Satellite
Communications Systems
This will
be the first of several articles on satellite communications systems. Before
I lead you into this exciting journey, I like to point out some limitations
and constraints I have put on these papers. First, the emphrasize and subject
matter of the articles will be on systems that serve mobile users. In this
regard, I will not cover the Very Small Aperture Terminal (VSAT) systems,
or those huge C-Band TDMA Earth station (e.g.., INTELSAT) systems. As a
consequence to this, direct broadcast services for special events such
as the World Cup will not be mentioned. However, I will find time for an
article on one VSAT network later. Being one of the earlier integrators
and testers for the K-Mart VSAT Data Network, these systems are quite "dear"
to me. Second, I am taking the user's point of view: that is, I am more
interested in looking at the services these systems will be providing.
The technical details are given only as an afterthought. Third, I am a
system engineer. Link budget, transponder intermodulation noise, and equivalent
isotropically radiated power (e.i.r.p.) are Greek to me as Greek can be.
Most systems
discussed are still in the planning stages. Available literature provides
very little detail on the architecture and other technical data of these
systems. More often, top-level rather than quantitative information are
given. Furthermore, due to closely held proprietary data, it is very difficult
to obtain technical information from the vendors. The detail information
obtained are very "soft": there are many occasions when specific
numbers vary from one documentation to another. For examples, the altitudes
of the satellite orbits are tens or hundreds of miles different from one
document to another, or the bit- error-rates for data transmissions are
several orders of magnitude different from one study to another. In general,
technical details of new designs or recent changes in the designs are seldom
published, leading to the variations in the technical specification in
subsequent documents as mentioned above.
The systems
are classified loosely into two groups: the geostationary Earth orbit (GEO)
and the mobile satellite systems (MSS). The MSS is further subdivided into
the low Earth orbit (LEO), and the medium Earth orbit (MEO). The main difference
between the GEO and the MSS is that the MSS satellites move relatively
to a point on the Earth, while GEO satellites stay stationary with respect
to the coverage area. Added to the difficulty in the definition, American
Mobile Satellite Corporation (AMSC) calls its Mobile Satellite (MSAT) system
the geostationary mobile satellite service (GEO MSS ?).
GEOs move
in an easterly direction in an orbit at an altitude of approximately 35,800
km directly above the equator, at the same speed as the Earth and appear
to remain over the same stationary point on the Earth's surface. Due to
the spacing requirement (GEO satellites are spaced 20 apart), only a small
finite number of satellites can be "parked" in this equatorial
orbit. A significant disadvantage of GEOs for real time applications is
the relatively long range of communications. A quarter of a second round-trip
propagation delay is expected, making voice communications a problem. GEOs
provide a broad region of coverage. However, coverage of the Earth's polar
regions is minimal or nonexistent. INMARSAT for example can cover to about
+800 latitudes. At such a high altitude, only three or four GEO satellites
are required to provide global coverage. GEOs have the advantages of a
relatively constant coverage under the satellite, and a relatively constant
pointing or elevation angle to the satellite from an Earth terminal. This
slow satellite movements lead to minimal Doppler effects (change of frequency
in the transmitted waves).
LEOs and MEOs
operate at a much lower altitudes, generally between a few hundred (LEOs)
to around 10,000 km (MEOs). The orbits are essentially circular or mildly
elliptical in shape. Because of the low satellite altitude, the coverage
area under a given LEO or MEO will be relatively small compared to a GEO.
LEOs and MEOs are moving relative to a point on the Earth, and will circle
the Earth in a few hours. All LEOs must operate in a store-and-forward
mode since it will require many LEOs to provide uninterrupted coverage
to a fixed region of the Earth. However, due to their shorter propagation
delay, LEOs are much more amiable to real-time communications. MEO may
be effective as a compromise between LEO and GEO. MEO system will require
fewer satellites than LEO, being at a much higher altitude. With delay
somewhere between LEO and GEO, MEO is also good for most cellular applications.
A MEO system
will need 12 satellites (Odyssey) to provide North America coverage, while
a LEO system will need 24 satellites (ORBCOMM - US coverage only), or 48
(GLOBALSTAR - global coverage), or 66 (IRIDIUM - global), or 840 (Teledesic
- global). Needless to say, network management (routing, switching, etc.)
is a challenge for these systems.
Crudely speaking,
the network architecture/topology of these systems is very much alike.
It consists of numerous hand- held or portable terminals (user terminals),
a small number of feeder or control/telemetry Earth stations (control center/gateway
stations), and the satellite constellation. The control/gateway stations
interface with the user Hosts via private or public networks.
The carrier
frequencies allocated to these systems are rather limited. They are licensed
in the US by the FCC (Federal Communications Commission). Competition will
be fierce. It is my belief that few of these systems will survive beyond
the concept phase. The mobile links (between the user terminals and the
satellites) will operate at the L-band (about 1600 MHz for uplink, and
about 1500 MHz for downlink), the feeder links (between the control center/gateway
Earth stations and the satellites) will operate at the S-band (about 2.5
GHz), the C-band (4-6 GHz), the Ku-band (12-14 GHz), or the Ka-band (20-40
GHz). These higher bands (Ku and Ka) are used for "cross links"
as well (for store and forward transmissions between satellites). The higher
bands offer greater bandwidths but the technology needed to implement them
is much more expensive.
Generally,
lower frequencies imply lower propagation "path loss" and atmospheric
attenuation. They also permit lower cost Earth terminal technology. But
the size (area of aperture) of the required satellite antenna, which is
proportional to the square of the wavelength, can be quite large. The size
of the antenna depends greatly on the allowable transmit power: the lower
the transmit power, the smaller the antenna. The transmit power in turn
dictates the maximum possible antenna gain, which affects the data rates.
The desire for mobility requires that the antenna on the user's hand-held
terminal to be quite small. Consequently, most MSS systems will provide
low data rates (1.2-4.8 kbps).
Another problem
facing these MSS providers is the multiple- access technology to be used.
While most systems will implement the code-division-multiple-access (CDMA)
technology, IRIDIUM opts for the time-division-multiple- access (TDMA).
For those who are not familiar with these terms, it is sufficient to say
that multiple-access allows many users to occupy the transmission medium
(the spectrum or frequency band) at the same time. TDMA accomplishes this
by giving each user a time slot in the transmitted frame. While the available
spectrum can be subdivided into several TDMA "bands", CDMA signal
is spread across the entire available bandwidth. CDMA techniques, used
for years in radio local area networks, are less susceptible to interference,
and are more robust. However, their implementation will be more costly.
I may add quickly that most MSS do not use a pure CDMA or TDMA scheme,
but rather a combination of CDMA and FDMA (frequency-division-multiple-
access) or TDMA and FDMA. Until there is some agreement in the multiple-access
technology used, spectrum sharing and licensing rules can not move forward.
Of the satellite
communications systems mentioned only International Maritime Satellite
Organization (INMARSAT) system is operational, with AMSC MSAT system running
close behind (perhaps as early as mid 1995). All MEO and LEO systems are
still in the planning stages with the anticipated operational date sometime
in the calendar year 1998 at the earliest. A big problem facing the LEO
systems is the large number of satellites required, hence higher investment.
Investor reluctance will remain a significant factor.
Despite the
many predicted problems, these MSS will become part of our lives: business
executives will find a need to conduct business using "cellular phone"
while on a cruise in the middle of the ocean.
Viet-Dung
Hoang, Ph.D.
[email protected]
For
discussion on this column, join [email protected]
Copyright
© 1994 - 1997 by VACETS and Viet-Dung Hoang
:
