In late November Google announced its participation in the final pre-launch funding round for O3b, a satellite network backed by SES and intended to provide broadband access to the developing world. The name refers to the “other 3 billion” – those of the world’s population without broadband connectivity – and the plan is to launch a constellation of Medium Earth Orbit (MEO) satellites that will provide low latency broadband to the area roughly between 45° N and 45° S.

Google was one of the founding backers of O3b, back in 2008, when it confidently expected to have its satellites in orbit by 2010. The first launch is now scheduled for 2012.

I mention this slipped schedule not to be snarky – I’ve missed the occasional deadline and so have you – but to recall the problems of a previous bold satellite venture, the Iridium system. Iridium was going to provide seamless international mobile coverage to business travelers, who would need only one phone and one phone number around the world.

  • This was a great idea in the late 1980s and early 1990s when planning for the system started. A lot less great in 1998 when Iridium finally went commercial: vastly increased GSM coverage and global roaming that actually worked took a lot of value out of the Iridium value proposition.

Will this happen to O3b? Not necessarily, but the global broadband gap is narrowing.

  • Until recently, the African continent was virtually cut off from the global Internet. Now, a frenzy of submarine cable building has people seriously talking about a repeat of the Great Atlantic Bandwidth Glut. True, this only means great connectivity at the landing points; getting bandwidth 500 km inland will be a challenge for a while – and an opportunity for O3b.


  • 3G service is growing in emerging markets, and is frequently used as a broadband access technology. Coverage is still limited, of course, and is largely unavailable outside urban areas.

O3B promotes the potential of its service as a backhaul technology, and in this it may find more traction than it does as an access method. There are complications, of course. Among other things, the fact that MEO satellites are not geostationary means that any uplink has to have active tracking antennas and some way of handling satellite to satellite handoffs.

And that, as my grandfather said about the automatic transmission, is a lot of stuff to go wrong.