What's the Catch?
The promise is space-based broadband nirvana. Reality No. 1 is that a lot of things can go right and wrong with these complex and expensive constellations before 2002. Businesses shouldn't stake their entire future on next-generation satellites, but they ought to carefully evaluate these systems as extensions to existing networks.
Reality No. 2 is that with individual constellation efforts costing as much as $14 billion, the race for mind share, with all of the ugliness, techno-sparring and hype, has already begun. Many satellite systems will live and die based on this verbiage and its ramifications for ongoing financing. While the number of competitors is considerable today, rapid and dramatic consolidation is expecte
d over the next few years.
Reality No. 3 is that now is the time for businesses to begin studying these new systems; learning the language and pros and cons of GEOs, LEOs and MEOs; assessing the cold, hard and often-distressing facts about satellite security (see "The Achilles Heel of Next-Generating Satellites," page 30); and asking the right technology questions about the effects of jitter on LEOs and latency on GEOs. It's also time to explore how hybrid systems plan to direct traffic to either LEOs or GEOs and what standards efforts are afoot to aid in such tasks.
Obviously, there is an enormous amount to learn: architecture, potential pricing and successfully handling the hurdles ahead. And in this highly competitive industry, extracting information can be extremely difficult. Perhaps the best place to start is with an explanation of what makes these satellite systems so different. The biggest factor is the evolution of technology that takes advantage of Ka bandwidth.
Traditionally, satellites h
ave relied on passive bent-pipe architectures that receive a transmission, then broadcast it across a huge GEO (geostationary earth orbit) cell. These footprints can take in large geographic areas. Next-generation systems will include new onboard processing systems capable of caching information, instead of simply rebroadcasting it back to earth.
This stored information is then switched to one of many small cells that overlay the satellite's footprint. Traffic is more precisely targeted to its destination and this "spotbeam" approach enables a frequency serving a single cell to be reused beyond that cell and those immediately abutting it. With spotbeam frequency reuse, as well as the new bandwidth now made available with Ka, symmetric links become economically feasible. The high frequencies in the Ka band also mean that less power and less expensive and smaller antennas can be used on earth.
Newer LEO (low-earth orbit) and MEO (medium-earth orbit) satellites--which, as their names suggest, orbit at lo
wer altitudes than GEOs--can provide smaller and more energy-efficient spotbeams than more traditional GEOs because of their proximity to earth. The leading LEO broadband constellations are Motorola-Matra Marconi's Celestri, McCaw and Gates' Teledesic, and the cross-investment and marketing effort of Loral Space & Communications' CyberStar with Alcatel Alsthom's SkyBridge. Both OrbLink from Orbital Sciences Corp. and Hughes' Spaceway (through a recent expansion) rely on MEOs or MEO-GEO hybrids. But spotbeam technology also is making its way to the high-earth orbits of GEOs. It may even make its way to the more traditional Ku band, though it's considered most effective for Ka.
Every constellation type, whether LEO, GEO or MEO, has trade-offs (see "You Say GEO, I Say LEO, Oh My, Oh MEO," page 86). LEOs require more satellites, all of which must be in orbit before service can be provided. LEOs, however, address the fundamental problem with GEOs--latency, or the delay caused in reaching high orbiting GEOs a
nd returning to earth. Because of their high altitude, GEOs typically require larger, bulkier antennas and tend to be more bandwidth constricted than LEOs. MEOs are a middle ground between LEOs and GEOs, with an orbit that in some instances must contend with greater radiation exposure from the Van Allen belts.
Newer LEO, GEO and MEO systems also plan to use proprietary spaceborne switching between satellites, relying on intersatellite links pioneered by the Department of Defense's Milstar (Military Strategic and Tactical Relay System) and used for space shuttle communications. These links are one of the toughest technical challenges in these already highly complex systems. Satellite-to-satellite communications will have to take into account issues like power differences between satellites, routing around congested portions of the sky, and beaming in on satellite targets that move within a range rather than along a precise path. This becomes even trickier with global LEO constellations, since their orbit a
lso requires constellations much larger in size--a minimum of about 48 birds versus about eight for GEOs. Further complicating the issue is the fact that ATM--or ATM-like protocols--is being used by most of the broadband providers. During a satellite-to-satellite handoff, ATM cells could get smeared between satellites. This might reorder cells--something ATM doesn't accommodate well.
The upside to intersatellite links is they promise an improved way for high-speed traffic to move beyond the boundaries of a single satellite footprint. Today, delays are inherent in systems in which traffic is first shipped to the sky only to return and travel along ground links until it can be shipped back up to another satellite and then down again to its destination.
Management also will be a challenge, given the complexity of these networks. On the other hand, some believe that broadband satellite systems will be simpler for users to control--primarily because some topologies will create a network with a single point o
f contact, the satellite provider. Excised is the finger pointing that goes with leasing lines from multiple providers in the United States or globally.
|
REPORTS
ANALYTICS
Follow 
