It's a bird, it's a plane -- no, it's a LEO Analysis/Commentary Since Arthur Clarke first prophesied them in the fifties and Howard Hughes first invested in them, satellites have been the wonders of man-made nature. Sputnik, the soviet satellite, spurred the space race in the late fifties, and now Microsoft's Teledesic wants to launch satellite constellations to create an Internet in space. Satellites circle the earth within the earth's escape velocity. That means, if they orbit too fast, they lose centrifugal force and swing away like pebbles catapulted by a slingshot. If they orbit too slowly or fail to gain enough altitude, they lose centripetal force and fall back to earth. It's a crafty balance between push and pull. So far, there has been no other way of sending satellites into space except by using boosters filled to the gills with explosive jet fuel. Once airborne, satellites orbit, for example, the equator or the North and South Poles, and may even eventually circle the whole surface of the earth in fleets, evenly spaced, forming a huge revolving dome. Teledesic suggests that the future depends on turning away from large geostationary satellites. These move around the earth in the same direction and speed as the earth's rotation, and hang like kites above a given coverage area. Teledesic feels the future lies with LEOs: low-earth orbiters that will communicate reciprocally and carry signals back to earth faster than the high-flying geos. Both kinds of satellite owners worry about the "latency problem," the time it takes for a wireless signal to travel up to a satellite and back. This pause in transmissions threatens to upset the activity of Internet packets, which must often make several trips back and forth between their point of departure and reception before completing a communication. The advantage of a Net in the sky is that anyone with a satellite dish will be able to enjoy access, whether from the jungles of Southeast Asia or the wilds of the Klondike. A wireless Internet, claims Teledesic, will match or surpass the advances of fiber optics. Teledesic wants to ensure that land-based protocols keep up with wireless ones, and of course, vice-versa. But satellites seem, at least for the moment, to be a money trap, much like the first intercontinental telephone line. Nineteenth century investors sank huge sums, only to be foiled by technical problems. Recently, the chairman of a Teledesic competitor, Iridium, resigned after investing $500 million against sales of $1.5 million. Iridium was involved in providing satellite service for wireless telephones. As if to fill the void left by Iridium, two new ventures propose to offer satellite radio broadcast service, using as leverage contracts with major automobile manufacturers to place special satellite-ready antennas in new automobiles. One problem: the satellites have yet to be launched. July 28, 1999