Let the Children be Fed First

This paper will show how Uganda Connect[1], together with a host of partners, friends and associates, overcame numerous obstacles in preparation to connecting the network of a World Links for Development school to the Internet by spread spectrum microwave, and how I and my colleagues attacked the problem of giving web access to students in Uganda, through:

Sensitizing stakeholders; gathering equipment for the network (and clearing it through customs); training a team to install and maintain the microwave link; setting up the link, working out topographies; and, having shown proof of concept, gaining enough support from stakeholders to allow the schools network to be established.

The paper will also describe the three-year process by which each hurdle was overcome, the importance of attending to details, tenacity in the face of difficulties, which leads to success and lessons learned which should encourage others doing similar projects. If we at Uganda Connect, with our low level of training and limited resources, could set up a wireless network for providing Internet connectivity you can too.

He was hardly noticed by the students gathered in groups, each around one of the dozen PCs, the head teacher of a secondary school who had brought the local Member of Parliament with me to visit the computer lab, so intense were they in their quest for new discoveries. Yet the knowledge they sought could only trickle through the Sun server's one telephone line link to the global Internet, if only for the few minutes allowed. Telnet was a hot topic for discussion! It was like going back in time, yet in March of 1999 Telnet was just about the greatest thing those Ugandan students had ever experienced about the Internet. And to see how excited they were about being connected to the rest of the world through their one telephone line made me a little hot under the collar that more was not being done to improve their connectivity!

Yet it had already been two years since the Uganda Connectivity Project team first visited the school, one of ten Ugandan schools participating in the World Links for Development programme. We'd come to the school in 1997 at the invitation of the head of the computer science department programme to discuss with teachers how Internet technologies could be used to enhance learning through information and communications technologies (ICTs).

The World Bank-sponsored programme had provided the school's computer lab with some high spec hardware, including several Pentium PCs. The dozen or so PCs, including a few 486 PCs and some 386 PCs, made an impressive sight.[2] The World Bank had also provided each school with a modem[3], and paid the monthly $60 local Internet service provider fee - almost a teacher’s monthly salary. They were nearly ready to enjoy the benefits of the Internet revolution that had begun to sweep the world - but not quite.

I asked the teachers how much experience did the students have with sending and receiving e-mail, browsing the web? The head teacher sheepishly explained that the school had only one telephone line that was used for administration and that, because their department had not been able to justify the cost[4] of using a telephone line for connecting PCs to the Internet they had not even used their e-mail account.

They hadn't connected to the Internet, so had only the vaguest notion of what the Internet was about, and therefore looked at me askance while I tried to project to them how much they were missing. As I wasn't getting much of a response I decided to liven up my presentation by giving them a rousing exhortation on the wonders of Internet Mail. By the time I was finished I was practically jumping up and down. I said, "The information revolution is only a local telephone call away to be experienced first hand for yourselves". I harangued them mercilessly on the urgency of getting connected. As far as I was concerned the Internet was the PC's raison, without which they were as the body without the spirit. I said that once they got the students on e-mail they would have more of a reason for learning typing and word processing and all the rest. E-mail was a motivator; e-mail had changed my life! I urged them to organize parent teachers association fund raising drives to pay for their connectivity. In the meantime they could go on-line at our workshops in the headquarters building of the ministry of education to get some hands-on experience.

Now two years had gone by since that first visit, and sadly little had changed. Sun Microsystems had donated a new server, but connectivity was still the bottleneck. We'd alerted ministry officials to the need, organized numerous workshops and seminars, but the students who, to us, represented the greatest hope for taking their society into the information age, still wanted the means to do so.

At a presentation I'd given to MPs in Uganda's Parliament to MPs in 1997, I projected web pages onto a large screen and showed how search engines would aid them in making better-informed decisions. The state minister for education was interested enough to ask probing questions, but others seemed only bemused. I could see we had our work cut out for us as far as sensitizing government officials.

In response to the need for a greater awareness of Internet technologies by government and the general public, I and my colleagues at Uganda Connect met with stakeholders from the Internet service provider community to form the Uganda Chapter of the Internet Society (ISOC). We organized public meetings where invited speakers discussed issues related to using Internet technologies for development. Our ISOC meetings received some good newspaper publicity and some were shown in the news on national television.

We also continued to make demonstrations at annual computer shows, and traveled to national fairs with our mobile communications truck. We also taught government officials both at our workshops in the ministry and privately.

It was an uphill struggle at the beginning, everything took longer than we expected and sometimes when the project seemed to get bogged we were tempted to believe that maybe our critics might be right, that the Internet was not a priority need in Uganda. But looking back I am sure that such presentations and demonstrations helped to lay the groundwork for building an information society in Uganda.

