nettime: TELECOMMUNICATIONS IN AFRICA - Michiel Hegener 2/2

[Geert Lovink <geert@xs4all.nl> (by way of Pit Schultz <pit@contrib.de>)]

Atlantic Ocean. The rest, including most telecommunications between
and within African countries, goes by satellite.
A loud roar turns into an even louder roar as Leonard Alloh, chef de
centre adjoint of the Akokro satellite ground station near Abidjan,
opens the door to the transmitter room. The voice of every caller to
Ivory Coast from outside West Africa or France travels in digital form
through these huge boxes full of electronic gadgetry. To bridge 40,000
kilometres with a radio signal, you need strong transmitters that need
constant cooling with powerful ventilators. Hence the noise.
The building where we stand seems insignificant alongside the two dish
antennas beside it, which relay signals from the atmosphere to cable
and vice versa. The first one, which was ceremonially switched on by
President Houphouet-Boigny in 1972, has a diameter of thirty metres and
is trained on Intelsat 605, which was launched on the Clarke orbit by
an Ariane 44L on 14 August 1991. Next to it stands an even larger dish,
built in 1989, communicating with Intelsat 603 at 34.5 degrees west
above the Equator. That is 400 kilometres north of the easternmost
point in Brazil and 35,786 kilometres above it. The 605 is ten degrees
further east. These are irrelevant details if you just want to phone
Amsterdam to Abidjan - until you see your phone bill. Of course, Akokro
ground station can communicate only with ground stations that also have
a dish trained on Intelsat 603 or 605. In order to communicate with a
country with no dish trained on either satellite, assistance from a
third groundstation, usually in the West, is needed: the signal will
travel through space twice: first from Akroko via Intelsat 603 or 605
to an Intelsat station in Europe, and then to its destination via a
second satellite. This type of "double-hop" phone call is quite
expensive, and the double time delay causes rather serious
inconvenience. Phone calls from the Netherlands never travel via more
than one satellite, but between developing countries they often do. One
solution might be for a country with no dish trained on the 603 or 605,
to build an additional one.
Fortunately, there is a much more intelligent solution: Pan-African
cooperation on satellite communications. The idea originated 15 years
ago, since when three paper organisations have come and gone. Till now,
the fourth such organisation, RASCOM, also seems to have spent much
more time on words than action. In 1992, the ITU published a
feasibility study of 5,000 densely printed A4 pages costing $8 million,
copies of which are available at $10,000 apiece. But here is a brief
summary free of charge: "Yes, RASCOM is feasible".
RASCOM's ambition is to carry all the satellite traffic within Africa:
from Ghana to Ethiopia, from Zambia to Egypt, and for instance also
between villages in Senegal, where a VSAT network is already in opera-
tion in rural areas. Essentially, RASCOM is a cooperative organisation
for the large-scale purchase of satellite capacity. All the African
countries are currently buying themselves a piece of satellite
capacity, albeit mostly from the same supplier, Intelsat. The only
notable exceptions are one or two North African countries that also use
satellites belonging to Arabsat (and occasionally Eutelsat) and some
countries like South Africa, which recently started buying capacity on
satellites belonging to the American company PanamSat (the worlds first
private satellite company with global coverage).
Stage One is for RASCOM to purchase a large amount of satellite
capacity from Intelsat and resell it to the member states, currently
numbering 42. Stage Two will then be for RASCOM to try to make use of
just one satellite, as opposed to the six used now. This would enable
people throughout Africa to see every television station in the
continent with just one dish on their roof. It would also put an end to
telephone calls within Africa being routed via Europe or America, the
douple hop connections - which now cost Africa $900 million per year.
RASCOM has been showing open interest in Intelsat 804, due for launch
in March 1997 at 22 degrees west above the middle of the Atlantic
Ocean. The drawback of this location is that it will appear very low on
the horizon in Mozambique, Somalia, and especially Mauritius - and the
lower the satellite, the further its signals have to travel through the
atmosphere. For these countries to communicate successfully with 804,
they would need larger, more expensive equipment.
RASCOM would therefore prefer a satellite located directly above Africa
- say at 33 degrees east, above Lake Victoria. On the other hand, the
advantage of 804 is that one of its antennas will be trained on the
east coast of the United States, allowing RASCOM to provide an
important intercontinental satellite link. America would fall outside
the footprint of a satellite at 33 degrees east, although all of Europe
would fall within it.
At present, the outlook is good. Speaking in Geneva, RASCOM
director-general Gounde Adadja recently announced that the organisation
hopes to have the $180 million necessary as startup capital to buy 50%
of Intelsat 804's capacity. Since RASCOM wants to be primarily an
African organisation, it demands that two-thirds of its funds be
contributed by member states. According to the latest information,
RASCOM hopes to submit the $180 million to Inselsat before 1 April
1996.
"Hello, how are you today?" says RASCOM's head of engineering and
operations, Kenyan James Rege, to his secretary as he enters her room.
"This is how I always start when I'm going to ask her a favour," he
quips unselfconsciously. He asks his secretary to dive into the
archives to retrieve two introductory parts of the feasibility study.
When she has found them, he gives them - all 624 pages - to me free of
charge, because a visit from a Western journalist to the RASCOM
headquarters in Abidjan is such a rarity. "We have to take advantage of
this opportunity to spread the word that RASCOM exists in a big way,
and that it will continue to exist for the benefit of the people of
Africa."
RASCOM's member states are free to buy satellite capacity from
suppliers other than RASCOM. But Christian Kow Sagou, RASCOM's head of
Marketing and International Relations, is quick to point out that this
is not desirable. "It's important to remember that RASCOM is not an
organisation that exists for its own benefit. We belong to African
governments. If we get a satellite, it will be their satellite - and
they will have to try to make a success of it." If the plan works, it
may also herald an African breakthrough for the Internet. If every
country in Africa has a dish trained on the same satellite, the
Internet will be able to provide multimedia communication for the
entire continent.
However, the ground stations will have to be linked in turn to good
terrestrial networks, using either cable or microwave transmission.
This is much more of a problem than satellite communication alone - ask
anyone who has ever placed a phone call to or from Africa. The national
Intelsat station is almost always near the capital city and connections
with it are always good, so it is much easier to phone Nairobi or
Bamako from the Netherlands than anywhere in the Kenyan or Malinese
countryside. The only exceptions are provincial towns with smaller
Intelsat stations (dishes with a diameter of five to ten metres,
depending on the number of lines needed etc.). Twenty African countries
are now using Intelsat for domestic communications, with anything from
one to dozens of extra ground stations.
Contrary to the belief of many telecoms users, there is an important
difference between satellite and cable communication. Cable networks
and microwave transmitters follow and emphasise existing geographical
structures. Cables connect areas of economic and demographic
concentration - roughly speaking, cities - which reinforces the
relative backwardness of the rest of the country. Satellite
communication, on the other hand, is geographically egalitarian; it
makes no difference whether your dish is located in the city or the
desert. Rural depopulation and overcrowded cities are a feature of
almost all developing countries, certainly in Africa. The absence of
telecommunications in rural areas has contributed to this trend. It
would be practically impossible to defeat rural isolation by laying
more cables - they would be too expensive and vulnerable to the
elements, and it would take too long. But satellite communication can
do the job fast and well.
Thanks to European colonialists, all the large cities of the African
coastal states are situated close to the ocean. To link them up, AT&T,
a world leader in the construction of underwater optical fibre cable,
has come up with a project to lay a cable all around the continent's
coast. If everything goes according to plan, construction will begin in
1997, and Africa ONE will be fully functional by 1999. Then all those
big coastal cities will be connected with each other and the rest of
the world. In many ways, this is a good - even courageous - plan. And
on balance, all of Africa will benefit from it. The only question is:
will Africa ONE do anything for the relative backwardness of the
interior, which is 95% of Africa?
This diversion brings us back to James Rege and Christian Kow Sagou.
"RASCOM's mission is after all to help Africa obtain better
telecommunications," says Kow Sagou. "The main goal is to get our own
satellite; but in the end, technology is a secondary concern. We aim to
become a very, very, very strong organisation." And this is why RASCOM
signed a memorandum of understanding in May 1995 with Africa ONE. This
memorandum also received the blessing of the Pan African
Telecommunication Union (PATU, founded in 1977), which manages
PanAfTel, an incomplete but reasonably good system of cable and
microwave transmitters interconnecting African countries. "One of our
aims is to turn Africa ONE into an African project - not just an AT&T
project," says Rege. And Kow Sagou adds: "RASCOM believes that African
states should provide at least half the total investment of 2,7 billion
dollars. If we can't manage that, then at least the management should
be in African hands. A lot of people ask: 'Why are you cooperating with
an optical fibre project - which is a competitor of the RASCOM
satellite system?' But that's not the way we see it. We see Africa as a
whole entity, and underwater cables are just as necessary as satellite
communication." Rege continues: "We thought about it long and hard
before deciding to cooperate. Suppose we did go into battle with AT&T.
An organisation that big could always turn around and tell you 'We're
going to do it anyway.' And they would! They'd just lay that cable and
connect every country that was interested." Kow Sagou disagrees: "Time
was when people came to Africa to do what they wanted and then took off
with the money. But those days are over. Practically every country in
Africa is a member of RASCOM. If AT&T approached any one of them with a
contract to sign, its first question would be 'OK, but does RASCOM
agree?'"
AT&T is itself keen on African cooperation. "It is very important that
non-coastal countries take part too. And this is why we're working
closely with RASCOM and the PATU," says Blaise Judja-Sato on the phone
from America. As AT&T Submarine Systems' regional sales manager for
West and Central Africa, Judja-Sato developed the sales strategy for
Africa ONE. His desk diary is full of meetings with representatives of
African PTTs, "but we haven't asked them to commit themselves yet."
>From 28 to 30 November 1995, he was in Botswana for negotiations with
representatives from land-locked countries, "to see how they can best
be connected via a few selected coastal states".
