LOWER COST BROADBAND SATELLITE
INTERNET BACKHAUL FOR AFRICA

By Henry O’Tani: www.wlan.org.uk
 

 
In fair judgement, the current end user charges made for internet connectivity outside of  Europe and N. America in
locations such as Africa and Latin America are “extortionate”.  That is to say, charges 10 - 100 times that of Europe for
the same service, exploit the lack of competitive choice and the desperate need of end-users for  service at any price  This
sad state of affairs is made possible by a virtual monopoly and absence of practical alternatives. 
 
It is of interest therefore to look into the technical feasibility of purchasing and using European Direct Broadcasting
Internet Satellites “out of area” such as to provide volume bandwidth to sites in Equatorial Africa and Latin America at European
or North American prices.
 
The first observation is that the Equatorial Region due South of Europe (Gold Coast of Africa) is only some 7.5
degrees “off axis” for a service footprint  beam pointed at U.K. latitudes.
 
This is demonstrated by the following scale drawing of the geostationary satellite’s position in relation to the planet Earth.
 
 
 
Fig.1
 

The beamwidth required for coverage of Europe is approx. 4 degrees…  At the distance to satellite, the equator is
about 7.5 degrees from the centre of the beam.. (axis) Antenna beamwidth is normally given as the width between
points of measured half-power.. or 3dB)
 
Fortunately for Africa, radio antenna technology does not permit the kind of sharply defined beam edges associated
with visible light, even at the 2.5cms wavelength of K Band signals...
 
On the equator, the geostationary orbit is DIRECTLY OVERHEAD….. A prime focus Parabolic dish antenna thus lies
more or less horizontal and can be carved or moulded into the ground…
 
The requirement is a radio reflective surface with a parabolic contour conforming to an accuracy of several
millimetres…  (At 11GHz an error of +/- 2 mm will be negligible)
 
An impressive example of a huge antenna built into the ground is the one at Arecibo shown below with a diameter of
305 metres …
 
 
Fig. 2.

 
No need to panic ….   Ours doesn’t have to be this big!
 
The actual loss of signal 7.5 degrees off the beam axis is unlikely to be more than 13 -19dB  (20-80th)  so for Gold
Coast ..Gambia / Cameroon a 4 to 8 metre diameter dish would probably be fine…  (a similar examination of the use
of a  low cost USA satellite in Brazil worked out as a 14 dB loss needing to be made up for. )

 
On site measurements of the actual satellite signals required will be necessary to determine how large the “ground
station dish”  must be…

 
Assuming a 1 metre diameter dish is completely adequate for service in the U.K. .. Basic Physics demands that moving
out of the main beam axis requires a larger dish to compensate for the progressive loss of signal. Doubling dish size
increases power by a factor of 4  (or approx. 6dB).
 
 

Fig. 3   Required dish size thus…

Here is a 16 metre diameter dish…

Fig. 4

 
 
 
If one is governed by the individualistic self-serving commercial ethic, dishes these size (16m) are far too large
for any individual…but are inconsequential for a large co-operatively organised (civilised & friendly) town or city community…
 
 
Communal Distribution
 
Once “backhaul” has been established it can be distributed at very low cost over tens and hundreds of kilometres using
the amateur and  community internet technologies developed by enthusiasts around the world as promoted by the
original community wireless internet site http://www.wlan.org.uk/operational_wlan_sites.html

 
 
 
DIY client equipment with 20 Km or more range can be made for under US $25…