Satellite Links Page 1

Satellite Links
Satellite links are used to provide communications over very large distances (global coverage).
This is achieved by using the satellite as a repeater. A ground station relays a signal up to the
satellite at a frequency known as the uplink frequency. The satellite receives this signal and
re-broadcasts it on a downlink frequency to another ground station. If digital communications
signalling is used, the signal may be regenerated before it is re-transmitted to Earth.
The analysis and design of satellite links is not unlike other links we have already studied in the
course. There are particular details in evaluating certain components of the link budget that
deserve particular attention: the characteristics of the satellite transponder, the fact that two
frequencies (and hence two link budgets) are used, and the need to factor in the orbital geometry
of the satellite to compute the free space loss.

1 Transponder Characteristics
A satellite receives a radio signal on its uplink receiver. It then downconverts this signal to the
downlink frequency, and amplifies it before sending the downlink signal to the transmit antenna.
These functions are carried out by the satellite transponder. Ideally, aside from the frequency
shift, you can imagine this process as a 2-hop link (discussed earlier in the course) with some
amplification between the links. At the satellite, apart from noise considerations, you would
expect the downlink signal to be G dB higher than the received uplink signal, where G is the
transponder gain.
In practise, it is not quite that simple. The transponder amplification stage is realized from a
solid-state amplifier or travelling-wave tube (TWT) which can only supply a maximum power level
due to device constraints. Hence, the transponder only behaves as a linear device over a range
of input power levels. This is demonstrated graphically in Figure 1. The power of the uplink
signal is shown on the x-axis. In this example, the transponder has a nominal gain of 30 dB, so in
theory, the output signal power should be increased by 30 dB. We can see that over the majority
of the characteristic, this is true. However, at a certain point, the output becomes saturated as
the amplifier is no longer capable of delivering additional power (it is essentially clipping the input
signal). In this example, the saturated output power is 35 dBW (about 3.2 kW). Well below
this point, the curve departs from the ideal linear characteristic (shown as a dotted line) and the
amplifier is said to be in compression or saturation.
Operating beyond the compression point of the amplifier causes additional problems. The amplifier
is no longer linear. A nonlinear device is capable of generating new frequencies from the frequencies
present in the input signal; in this case, the input communication signal will be distorted to
produce intermodulation products which extend beyond the bandwidth of the input signal. This
can create interference problems: if you imagine a single channel in the input signal as occupying
a specific bandwidth, then the spillover will produce unwanted signals in the adjacent channels, or
interference. This is essentially noise, so in effect we are increasing the noise in the re-broadcast
signal when the amplifier is operated in compression.
Therefore, it is important to limit the input power to the transponder, which equates to managing
Prof. Sean Victor Hum Radio and Microwave Wireless Systems
Satellite Links Page 2

Figure 1: Satellite transponder input/output power characteristic

the EIRP of the ground station carefully. The goal is to obtain as much output power available
from the satellite while minimizing the signal distortion. This usually means reducing or backing
off the input power, so that the amplifier is operated just below compression. This point gives
decent output power while producing manageable levels of distortion/interference. The saturated
input power density may be specified for a satellite, as opposed to the saturated input power.
In this example, the output begins to compress at an input power of about 5 dBW. Increasing the
input power beyond this point does not yield a linear increase in output power: for example, at
an input power of 10 dBW, we would expect 40 dBW of output power and yet only are able to
achieve a power of 35 dBW. The output is fully saturated at this point.
A suitable input power level could be right before the saturated region is entered, along with some
margin. If the input power level was chosen to be 4 dBW, we would say the the input power
back-off (IPBO) is 6 dB, since the power has been reduced 6 dB from the saturated power point.
Correspondingly there is an output power back-off (OPBO) which in this case is 2.5 dB; that is,
the output is 2.5 dB from being fully saturated.
A common empirical formula that is used to model the input/output characteristic described

Prof. Sean Victor Hum Radio and Microwave Wireless Systems

Satellite Links Page 3

above is
Wout 4Pin /Psat
= , (1)
Wsat [1 + (Pin /Psat )]2
where Wsat is the saturated output power from the transponder, Wout is the actual power power
produced at the output of the transponder, Pin is the input power density to the transponder’s
receiving antenna, and Psat is the power density at the transponders’ receiving antenna that is
known to saturate the output of the transponder.