Politically, this is a very problematic area, because the necessary
transit agreements require a great deal of trust. Lack of trust was one
of the main reasons why the PATU microwave transmitter network was not
a resounding success. It also partly explains why satellite
communications have flourished so well in Africa: none of your
neighbours can cut off your link with the outside world when relations
are bad. On the other hand, Africa ONE's enormous capacity - 40
gigabits per second in both directions - is bound to look attractive to
investors. Judja-Sato still has no idea how Africa's 15 land-locked
countries will be connected with Africa ONE, nor whether AT&T cable
constructors will do any digging on land, although the results of the
negotiations in Botswana were encouraging. AT&T itself does not want
any satellite groundstations to close. "We are not out to connect all
of Africa to the cable and disconnect them from the existing
infrastructure," says Judja-Sato. "If only for diversity's sake, it's
good to have dual technologies at work." In the back of Judja-Sato's
mind, no doubt, lingers the outside chance of cables snapping or being
cut, even though Africa ONE has the advantage of circularity; that is -
if the cable breaks at one point, traffic will still be able to travel
in the other direction. But a breach in one of the branch cables could
cut off entire countries inland. Geostationary satellites are beyond
the reach of terrorists and enemy armies; if one breaks down, you just
train your dish on another one.
Africa ONE is incidentally not the only cable project on offer to the
continent. Siemens and Alcatel are both working on projects to link up
West African countries in an underwater Infobahn: Afrilink and Atlantis
2 respectively. There are plans however to integrate Atlantis 2 with
Africa ONE. Afrilink, the $500 million Siemens cable, is intended to
link all countries between Senegal and South Africa with each other,
Europe and the USA. However, there is little chance that any part of
the continent will be served by more than one cable, let alone three.
As Samson Brou Yapo explains: "Ivory Coast will have to make a choice.
Technically, the AT&T plan is the best, but it may be a lot more
expensive than other cable projects. And the cost depends heavily on
what other West African countries do. We don't have enough information
yet."
While technical and financial pros and cons undoubtedly cause some
division and confusion in the African telecoms sector, the overall
mood is one of idealism and hope. "You know," says Judja-Sato after
replying to my last question, "the problem is so enormous. In Cameroon,
where I was born, it's a nightmare to try to phone a neighbouring
country. Time and time again, you don't get through, and what's more,
it's terribly expensive. In Africa, travelling can be faster and
cheaper than phoning. The postal services don't work. The universities
can't get through to databases. All of this has got to change. And
Africa ONE can help."
Similar thoughts are what drive the Malinese Hamadoun Toure, Intelsat's
group director for Africa and the Middle East. "Africa ONE would be a
very valuable project if it were conceived and funded by Africans," he
comments from the first floor of Intelsat's stand at Telecom 95 in
Geneva. "But rural telecommunications are the biggest problem, and you
can't solve it with cables. Toure has had no contact with the Africa
ONE team, though he has met people from RASCOM, if only because
Director General Adadja is in town. Speaking to CommunicationsWeek
International, Adadja commented: "Walking around this exhibition, I ask
myself: 'Am I on the same planet? Is this global village not creating
outcasts?'"
Toure is keen to point out what Intelsat is doing to prevent such a
situation. "We have submitted a claim to the ITU for the 33 degrees
east position so that we can put a satellite there for exclusive use by
RASCOM," he says after a presentation on the interim Intelsat 804 plan.
"At the moment, we're conducting a feasibility study. In December 1995,
we'll decide. And in 1998, a satellite could be operational at 33
degrees east." (In December 1995 the Intelsat Board of Governors
decided indeed to go ahead with the 805 at 33 degrees east.) Earlier
than that, in March 1996, Intelsat will introduce demand-assigned
multiple access (DAMA), first via the 605 and later via other
satellites. DAMA is now the big buzzword in satellite communications.
Brochures that don't mention it are becoming collector's items. DAMA is
technologically very complex, but the idea is very simple: a certain
amount of bandwidth, or satellite capacity, is shared by a number of
ground stations in varying proportions. Whichever station needs the
most capacity at a particular moment will get the most; if a station
needs nothing, it will get nothing; and when it needs just one line,
that is what it will get. Setting up a phone link with DAMA takes about
threee seconds. At present, Intelsat customers still hire fixed amounts
of capacity for long periods, and the rate is the same however many
calls they make (it is possible to hire satellite capacity for just
half an hour - but the rate is higher, and you have to specify in
advance which half hour you want). The result is that money is wasted
when calls are few and far between, and at busy times the hired
capacity is overloaded - especially in poorer countries, which hire the
minimum.
With DAMA, therefore, they will receive a fairer bill. However, ground
stations will first have to be equipped with DAMA equipment, which
costs about a quarter of a million US dollars per station. "We're
introducing DAMA at the large ground stations first," says Toure.
"It'll be very cheap - 10 cents a minute for a 64 kilobit channel. At a
later stage, Intelsat's DAMA service will also be available for VSATs.