk about putting up antennae for WiMAX, and all other services looking to fill the digital dividend, we're going to have to find smarter places to put our aerials.
Which is where Space Data Corp comes in: it owns some spectrum in the 900MHz band, and rather than build a network of transmitters it just attaches its cell sites to hydrogen balloons and pays local farmers $50 a time to launch them, usually one every day keeping around ten covering the whole southern US from a height of around 24km (between 80 and 100 thousand feet). The balloons last a day, after which they pop and the kit parachutes down to be collected by hobbyists, with GPS equipment, who get $100 a time for returning the $2,500 equipment package.
When he we asked Google if it indeed has an eye on Space Data, it didn't respond. But its interest isn't surprising, especially if the search megabeast is serious about grabbing a chunk of 700MHz; but the more meteorologically-inclined will have noticed that Space Data Corp.'s model requires a reasonably predictable wind pattern, making expansion into the UK - or indeed deployment beyond the southern US - unlikely.
Over the years we've seen various plans for aerial broadband, encompassing everything from remotely-piloted blimps to manned aircraft circling day and night, but outside from Space Data Corp.'s success the rest of the world is still relying on steel towers with antennae bolted to them. So most likely we'll just have to expect to see a lot more of them in future even if, in the UK at least, we'll have no idea what they're doing there. ®