2 Link Budget
As discussed, the link budgets on the uplink and downlink are treated separately, since different
transmission distances may be involved, and there is a significant different in frequency/wavelength.
On a per-Hz basis, the link budget equations are:
 
C GS
= EIRPG − F SLU − LU + + 228.6 (units: dB-Hz) (2)
N0 U TS
 
C GG
= EIRPS − F SLD − LD + + 228.6 (units: dB-Hz) (3)
N0 D TG
where the subscripts U and D refer to the uplink and downlink, respectively, and the subscripts
S and G refer to the satellite and ground station respectively. Note that the EIRP of the ground
station is selected so that the transponder is suitably backed-off from saturation. The terms LU
and LD refer to other losses on the uplink and downlink, such as atmospheric absorption, rain
attenuation, etc.
Free space loss forms the bulk of the loss in the link budget. This is especially true for satellite
systems due to the huge distances involved which are typically tens of thousands of kilometres
from station to station. Hence, it is important to understand the orbit geometry of the satellite so
that the link distance and hence free space loss can be estimated accurately. The calculation of
the link distance requires an understanding of orbital mechanics, which is discussed in a separate
note.
Knowing the link distance the free-space loss can be calculated on the link being considered
(uplink/downlink). The link losses on the uplink and downlink are not the same, even if the
distance r is the same, because the uplink and downlink frequencies are not the same. In fact,
usually the downlink is at lower frequencies. This is done for multiple reasons:

• Atmospheric attenuation is less at lower frequencies.

• The ground station is not power-limited like the satellite, and hence is better equipped to
deal with attenuation on the uplink. The maximum EIRP of the ground station is ultimately
limited by the satellite’s transponder (see the next section)
• The G/T on the uplink is very poor, since the satellite “sees” a warm earth. Hence, there
is little point in using a lower frequency with less attenuation since the SNR is going to be
relatively poor, anyway.
• Improvements are thus best reserved for the downlink.
Prof. Sean Victor Hum Radio and Microwave Wireless Systems
Satellite Links Page 4

3 Satellite Orbits
As discussed in a separate note on orbital mechanics, the time T it takes the satellite to transit
through one orbit is determined knowing the gravitational force produced by the Earth, and the
distance the satellite is from the centre of that mass a. The orbital period is equal to
2πr3/2
T = √ (4)
µ
where µ = mE G is the product of the Earth’s mass me and the universal gravitation constant G.
This results a constant µ = 3.986 × 1014 N · m2 /kg which is known as Kepler’s constant. Knowing
the orbital altitude of the satellite H, we can determine a = Re + H and find T . The orbital
period for various different orbits and some well-known satellites are summarized in Table 1.

Table 1: Satellite orbits

Point of interest: GEO orbit Geostationary orbit is a special case of a satellite orbit meeting
the following conditions:

1. The orbit is perfectly circular, with an eccentricity of zero unlike LEO, MEO, and other
orbits.
2. The altitude of the satellite is such that the orbital period of the satellite is 23 hours, 56
minutes, and 4.1 seconds, which is one sidereal day which is the time it takes for the Earth
to rotate about its axis exactly once.
Prof. Sean Victor Hum Radio and Microwave Wireless Systems
Satellite Links Page 5

3. The satellite’s orbital plane must coincide with the the plane of the equator.

Under these conditions, the satellite appears perfectly stationary with respect to an stationary
station on the Earth’s surface. This makes this orbit highly coveted for telecommunications and
broadcast applications, such as television. The geostationary orbital altitude is, from Table 1, H =
35, 786 km. It follows knowing that Re = 6, 378.2 km that the ratio ReR+H e
= 0.151 = 1/6.61.
Another way of remembering this is that the geosynchronous orbit altitude is approximately 6.6
Earth radii.

