Last week, newcomer O3b Networks announced it was launching a new global network of 16 low earth orbit (LEO) satellites, carrying 2,133 transponders and designed to offer IP and cellular backhaul for telcos and ISPs across the developing world. As was the case with a recent trans-Pacific submarine cable, most media coverage has swept past any other impact to focus on the fact that one of the new venture's backers was high-tech celebrity player Google. We don't necessarily disagree that "if Google did it, it's news," but let's look at the news on its own merit first.

Certainly, this is a significant announcement, comprising a substantial new satellite network backed by some known and unknown players - namely, Liberty Global, one of the world's largest cable operators, and HSBC, the financial giant with a history of self-provision for its IT needs. The release speaks of access to speeds up to 10 Gbps, with a total aggregate capacity of 160 Gbps. Most importantly, Gartner has learned that these will be Ka band satellites, which puts O3b's network infrastructure decision on the leading edge of the latest technology for these underserved regions. The use of Ka band has the potential to impact price, available capacity and network architecture for most of the regions announced for coverage.

Although it purports to serve "Asia, Africa, Latin America and the Middle East" and the founders speak of "the next 3 billion," it is unlikely that this will be a global network to start with. O3b quotes five satellites as needed to provide "global" coverage, but that's more likely a bare minimum to blanket the southern hemisphere. At any rate, the 16 satellites announced in this launch pale in comparison to Globalstar's 48 LEOs and Iridium's 66. We tentatively believe this "Asia, Africa, Latin America and the Middle East" network will focus its first phase on Africa. Such a focus would make sense, as this is easily the least penetrated telecom region in the world (and the fastest growing). The impact on pricing would be significant, assuming that O3b and its partners can bring the vision of a data/backhaul network to fruition on LEO-based satellites. The use of such a network for data is the key differentiator - the other incumbents may also be planning upgrades to capacity with this in mind, but only time (and perhaps the pressure brought by a successful new entrant in O3b) will tell.

With several new submarine cable projects under way for the region, which in total could dectuple the available bandwidth, satellite needed a competitive response. The immediate result for end users should be very beneficial in terms of lower prices, increased choice and access to new applications. Africa has historically been capacity-starved, and the easing of bandwidth constraints will bring not just more of the same old service to a larger base, but it will also enable the appearance of new services and applications that previously could not have been considered. Mobile VoIP is only one prominent example. Importantly, all the new cable construction has fostered growth of the open-access principle. This may finally displace the proprietary ownership and distribution of capacity by incumbent providers, with higher hopes for transparent pricing and market-based rules.

Now, back to Google. The search giant, which was rumored in recent weeks to be talking with carriers and other partners about a submarine cable, has now backed a satellite play for the world's underserved regions and populations. We believe Google's aim is largely to commoditize carrier strengths in the network, and reach users via its search portal to provide them with applications that keep them on Google's servers. Its vision is one in which folks with Internet can answer all their important needs through Google - Google has worked with carriers in the past, but overall, it seems to prefer a future in which operators don't differentiate against off-network traffic and are content to take their access fees like a good utility and stay out of the application arena. The open-access discussion we just referenced plays directly into this tactic. If Google can make even one dollar in increased ad revenues per user, using its search and location technologies to provide attractive targeting, then the population of Africa alone will make a large dent in the $100 billion target the company has set for its growth.

For more information, consult the Next Generation Satellite technology profile in "Network Service Provider Infrastructure Hype Cycle 2008".