Google is one of the many investors who together have put $60 million into O3b Networks, a St. John, Jersey, Channel Islands-based startup that is looking to offer Internet services in the emerging world, especially areas that are far away from the sub-sea networks and major backbones. Apart from Google, other investors include HSBC Holdings, Allen & Co. and Liberty Global, according to the Wall Street Journal. It is not clear how much money Google (s GOOG) invested in the company. The company lays out its reasons in a press release, but that I am skipping because it doesn’t really say anything.
One thing is clear: O3b Networks will need a lot more than $60 million. The project, the brainchild of Greg Wyler, is going to cost $650 million and will require 16 satellites; the service is due to start by the second half of 2010. Wyler apparently has a lot of telecom experience in Africa. Liberty Global, a company owned, in part, by legendary media mogul John Malone is going to help develop the project.
Given how many of the previous attempts have been non-starters — whether for economic reasons or simple technology constraints — the claims by this company are quite audacious. Here are some facts about their service:
* The service is strictly wholesale and will sell bandwidth to local ISPs, fixed-line and mobile providers for, say, cellular and WiMAX backhaul.
* The company is also going to offer Google Apps to its potential customers.
* They claim speeds of up to 10 Gbps and low latency. They can do that for three reasons:
* They have a dedicated Ka-band satellite and as a result, the large available Ka-band spectrum can help deliver bandwidth at speeds of Gigabit/second and higher.
* O3b Networks uses parabolic antennas, which reduce latency.
* Their birds — 16 of them — are going to be positioned at 8,063 kms from the Earth, which allows them to add new satellites.
* The coverage zone is between +/- 40 degrees of latitude.
I’m intrigued by this startup because it does make sense to offer connectivity in remote areas. It also makes sense because Africa is one of the booming cellular markets and one where there is a need for cellular backhaul infrastructure. In remote areas, voice is going to be the killer app for a long, long time. The problem is that this company will always compete with fiber networks in terms of pricing, and that might put them on the back foot.
25 thoughts on “Google Invests in Satellite Broadband Startup”
There are a number of “content distribution to the edge” scenarios where multicast satellite downlinks are more efficient than fibre (assuming that fibre is where it needs to be, without additional last-mile connectivity to fibre PoPs).
If Android phones are going to dual-mode (GPRS/3G and WiFi), a much more viable content distribution strategy to mobile handsets would be via SMS together with WiFi hotspots/hotzones (with mesh WiFi to satellite backhaul), with some kind of local content servers/caches at the satellite points. SMS could be used to alert users (via cell-ID tracking or even GPS) as to the content available at the hotspots/hotzones.
Remember that the vast majority of mobiles in Africa are on pre-paid usage, and pre-paid GPRS/3G data-usage is still a very new thing. WiFi backhauled by satellite solves a lot of problems for Google, especially having to deal with the local mobile operators. A lot of these mobile users in Africa are already big on Opera Mini.
Since the longer-term potential for Android is for (lots of) low-cost handsets from China (which will find their way to Africa), I can only see good from this Google investment. Google might also do well to sponsor development of Android applications in Africa.
I don’t see how they would run into competition with fiber in “areas that are far away from the sub-sea networks and major backbones”. If the areas have fiber, wouldn’t it mean that they are close to a major backbone (or that a major backbone could be built using cable)?
This is very interesting news for african people but i am afraid that project would failed just like 100 dollar laptop project…
I am more hopeful about the venture as a whole.
There is considerable evidence, at least in India, that the areas seen as remote by urban people are doing much more than ‘voice’ over the web. Mobility will enhance that experience.
Operators not being keen to lay fibre is one of the reasons why these areas are still seen as ‘remote’. The African experience is that the absence of such fixed infrastructure catalysed the wide adoption of mobile. A mobile web offering therefore makes sense for Africa too.
I do however think that Google’s strategic interests in this project need closer scrutiny, on which I wrote a post. I linked yours to provide a good perspective too.
Google became self-aware at 2:14am EDT August 29, 2009.
How soon we forgot Iridium . Eienstien was right there is no match to human stupidity
Interesting..we just saw another report about Google sending a satellite into space to collect user’s data as opposed to helping “emerging worlds” get their service. Funny at the crazy conflicting reports I read on a daily basis!
