The company said it could bring phone and other services to rural, suburban and sparsely populated areas as well as to urban areas lacking sufficient broadband infrastructure, and do it faster and more cost-efficiently than by either laying copper cables or building cellular stations.

Mr Brun said SkyBridge expected Asia to be its second-largest market after North America, with at least 10 per cent of Asian revenues coming from the mainland.

Most Chinese users would be corporate, at least for the first decade, but home users would join because of the low prices, he said.

Sending and receiving data should cost SkyBridge partners - such as local telecoms or other service providers - about 30 US cents per megabyte.

Home users will be able to receive data via fixed satellite dishes at up to 20 megabits per second, while businesses should be able to download data at up to 100 mbps using larger dishes.

Upstream rates will be slower, at a maximum of 2 mbps for consumers and 10 mbps for business users.

Providing a three-minute phone call anywhere in the world would cost the carrier just 40 cents, meaning that even if the service provider marked up the price to end-users for a 100 per cent profit, it still would charge drastically less than the $3 per minute for an international voice phone call through Iridium, a recently launched satellite service.

Initial investment in an antenna and a set-top box should be about $700, based on manufacturing and distribution costs.

SkyBridge's strategy is to take advantage of all interfaces, including the PC and television.

Alcatel-backed SkyBridge wants to launch full commercial services in 2001, two years after another Alcatel-backed consortium - Globalstar - is supposed to launch its own LEO satellite system.

Globalstar differs in that it will provide only narrowband voice services to mobile telephone users.

If SkyBridge keeps to its plans, its services will be available one year ahead of Teledesic, Motorola and Bill Gates' similar project which is planned to launch after 2002.

Speed to market is crucial, because in addition to competing against other satellite systems, SkyBridge also is competing against ground technology such as asymmetrical digital subscriber line (ADSL), which enables up to 8 mbps transport of digital video, audio and data over existing copper telephone lines.

ADSL already is being used in video services and high-speed Internet access in Singapore and several mainland provinces.

Satellite is considered more appropriate where copper lines have not been laid, because it offers immediate access.

SkyBridge's target is to garner 20 million users worldwide - one-third home users and two-thirds corporate - out of the expected 400 million broadband users in 2005.

Mr Brun said these forecasts took into account the current economic crisis and were conservative.

'The crisis has an impact on capital expenditure, but access will remain indispensable and demand will remain,' he said, adding that despite facing financial difficulties, the Japanese investors in the SkyBridge consortium were still backing the project.

Analysts continue to express concern about satellite projects such as SkyBridge, saying they are risky due to the heavy investment needed.

One recent example is Iridium, a $5 billion global satellite system for mobile voice communication and short messaging. After much delay, it was launched on November 1, amidst concern about poor performance and higher-than-expected competition from cellular networks.

One year after the SkyBridge project kicked off in 1997, it decided to increase its global bandwidth capacity by almost 50 per cent, from 144 gigabytes per second to more than 200 gbps.

The number of satellites also grew, from 40 to 80, raising the cost of building the system to $4.2 billion.

SkyBridge partners will receive licences to offer its services in return for investing a combined $2 billion to build satellite dishes on the ground.

'We do not want to bypass local operators or compete with them,' Mr Brun said.

This approach is fundamentally different from that of the $9 billion-plus Teledesic project, whose 288 satellites can swap data, completely bypassing the terrestrial network.

Another key difference is that SkyBridge is much simpler. It has obtained from the International Telecommunications Union the right to broadcast on the Ku-band (10 to 18 gHz), a frequency band already heavily used by geo-stationary satellites and earth stations.

This means the technology is well developed and the components inexpensive.

But above all, SkyBridge designed its system so none of the important control technology is on the satellites. All the intelligence, such as the asynchronous transfer mode switches, is in the earth stations, enabling service providers to upgrade and design services more easily.