That head teacher who had first invited us to his computer lab later became the national coordinator for the World Links for Development programme. He was invited to make a presentation of the World Links programme at one of our ISOC meetings. And one of the ministers we had taught became a leading champion for Internet connectivity in government. Only a year after our first Internet Society meeting the level of interest and awareness in using Internet technologies for development had soared. Members of Parliament were now clamouring to know more; how could they get their constituencies connected?

We had helped to pioneer the use of GSM data in Uganda, and been working with CelTel, the national cellular telephone provider, in demonstrating Internet connectivity over GSM mobile telephone. Company directors had been reluctant to offer the service commercially fearing demand would be too low to justify the investment. But they had not reckoned on the soaring interest in the Internet. After one year of our demonstrating Internet connectivity over GSM they announced at the AITEC computer show in 1998 that their GSM data service was available commercially.

The headmistress of Namagunga Secondary School, one of the World Links schools near Jinja, east of Kampala, who was receiving training at our workshops told us that their telephone line was too noisy to support the connection to the Internet server in Kampala. She had paid for the phone line with her own money. She asked what we could do to help. We met with directors of CelTel to suggest that the company sponsor such schools with so many hours a month call time. They agreed, and by the time of writing several schools were enjoying 30 hours per month connectivity, by which students were able to browse the web at a walloping 9.6 kbps! As far as I know those were the only schools in Uganda able to do any appreciable browsing.

While in Geneva to present papers about the Uganda project at a couple of conferences I attended a meeting at CERN of the Geneva ISOC's Special Interest Group for Development. I was thrilled to hear Dr. Gideon Chonia[5] tell how a spread spectrum microwave radio network connected schools in Accra, Ghana, providing them with high bandwidth wirelessly, thus without having to pay the high telephone charges. The network was able to provide bandwidth to the schools inexpensively because businesses connected by the same network indirectly subsidized the schools network by paying a commercial rate. I think the thing that hit me right away was that very little investment was needed to provide for a 64 kbps link to one’s ISP. Previously that kind of connectivity could have only been provided over a leased line from the national telco, rented for thousands of dollars a month. It had taken a lot of copper and a lot of telephone poles and lot of years to build the older infrastructure. Now one could put up a small mast on the top of one building and connect another building tens of kilometres away in a matter of hours and for only a few thousand dollars. Wow!

I immediately set out to set up a similar network in Uganda. Since we run the project on a shoestring and our funding comes in small grants from companies, foundations and private donors, we regularly beg equipment from manufacturers allowing us to keep our costs down. I got to work right way contacting donors. We were already working on a project to connect up country telecentres by an HF radio data network[6] and the equipment we would need could be shared. At Telecom Interactive 97 I'd met with a director from Cisco Systems and asked whether they could donate a couple of routers for the project.

One of our team was meeting with a minister in Kampala to discuss the project, and I sent a draft a memorandum of understanding, as an e-mail attachment, proposing a spread spectrum network for connecting schools Uganda, in which we asked the government to arrange for concessions on license fees.

Apropo of which, at the ITU's World Telecommunications Policy forum in March of 1998 a colleague and I met with members of the newly convened Uganda Communications Commission (UCC). We urged them to take the lead in African telecoms regulatory policy towards the spread spectrum ISM bands by following the example of the U.S. and European Union in making them unlicensed.

The World Bank's Bob Hawkins was on the ANCARA stand together with our team for Telecom Africa 98 in Johannesburg. Students from one of Uganda's World Links schools would participate in a videoconference with students from Holland and South Africa. While the videoconference demonstrated future possibilities and showed the students' readiness to use some of the leading edge technologies for learning, I felt the application was a little premature. More appropriate, I thought, to provide the opportunity to many students at several schools to browse the web consistently. We discussed our idea of connecting the World Links schools by spread spectrum.

While working with a colleague in preparation for INET98 we had a request from track leaders for the Network Training Workshop (NTW) for some Cylink spread spectrum microwave modems. We contacted P-Com, who agreed to the free loan of four modems for setting up a network in Uganda. We would loan the modems to the NTW track leaders for demonstration.

Before taking the modems to Uganda we set up a network in Geneva, to run it through its paces so we'd be able to consult with Gideon and others if we ran into any difficulties. But we also demonstrated the network to sponsors, IT managers and communications experts from international aid agencies (photo below). In a conference call organized by our Swiss ISP partners, PSINet (who had provided the bandwidth for the Geneva demo) we discussed the project with the director of Uganda ISP partners SwiftGlobal, who readily agreed to sponsor the project by providing bandwidth.

