Well, what do you know! It turns out there isn't enough fiber and 
electronics in the ground and under the sea to meet the world's soaring 
network communications needs after all.
     As a result, an industry that was almost given up for dead - the 
commercial satellite industry - is back with a vengeance. By 2002, the 
number of satellites serving various communications functions is expected 
to rocket from the current 150 to about 1,700.
     But whether the dozen or so new satellite projects will have anything 
meaningful to offer enterprise network managers is another question. 
Bandwidth, latency, coverage, and, especially cost remain as big or bigger 
issues as they do with terrestrial services.
     Circling the globe
     There are two principal ways to divide the major satellite services 
that are beginning to get up and running. The first is according to the 
height of their orbit.
     Most of the current satellites operate in what's known as 
Geosynchronous Earth Orbit (GEO). They track the rotation of the Earth in 
order to remain stationed above the same spot. Because they are launched as 
high as 22,300 miles above the Earth, they can 'see' a good portion of the 
planet. A single GEO satellite can receive and send signals to all of North 
America, for example. Three to five GEOs operating in tandem can cover the 
globe.
     The principal drawback of GEO satellites is la-tency. The height of 
their orbit generally results in a minimum propagation delay of half a 
second. Nevertheless, they are coming back into favor for services such as 
bulk file transfers.
     Three GEO satellites form the backbone for the new DirecPC Commercial 
service to be offered by Hughes Network Systems, Inc. This is an intranet 
access service that can offer raw downlink rates of 3M to 24M bit/sec, 
though with considerable latency (see IntraNet, May, page 12).
     By contrast, the current rage is in Low Earth Orbit (LEO) satellites. 
These smaller satellites sit only 400 to 1,200 miles above the Earth. As a 
result, they can receive and transmit signals within milli-seconds from 
antennas as small as 60 cm, or roughly 2 feet long.
     But because LEOs can only see a small part of the Earth's surface, LEO 
operators must launch a large number of them and develop intersatellite 
links analogous to the telephone network for handing off traffic.
     For example, Motorola, Inc. this year has launched 22 of the 66 
satellites it will require for its Iridium network, expected to be the 
first commercially available LEO system when it begins operation about a 
year from now. Another prominent LEO network, the Teledesic 'Internet in 
the sky' personally backed by cellular phone pioneer Craig McCaw and 
Microsoft Corp. CEO Bill Gates, will require a staggering 288 satellites, 
none of which have been launched yet.
     Because each launch costs millions of dollars with no assurance of a 
payback, critics complain that LEO projects could result in a trail of 
'space junk' girdling the plant. In addition, many such projects have been 
enmeshed in licensing controversies with the Federal Communications 
Commission and foreign authorities. To sort out the remaining LEO 
applications, the FCC is considering auctioning its remaining satellite 
licenses, further adding to the cost.
     Some operators are seeking a middle ground with Medium Earth Orbit 
(MEO) satellite networks at orbits of roughly 5,000 to 6,000 miles. The 
most prominent MEO system is Odyssey, largely backed by Teleglobe, Inc., 
the former monopoly carrier of international traffic originating in Canada. 
Odyssey plans to establish 12 MEO satellites to carry voice traffic and act 
as a switched 64K bit/sec dial-up data carrier at close to $1 per minute.
     Voice or data
     But unlike Odyssey, most of the emerging satellite services are 
concentrating their efforts in one of two directions. Some want to carry 
voice traffic around the world accessible with a single handset. They would 
compete with domestic cellular and digital personal communications services 
(PCS) systems still dogged by incomplete coverage and incompatible 
protocols. Others see Internet and intranet access as their ticket to 
riches.
     Unfortunately for corporate data network managers, the voice-oriented 
systems are likely to come first. The reasons are much the same as those 
found in the traditional wireless industry.
     Because their orbits are so low, LEO satellites are only expected to 
last five to nine years before gravity pulls them into the Earth's 
atmosphere, where they will burn up. That's even shorter than the 10-year 
term of the FCC's PCS licenses, the first of which were issued in 1995 with 
billion-dollar price tags.
     The response of the first PCS licensees, notably Sprint Corp., AT&T 
and the regional Bell operating company consortium PCS PrimeCo, has been to 
attack the mass market for phone calls and voice mail rather than roll out 
the data applications originally hyped by PCS evangelists, including 
outgoing FCC Chairman Reed Hundt. While the FCC has garnered nearly $20 
billion in wireless licenses, Aber-deen Group, Inc., a Boston-based market 
research firm, estimates that the emerging satellite services will invest 
more than $40 billion in their networks.
     The first two LEO systems, Iridium and Globalstar, hope to sign up the 
so-called 'mobile professional' crowd willing and able to pay premium 
prices for phone calls. Dial-up data will be available, according to 
Aberdeen, but only at the rate of 2.4K bit/sec from Iridium and 9.6K 
bit/sec from Globalstar.
     Teledesic's data focus will come from the expected use of wide radio 
channels at ultrahigh frequencies. Teledesic may draw competitors, but 
analysts say those potential competitors really are lying in wait to see 
how the McCaw/Gates superventure goes.
     Aberdeen Director of WAN Research Traver Kennedy says a key challenge 
for all of the new providers will be strikingly similar to those faced by 
terrestrial WAN and Internet service providers: coming up with specific 
quality-of-service parameters, particularly availability, latency and 
bandwidth, at an attractive price.