There are some issues with this article. Mostly centered around the bullet points. Some of it is simply formatting, but some of it makes very little sense the way it is written.
The low latency will come from the orbit that is chosen, less distance to the satellite will mean less latency. However to increase bandwidth you need to use directional (versus omnidirectional) antennas and therefore require parabolic dishes that actively track the satellite. These are orders of magnitude more expensive to purchase and operate than fixed VSAT antennas. The orbit may allow for additional satellites, but as you add additional satellites complexity for routing increases exponentially.
In response to Prashant Singh’s comment regarding Iridium, this project is more an evolution of existing VSAT technology. The significant difference is that the satellites will be communicating with fixed ground stations that will most likely use multiple dishes for continuous operation rather than mobile handsets with omnidirectional antennas.
It still faces the fundamental problems that Iridium did in both the political quagmire of offering services in multiple countries and the long development and deployment time for satellite based systems having a tendency to make them obsolete at or near service launch.
I’ve written a critical review of the project at http://tinderblog.wordpress.com/2008/09/09/google-broadband/
“I think that their US$ 750 million investment in satellites will be useless within 20 years of launch due to degradation (NYT claims a 10-15 year life for MEO satellites), whereas building more land capacity in Africa would have much longer-lasting benefits.”
“Google clearly wishes to use this project to enable broadband Internet access in developing regions, but many other things must be in place, including fixed power infrastructure, PCs or OLPCs, technical support and skills, and demand and useful content and services for areas with lower literacy, before that can happen.”
“They have a dedicated Ka-band satellite and as a result, the large available Ka-band spectrum can help deliver bandwidth at speeds of Gigabit/second and higher.” I believe that Ka-band has issues with rain fade, and the weather in Africa is often bad and likely to get worse, particularly at times when people might need a phone service for emergencies
“O3b Networks uses parabolic antennas, which reduce latency.” I think the key to reducing latency is the low orbit of the satellites, and I can’t see how the antenna shape makes any difference.
“Their birds — 16 of them — are going to be positioned at 8,063 kms from the Earth, which allows them to add new satellites.” You can add new satellites in any orbit, and I don’t see any plans for more than 16 of them.
Even if it is a LEO (or MEO, who knows) not a GEO system without spectrum – O3b Networks is just a pipe dream. I remember in the 90’s when Teledesic sponsored by Mr. McCaw and by another billionaire Mr. Gates, it took them many years to convince the ITU and the GEO satellite operators that they future LEO Teledesic satellite system will NOT create a harmful interference to existing and future (with priority ITU filing) GEO systems through elaborate beam switching scheme on-board of each LEO satellite to prevent interference to GEO systems. I believe this cellular entrepreneurs from O3b are for rude awakening to the process called ITU satellite coordination (see a recent ProtoStar 09/05/2008 press release: ProtoStar and Russian Intersputnik Finalize Joint Orbital Location-Services Agreement). With all my respect to Google, HSBC and Liberty Media, I remain skeptical, even if they have a protected (coordinated) spectrum – to built and launch all 16 LEO (LEO or MEO – they don’t tell what it is) satellites by 2010 for $650 million and have them operational (to build, launch and insure all for $40 million per satellite!) the numbers and schedule don’t make any sense. This is my very humble opinion.
Looking to the picture of the satellite it is obvious this satellite is using the Globalstar LEO bus manufactured by Alenia for Loral/Globalstar LEO satellite telephony system, which failed miserably in a marketplace because of service cost vs. GSM system. The biggest problem is going to be a business viability of such system which much more expensive than GEO satellite system, and operational/replacement costs will kill or will force them into the destitution. Unlike always profitable GEO satellite systems, LEO/MEO systems are fundamentally weak, look what’s happened to Iridium, Globalstar, Teledesic, ICO etc., etc, most of them are gone, few of them surviving only because of DOD contracts, but each of them made losses to investors in ten’s of billions of dollars.
Please don’t forget about us in the South Pacific. We are not many, so laying cable to us makes even less sense.
Om, heard you on The World today! You did a great job!
I’m not privy to the technical ways and wherefores, but in light of Craig Rubens’ remark, I think a moment of recognition of Arthur C. Clark in light of a project like this would be appropo.