We planned to demonstrate the spread spectrum wireless network at the at the annual 1999 AITEC annual computer show in Kampala. We had been able to bring some of the equipment which we would need with us, but much of it, including some recycled Pentium PCs donated by the Reuter Foundation, was still waiting to be sent by airfreight, pending our getting the necessary tax exemption certificates from the government. The government was very helpful, but the process is exasperatingly slow at the best of times, and time was ticking away.

In the end we were saved by our friends at NCR who loaned us the equipment we needed; another company loaned us a roll of Cat5 cable. By the Press and VIP day our communications truck was parked at the main entrance to the Sheraton hotel, linked to SwiftUganda's server by semi-parabolic grid antenna to our Cylink[7] microwave modems with a 64 kbps pipe. The truck also provided connectivity to the Internet Society stand inside the hotel.

While setting up the equipment in the truck I met with a UCC commissioner who told me that the body had decided to make the 2.4 GHz an unlicensed band. He was still a little concerned that ISPs who used the band would be able to police themselves through the voluntary disclosure of each one's microwave installations and frequencies on a UCC listserve. He hoped Uganda might thus avoid the loss of the band, as had happened in some neighbouring countries because authorities there had failed to properly regulate the use of the band.

At our Uganda ISOC stand we offered web mail addresses to visitors, sending them to the communications truck (photo below) outside where they could queue up to have a twenty-minute free browsing session on one of the five PCs connected by our microwave link. We were pushing spread spectrum, getting out the word that wireless connectivity was the way ahead for developing country Internet. Visitors to our stand included government ministers and an international television crew who filmed our demo to be included in a documentary.

Now that our LAN at the Ministry of Education workshops was connected by a 19.2 kbps microwave radio link our trainees began to migrate from concentrating on improving their word processing skills to Internet related skills - so much so that if our link goes down for any length of time, the number of trainees begins to drop off.

Students Starved of Knowledge and Information - Let the Children be Fed First

We knew that if we could only get some of the leading schools connected with a 19.2 kbps link, so that each student working on one of the PCs in their computer labs would be able to browse simultaneously it would cause an information revolution, first in the school, then in the community, and finally in the ministry. So far the revolution was only waiting to happen.

We decided to set up a demonstration school. The aim of linking one of the schools to the Internet by microwave spread spectrum was to demonstrate proof of concept, as already discussed, an essential step in sensitizing stakeholders, teachers, school heads, parents, aid organizations and sponsors, government officials, telco and ISPs, to gain their support. Once accomplished it was hoped that other schools would decide to participate in a schools network. A sustainable bandwidth sharing subscription scheme could then be set up between schools (initially with sponsorship) able to pay for itself through the cost savings[8] afforded by its more effective means of distributing web-based knowledge.

I visited with the head teachers of several of the World Links schools, accompanied by a Member of Parliament friend who was keen to facilitate the acceptance of the idea. While presenting the project we were careful not to give false hopes but only advised that such a demonstration school would be set up. Each head teacher was by now acutely aware that connectivity was the bottleneck, and so the project easily gained acceptance. While we were at the school we studied the topography to see whether the school was line-of-site to either our antenna at the ISP or to a planned repeater site on one of the larger hills surrounding Kampala.

Our visits to the schools coincided with a progress report released by the national co-ordinator for the World Links programme in which he'd indicated that all but one of the ten schools had some form of connectivity. We had already met with the headmistress of that school, who was most enthusiastic, but we were concerned that the topography was not going to allow the connection without some extra effort. We would need to set up a repeater site.

Our partners, SwiftUganda, were using the Wi-LAN Hopper and, in order to fit in with their overall plans for extending the network to other schools as well as businesses, we would need to acquire a couple of Hoppers. SwiftUganda were prepared to provide bandwidth for the demo, but could not to provide the additional microwave equipment we'd need. We got in touch with Wi-LAN in Canada, who immediately responded by offering to provide us with two of their top-of-the-line Hoppers[9] on free loan.

Bringing communications equipment into a landlocked developing country as Uganda is not easy. Although we had good relations with the government, including the customs authorities, the delay in clearing the equipment through customs was a setback. We missed a window of opportunity to set up the demonstration school before Jed[10], our system administrator, had to return to Europe, partly due to that delay, and also because one of our Cylink modems went down just when we might have completed testing the link to the school.

Running the project on a not-for-profit basis may have contributed to the delays, yet I’d have to say that it has served to help us to be more sensitive to the needs of the community and schools. For I know that if we are to be successful in putting together a package of connectivity, such as we envision, in which each school subscribes to the network with a fee that pays for bandwidth according to each school's use, it must be affordable by a sufficient number of the schools.