4 Orbital Geometry
Satellites generally orbit the Earth along an elliptical path and this ellipse is actually inclined
relative to the equatorial plane of the Earth. Satellites also orbit the Earth at different speeds,
since the orbital periods depend on the orbital altitude of the satellite. Generally, to locate a
satellite at an instant in time requires knowledge of the geometry of the ellipse relative the the
Earth as well as times that the satellite passes reference points along the ellipse. The parameters
are known as the Keplerian elements of the satellite and are precisely known by satellite operators.
The relation of these elements to the elliptical geometry we have described here is beyond the
scope of this document, but it suffices to say that in general the distance from a ground station to
the satellite depends on time, and that the satellite may not always be visible (above the horizon).

satellite

nadir direction

zenith direction

C
sub-satellite point
Earth

Figure 2: The sub-satellite point

From the perspective of a ground station, what matters is the location of the sub-satellite point
on the Earth’s surface which is a point on the ground in the nadir (downward) pointing direction
of the satellite. This point is shown in Figure 2. This point can be described by its latitude and
longitude. The Earth station also can be located by its own latitude and longitude. Combined,
these parameters are:

• Le , the Earth station latitude (the number of degrees the station is north of the equator);
• `e , the Earth station longitude (the number of degrees the station is west of the prime
meridian);
Prof. Sean Victor Hum Radio and Microwave Wireless Systems
Satellite Links Page 6

• Ls , the latitude of the sub-satellite point; and

• `s , the longitude of the sub-satellite point.

Consider the link between a ground station and orbiting satellite, shown in Figure 3.

Figure 3: Orbit geometry

The angle ψ is related to the coordinates of the Earth station and the sub-satellite point according
to
cos ψ = cos Le cos Ls cos(`s − `e ) + sin Le sin Ls . (5)

The remaining angles shown in the diagram are easily related by applying the law of sines to the
triangle formed by the orbit geometry. Hence,
Re Re + H r
= ◦
= . (6)
sin β sin(α + 90 ) sin ψ
Since sin(α + 90◦ ) = cos α,  
−1Re + H
α = cos sin ψ . (7)
r
α is the elevation angle (angle above the horizon) that the Earth station antenna needs to be
pointed to to make contact with the satellite. Therefore, knowing the coordinates of the Earth
station and the sub-satellite point, the appropriate elevation angle can be found. An additional
angle, the azimuth angle, is also needed, but is not discussed here.
Knowing the angle ψ we can find the total distance r the satellite is from the ground station.
Using the law of cosines,
r2 = (Re + H)2 + Re2 − 2Re (Re + H) cos ψ, (8)
or " #1/2
 2
Re Re
r = (Re + H) 1 + −2 cos ψ . (9)
Re + H Re + H
Prof. Sean Victor Hum Radio and Microwave Wireless Systems
Satellite Links Page 7

The elevation angle can then be found knowing sin(α + 90◦ ) = cos α, so

(Re + H) sin ψ sin ψ

cos α = = 1/2 . (10)
r  2
1 + ReR+H
e
− 2 ReR+H
e
cos ψ

If instead, we wish to determine ψ for a given elevation angle α in advance, we can use the fact
that β = 90◦ − (α + ψ), sin β = cos(α + ψ) and
 
−1 Re
ψ = cos cos α − α. (11)
Re + H

The satellite is only visible from an earth station is the elevation angle α to be above some
minimum value, which is at least 0◦ . From Figure 3, this requires
Re
Re + H ≥ (12)
cos ψ
or
Re
ψ ≤ cos−1 . (13)
Re + H
For a nominal GEO orbit, ψ ≤ 81.3◦ for the satellite to be visible.

Prof. Sean Victor Hum Radio and Microwave Wireless Systems