Connecting the other three billion
Developing countries would benefit from lower-cost Internet connections. A US millionaire thinks he has a way to provide them, writes E&T.
Information and communications technologies have their benefits, but they have not reached all equally, with an obvious concentration of resources in developed countries. It is this that led, in the mid-1990s, to the introduction of the term 'digital divide', which according to the International Telecommunication Union (ITU) addresses 'troubling disparities in terms of access to information technology'. Originally coined with respect to computer access, the term has evolved to include access to the Internet and broadband technologies.
As with 'global warming' and 'the war on terror', labelling the problem has increased its visibility in the media and its importance to politicians. Despite this, there are still parts of the world with less than one telephone line per 100 inhabitants (compared with about 60 per 100 in the US and UK), and very limited access to the Internet.
A report by the World Bank, entitled 'Global Economic Prospects 2008: Technology Diffusion in the Developing World', says that "Internet services have not spread uniformly across the developing world. Although Internet penetration has risen by 41 per cent in sub-Saharan Africa from 1999 to 2005, the region still has the lowest penetration of any developing region in the world."
This is partly due to a lack of infrastructure, the report says, which makes these technologies prohibitively expensive. It compares, for example, call centres in Kenya that pay $7,000 per Mbit/s of bandwidth to be connected by satellite with those in India which pay $500 per Mbit/s for a fibre-optic cable connection.
Cable is usually cheaper than satellite, but there are parts of the world that will never be cabled, either because of their geography or because their populations do not live in easily wired, urban concentrations. It is this aspect, and the fact that there are still so many poorly served people, that forms the keystone of the business plan of a new venture called O3b Networks.
O3b Networks, which first came to public attention in September last year, is based in Jersey and backed by Google, Liberty Global and HSBC. Its slogan, "Connecting the other three billion", explains its name and summarises its mission: "to make the Internet accessible and affordable to everyone on the planet".
O3b Networks says it is building a new fibre-quality, satellite-based, global Internet backbone for telecommunications operators and Internet service providers in emerging markets, providing them with a "low-cost, high-speed alternative to connect their 3G, WiMax and fixed-line networks to the rest of the world".
It plans to do this by deploying at least 16 relatively small communications satellites into 8,000km-altitude medium Earth orbits around the equator. A contract for the 16 satellites, with options for more, was announced by O3b and its contractor, Thales Alenia Space, in September, but the manufacturer started work on the design, under contract, in 2007. The first eight satellites are due to be delivered and activated by late 2010.
According to O3b, the system will "combine global reach and the speed of a fibre-optic network", allowing "billions of consumers and businesses in more than 150 countries to benefit from high-speed Internet connectivity for educational, medical and commercial applications".
The satellites' global reach will be provided by the footprint of their antennas, which will cover up to 45° north and south of the equator where most of the world's 'other three billion' reside. The speed is provided by Ka-band satellite payloads capable of throughputs "with a combined total capacity in excess of 160Gbit/s", according to the company. The Ka-band spectrum has been chosen to provide high bandwidths, and because it is relatively new territory for satellite communications.
The O3b constellation will overlay the existing terrestrial network of cables, microwave links and cellular towers by joining the towers together wirelessly, avoiding the excavation of thousands of very (very) long trenches. Picture the satellites as multi-legged spiders, moving constantly across the landscape, connecting with individual cell towers for a while and then roaming on around the planet.
Each satellite in the network will have 10 spot beams, each delivering in excess of 1Gbit/s to the cell towers. O3b expects telecoms companies to update these towers, of which it estimates there are 2.2 million in the network's coverage area, to receive its signals. By way of encouragement, it launched Quick Start Africa, its first turnkey offering, in November. By providing an integrated package of ground equipment, space segment and teleport access at special pre-launch pricing, Quick Start is "the simplest way for telcos and ISPs to connect their networks to the global Internet," said John Finney, executive vice president of sales and marketing for O3b Networks.
Just eight days after the Quick Start Africa launch, O3b revealed its first two multi-year contracts, one with Netcom Africa, Nigeria's leading telecommunications services supplier, the other with Galaxy Wireless Communications Ltd, also based in Nigeria.
Geostationary orbit or low latency?
One of the key things that sets the O3b system apart from standard satellite communications is its orbital infrastructure. Individual satellites in a 36,000km-high geostationary orbit can provide communications links to about a third of the globe, so placing three spacecraft equidistantly around the equator gives whole-Earth coverage.
The fact that the satellites are stationary with respect to the Earth's rotation means that customers' receiving antennas can remain fixed. The problem with placing satellites in any orbit of lower altitude is that this advantage evaporates; the solution is to launch a multi-satellite constellation to ensure that at least one is always in view of the ground antenna.
The usual technical argument in favour of low-orbiting satellites over those in geostationary orbit is one of latency, or signal delay. Because the satellites are closer to the Earth, the delay is less. However, according to communications satellite consultant and author Joe Pelton, "while the lower latency of transmission is potentially of some value, there are problems to consider". When the low elevation angle of the satellite, at the point of hand-off to the next satellite, is taken into account, "the transmission distances are not all that much shorter", he says. Moreover, he adds, the high signal attenuation of Ka-band frequencies by rain, the constant repointing of multiple beams, and managing hand-offs at transmission speeds of up to 1.25Gbit/s "are pretty severe technical challenges".
