Really Rocket Science

Hey, that got my attention: the sub-head in a story on MyBroadband in South Africa, an interview with Greg Wyler.

O3B Networks has been quietly preparing itself over the last 12 months for the moment last week when it announced that it was going to be offering cheap, low latency satellite bandwidth that can cover any part of Africa by 2010. It has put in place early finance with Google, Liberty Global and HSBC.

Russell Southwood talked to the entrepreneur behind the project Greg Wyler.

Q: What’s the overall technical configuration of what you’re going to do?

It’s a number of satellites flying over the equatorial area (of the earth). Because they’re approximately 5 times closer to the earth than geo-satellites, the latency is reduced by approximately five times. It’s a constellation of satellites?

Q: Why do you think you’ll succeed where others have failed with this approach?

There are two major things that are different (to previous projects). These other projects were designed to reach the developed world. It required many satellites because that had to be inclined to reach these areas. It would take 840 satellites to cover the whole of the earth in this way. We only need a minimum of five and therefore the cost is much lower.

The second thing is that this is a fully designed system with a fixed contract for 2010. As a single system designed just for trunking for telcos and ISPs, we know the costs, the delivery date and the performance. It’s a well thought out and structured process.

Q: What sort of coverage will the constellation of satellites offer Africa?

It will be the whole of the continent eventually. In the first launch, there are only 8 eight satellites which will give us 30 spots. Each spot is 500 kilometres in diameter. Each terminal can receive and transmit 1.25 gbps. The spots can be placed virtually anywhere on the continent. A spot on fibre is one physical landing station that then has to be connected to a whole city. Our spots can cover a whole city immediately or for instance, cover the whole of Nigeria.

We’re not planning to cover every square meter but we have the ability to cover any square metre you might want to specify. The unique beauty of the system is that it has the speed, latency and cost of fibre but we can take it immediately to where the customer wants it. Landlocked countries can get access to cheap international connectivity (without tackling existing transit issues).

Q: What price is the bandwidth going to sell at?

It will be in the range of US$500 per meg or below. It will be competitive with fibre but it’s not our intention to compete with fibre. Carrier class telecoms going out over the new fibre routes will require redundancy and we can complement any fibre network. Bit for bit the system is the equal of fibre: it’s better in some areas and not in others. For example, we can reach directly to the cell tower where build is needed for things like WiMax and EVDO because the spot beam of 500 kilometres can reach anything from 3 to 100 cell towers.

Q: What do you think the impact will be on the existing satellite business in Africa?

Existing satellite operators are currently very capacity constrained. In discussions we them, we see them flipping over their transmission (customers) to us to free up the geo-satellites to do WANS (multiple remote stations) and video distribution which is what they do best: one way applications like video distribution.

Q: What’s the latency going to be like on the system?

The path is five times less than for geo-satellites. It 123 milliseconds between the African port and some port in Europe connected to the global Internet. It’s comparable to fibre and in some cases will be quicker. And it’s certainly not the 600 milliseconds of geo-satellites.

Q: What markets are you aiming at?

Telcos for transmission backhaul will be the core of our market. FI2 will be our backhaul product and when the telco orders it, we will drop in a landing station at the telcos location. If they want a gigabit, we give them a gigabit. There is no CAPEX involved for the telco. This allows them to focus on their own network.

The ISP products are similar except the decision to put down a dish depends on whether the traffic is less than an STM1 or not. If not, we give a VSAT terminal for direct connectivity. For more than an STM1 we will give a 3.5 metre dish and a carrier class landing station.

The highest level product is FI2 Cell which illuminate over the whole 500 kilometres of the spot with 250-300 mbps capacity to and from cell towers. These will connect with a US$2,000 terminal at the base of a cell tower or a WiMAX base station. We will share this pool of bandwidth dynamically. This will allow the telco or WiMAX operator to place their towers without regard to line of sight. It might save the operator as much as US$1,000-1,500 a month.

Q: What’s your deal with Google? What’s their stake?

They are equity partners but they haven’t released exact figures because there are three very large global companies working together as a team: Google, Liberty Global and HSBC.

Q: How’s the overall fundraising going?

The first eight satellites will cost slightly less than US$450 million. We have about 12% of that in equity so far and we’ll be raising another 20% or more. The debt will cover the rest.

Q: What was the market reaction when you announced?

Our web site got over 380,000 hits and there were 400 newspaper articles. The company name O3B stands for the Other 3 Billion, the people who are not yet connected. Our task is to support ISPs and telcos reach directly to these people.

Light discussion here, and Slashdot's gets into the details.