Low Earth Orbit Satellites
Geosynchronous satellites are not the only satellites orbiting the broadband access space. Low Earth Orbit (LEO) satellites have garnered a large amount of interest and press because of their high concept and the participation of heavy-hitters Bill Gates and Craig McCaw. Some conflict has arisen between LMDS and LEO supporters regarding spectrum allocation. Though it is unlikely that the bandwidth of these systems will exceed 1 to 2 Mb, their potential role in RBB is to provide a ubiquitous return path for other one-way technologies. Another possible effect is to siphon off investment and consumer interest from higher-speed services.
Background
Unlike their geosynchronous brethren, which operate at 35,800 km above the equator, LEO satellites orbit the planet at low altitudes. Depending on the system, the altitude of these orbits ranges from 780 to 1400 km, which is above the earth's atmosphere but below the Van Allen radiation belt. At these altitudes, a LEO satellite is in view of 2 to 4 percent of the earth's surface, which means that the footprint of any given LEO satellite is 4000 to 6000 km in diameter.
The low altitude provides two advantages for the consumer. First, latency is short. Phone calls through GEOs incur a 239 millisecond (ms) one-way delay; a round-trip delay is therefore nearly half a second. This causes annoying pauses during voice conversations. One-way latency for LEOs is about 6 ms. The second advantage is lower power consumption. LEOs travel so low that a handset requires very little power to reach it, as compared with a GEO.
Given these advantages and the recent advances in satellite launch technology, a number of companies have initiated development programs. Among these are Globalstar, Motorola, Iridium, Skybridge, and Teledesic. The objective of Globalstar and Iridium is global voice service. The objective of Skybridge and Teledesic is 2 Mbps data service.
Teledesic requested use of paired spectrum in the 28.6 to 29.1 GHz (uplink) and 18.8 to 19.3 GHz (downlink) band segments for its service links, and the use of 27.6 to 28.4 GHz (uplink) and 17.8 to 18.6 GHz (downlink) band segments for its "gigalink" gateway terminals. Teledesic proposes to operate intersatellite links in the 59.5 to 60.5 GHz and 62.5 to 63.5 GHz bands to interconnect each satellite with eight other satellites in the same and adjacent planes (www.fcc.gov/Bureaus/International/Orders/1997/da970527.txt) .
SkyBridge, considered by many to be Teledesic's largest competitor, is utilizing the more familiar Ku band (14 GHz up/12 GHz down) for its 80-satellite LEO constellation and expects to be the first broadband system operating when it launches service in 2001.
LEO Architecture
Because LEOs fly at low orbit, they move with respect to the Earth. This means that when as a user on the ground communicates with the LEO satellite, the LEO satellite passes over the horizon whereupon communication stops. To maintain the session, the original LEO must hand off the session to the following satellite. By succeeding handoffs, the session can be maintained. The frequency of handoffs is determined largely by the distance of the LEO from Earth, which determines the footprint size of the LEO. The lower the satellite, the faster it moves with respect to Earth, creating more frequent handoffs. There are possibly hundreds of satellites communicating with each other, handing off user sessions. The original Teledesic system had 840 satellites until a modified design reduced that number to 288. Higher orbits reduce the number of satellites, which is an important component of system cost.
Two different approaches are being taken regarding switching. Globalstar and Skybridge take the "bent pipe" approach, in which traffic is sent from the user up to the LEO. The LEO handles minimal processing of the bits and returns the traffic to the ground as soon as possible. All switching decisions are made on the ground.
Iridium and Teledesic switch traffic in the satellite. Each satellite contains an ATM or packet switch to forward packets to other satellites. When the final satellite is reached, that satellite beams down directly to the end user or the ground station. Switching in the sky minimizes the cost of ground stations and ground switching equipment. It also reduces the requirement to coordinate with local telephone companies for gateway services and settlement payments. However, switching among moving satellite is an enormously complex software problem.
Challenges to LEOs
Not surprisingly, the futuristic and spectacular concepts behind LEOs face some down-to-earth problems. Switching or routing in the sky is rocket science, but there are also more prosaic problems. One of the major initial problems is there is comparatively little military and commercial experience in LEO technology when compared with geosynchronous satellite technology. Therefore, there is a steep learning curve. However, the more prosaic problems are serious as well.
Telephone Interconnection
Voice service will require cooperation with land-based telephone companies. If an LEO customer initiated a phone call to a Pacific Bell user, at some point PacificBell must agree to accept the call. Similarly, LEO operators must sign settlements and interexchange agreements with telephone companies worldwide.
Spectrum Use
Conflict arose between Teledesic and LMDS operators about use of the 29.1 GHz spectrum (Ka band). Eventually, discussions with the FCC resolved the issue. However, LEOs are global systems, so spectrum-utilization problems could recur in other countries that use the same spectrum. Little use of Ka band occurs worldwide, but full global coverage requires spectrum agreements among the S-band, L-band, and Ka band. In addition, radio astronomers are concerned about LEO impacts on their frequency measurements.
Launch Capacity
Finding satellite launch space will be a challenge. More than 1700 satellite launches are planned for the next 10 years for uses other than LEOs. LEOs will require the launch of more than 400 satellites over the next five to seven years. Considering that there were 22 rocket launches worldwide in 1996, which placed 29 satellites in orbit, the launch business needs to add capacity quickly to meet launch demand. To make things worse, about 10 percent of launches fail. Insurance premiums in some cases are nearly 25 percent of payload value. Incidentally, only about 30 percent of satellite launches are handled by the United States. About 60 percent of the world's launches are made by the European space consortium, Ariane. Others in the launch business are China and Russia.
Cost Issues
In general, there is widespread skepticism about lifecycle costs for LEOs. The original estimate provided by Teledesic was $9 billion to launch an 840-satellite constellation. Independent industry observers made much higher estimates. Over the past year, Teledesic reduced the number of satellites to 288 by increasing altitude of the orbits. The following list details some of the cost concerns of particular relevance for LEOs:
Capital costs for satellite development, construction, and launch. For example, data communications equipment must be modified for operation in space to accommodate environmental factors such as temperature and radiation. This precludes the use of commercial, off-the-shelf data communications equipment. These modification costs are unknown at present.
Requirement for technical innovations, such as satellite-to-satellite communications at multimegabit rates and satellite-based switching.
Continuing requirement for ground stations and associated settlement costs.
Damage due to solar activity and small projectiles.
Development of new handsets and customer terminals.
The need to relaunch satellites. The lifetimes of LEOs are shorter than that of GEOs.
The Iridium Experience
Iridium is the first commercial LEO venture, and it provides global telephone service. Unfortunately, the handsets are bulky, the prices are high, and the costs of the satellites are daunting. As of this writing, technical and business problems have forced this highly publicized service into Chapter 11 bankruptcy. The fate of Iridium may have a negative impact on the broadband LEOs.
Nonetheless, the concept of LEOs exhibits a boldness not found except in science fiction. LEOs offer global roaming, a simple worldwide dialing plan, and an instant voice infrastructure for developing countries. In addition, Teledesic and Skybridge are offering megabit service. The impact on RBB is simply that every one-way technology has the possibility of a low-latency return path, which can be used everywhere in the world. More broadly, if successful, LEOs can substantially alter how people think about global communications.