Critics point out that we've been here before, and with less than stellar results. The 1990s saw a number of satellite constellation proposals, all promising to revolutionise satcoms in one way or another. Iridium and Globalstar launched dozens of satellites designed for mobile applications, but were thwarted by the 'boom and bust' cycle of the decade, the telecoms downturn and the fact that terrestrial mobile phones evolved much more quickly than anyone expected. Both companies went bankrupt and, in the fire-sale economics of the times, were acquired for a pittance and redeveloped. A similar system, ICO Global, managed to put a single satellite into orbit, but ran out of money and has been mired in a legal dispute with Boeing, its contractor, ever since, while the rest of its satellites languish unfinished in storage. At least a dozen other constellation systems were proposed around the same time, but only historians remember their names.
Interestingly, the O3b system uses some of the bandwidth freed up by Teledesic and incorporates the same interference-avoidance techniques designed for SkyBridge, but it can't escape the constraints of the financial markets, especially in today's uncertain environment. Industry commentators have suggested that backers will need to raise around $600m to build and launch the system, so even considering the initial $60m investment and the company's expectation of raising "four or five times that" by the end of 2008, it still has a way to go.
Even when it gets there, there will be competitors such as ViaSat, which plans to launch a single geostationary satellite in 2011, offering a throughput of 100Gbit/s at Ka-band - more than half what O3b promises with 16 satellites.
Given the history of failed commercial constellation systems, in contrast with the success of the government-backed GPS constellation, what makes O3b think it can do better?
First, a little background. O3b Networks was founded by American millionaire entrepreneur Greg Wyler, who sold his semiconductor cooling-technology company to dedicate himself to providing affordable Internet access to the underserved billions in developing nations. Wyler could have retired comfortably on the $15m he made from the sale of his first company, but a family tragedy led him to reassess his life.
By chance, Wyler met a Rwandan government official who complained to him about the prohibitive cost of Internet access in his country. As a result, in 2006, Wyler's communications company Terracom introduced a state-of-the-art terrestrial wireless network in Rwanda, one of Africa's poorest countries. "We simply have to bring the Internet into each of these huts," he said at the time, "and the rest will fall into place." It seems, however, that a terrestrial solution was not enough, and O3b was born.
Asked recently about his motivation for getting involved in a satellite system, Wyler said that developing a satellite network is "the only cost-effective way to quickly and efficiently roll out a low-latency connection on a global basis to connect carriers to the global infrastructure."
The company's partners are understandably supportive. Google product manager Larry Alder said: "We believe in O3b Networks' model and its goal of expanding the reach of the Internet to users who currently have limited and expensive connection options, as it complements our mission of organising the world's information and making it universally accessible and useful." Liberty Global's CEO Michael Fries said: "Using innovative modern satellite technology, O3b will make fibre-quality connectivity available throughout most of the world, without having to lay any fibre."
Before getting carried away with the potential of this new communications infrastructure, it is worth reminding ourselves how, back in 1998, Teledesic described itself as "the first satellite communications network to enable affordable, worldwide access to fibre-like telecommunications services such as broadband Internet access, videoconferencing and interactive multimedia."
CEO Steve Hooper was typically upbeat regarding the system's promise.
"Teledesic's ability to provide guaranteed end-to-end quality of service anywhere in the world will continue to set us apart," he said. "Our ability to execute our plans in 1999 will bring us closer to realising the world's first Internet-in-the-sky."
Satellite consultant Pelton offers his own assessment of O3b. While believing that Wyler "truly wants to do the right thing to help 'the other three billion people'," he warns of "the challenge of finding a viable and cost-effective solution," in terms of both satellite design concept and business model.
"O3b has some things going for it that Teledesic, Iridium and Globalstar did not," he says. For example, establishing the network with eight satellites and increasing to 16 only after its viability has been verified is "a much better concept than [its predecessors'] that required building and launching 50 to 900 satellites before revenues could start to come in the door.
"Also, concentrating on Internet services rather than telephone services makes a good deal of sense for the equatorial market that is leapfrogging traditional telephone concepts and going directly to Internet and voice-over-IP services." Having Google and Liberty Global as business partners cannot hurt, either.
Analysis of today's emerging space industry shows that computer and Internet entrepreneurs have a new-found affinity for commercial space developments, an almost 'born again' belief that they can improve on the old ways of doing business.
Notable examples include: PayPal developer Elon Musk's efforts to build rockets and launch supply capsules to the International Space Station; Microsoft co-founder Paul Allen's backing of SpaceShipOne, the prototype for Virgin Galactic's suborbital tourism vehicle; and 3D graphics pioneer and 'Doom' game developer John Carmack's ventures with reusable rocket technology and lunar landing vehicles.
However, the involvement of serial entrepreneurs does not guarantee success, as shown by the misguided adventures of Bill Gates and Craig McCaw with Teledesic. It remains to be seen whether Wyler and his colleagues and partners can raise sufficient financing and reinvent the satellite constellation concept.
But hope remains that O3B's 'spiders' will help drag the other three billion into the World Wide Web. As Pelton puts it, "the scale of the vision and the clearly humanitarian scope of what Wyler and company are trying to do makes you wish very hard that they can succeed".