Training the trainers and teaching others to teach others are central to the project’s aims of creating reproducible and sustainable models. And so when we considered how such a network might pay for itself we had to consider the cost of maintenance and support. We’d had some encouraging experience over the years with picking up one skill after another that we needed despite some saying it couldn’t be done, that we ought to let the experts do it for us. When we started the project we didn’t even know the basics of DOS and Windows, but took the plunge and soon found out that they could be learned. With the help of some friends we learned enough about them to get the job done, even to be able to teach others. Then we learned the basics of office suite programs so we could use them and help others use them too. Configuring dial up networking and TCP/IP seems now to be almost second nature to us, but we got lots of help from friends at our first ISPs in the beginning. Our colleagues at the World Food Programme's Technical Services Unit in Kampala continue to coach us on the intricacies of the HF radio data network.

And so we knew that somehow we’d be able to get the spread spectrum microwave radio network up and running. We consulted Gideon Chonia (photo below) at an early stage, and then PSINet’s system administrators when we needed help configuring the routers; we studied the equipment manuals, and we did a fair amount of scary work adjusting semi-parabolic grid antennas very high up. I now have the conviction that if we, who were not specifically trained to be electronic or communications engineers, could somehow pick up the necessary skills in a short period of time, in order to be able to install these kind of networks, just about anyone can, with the right kind of assistance and support.

Jed, had been working with technicians from SwiftUganda to test the link. Arrangements were made to install a microwave radio on the Uganda Television antenna mast. The school’s management had approved needed expenditures for installing a small mast as well as making necessary improvements to security. Jed had been training the Uganda Connect team to be able to troubleshoot the link should problems arise while he was out of the country. Within one week of Jed’s departure for Europe the link went down, and the team were able to run through a checklist Jed had left with them, received further instructions from Jed by e-mail, and managed to get the link up with only a little help from SwiftUganda’s technical support team.

We are nearly ready to send another team to Uganda, including Jed, who plans to sort out the difficulties he had encountered with setting up the link to the school. To give you a better idea of what that will entail I include a portion of his report:

"My previous testing of the link was cut short by technical problems at the repeater site on Uganda Television mast. The repeater site consists of a Wi-LAN radio receiving its link via a directional antenna from the roof of the building housing the ISP. A second Wi-LAN radio pipes that link into an Omni directional antenna serving an approximate radius of nine kilometers.

"Gayaza High school is located on the outer limits of that 9km radius. The most cost effective solution would be to use one Wi-LAN radio at the school with a directional antenna to receive the link. For this to work we need a clear line-of-site from the school to the repeater site. The clearest line-of-sight is at the school's gatehouse. The gatehouse about 200 meters from the computer lab. Without the use of a hub, the maximum range of the Cat5 cable we will use to connect the radio with the lab is 100 meters. Since there is no building at the point where would need to install the hub, a weatherproof box with power supply would have to be installed for a hub. Another inconvenience is that the main drive passes between the gatehouse and computer room, and therefore conduit would have to be laid under the asphalt.

"The school administrators have agreed to make the necessary installations. But we must first test the link so see if the signal from the gatehouse is strong enough. If I find that we have a very strong signal at the gatehouse, there is a possibility that we may be able to set up the radio from a more convenient location for which the line-of-site to the repeater is slightly obscured by trees. That site is a classroom only 75 meters from the computer lab. If the signal is strong enough from the classroom then we would no longer need to install a weatherproof box nor tear up the school's drive.

"If it turns out that the signal at the gatehouse is not strong enough, then we would have to use the Cylink modems to do a directional-to-directional link to the repeater site. This would entail our installing a Cylink and Cisco 1601 router on the Uganda Television mast. And since this is almost certain to provide a good strong signal I would be able to set up the link on the classroom roof."

I am looking forward with great expectancy to seeing the link to the school up and running.

I had thought to entitle this paper "How We are Overcame the Hurdles". But even after we have successfully connected the demonstration school, shown proof of concept and gained enough support from stakeholders, teachers, students and parents, business, aid agency and government officials, to provide a high bandwidth network to several schools in the Kampala area, much work will still lie ahead. I am sure that the benefits enjoyed by students and teachers at those schools will show that much, much more needs to be done to enable schools throughout Uganda and other developing countries to be filled with the knowledge and information that is offered by the global Internet. I've tried to show how our NGO overcame a number of obstacles in bringing knowledge networks within closer reach of students and youth in a developing country, those most in need. I hope that I will have encouraged others to help to make it happen.

[1] Uganda Connect, or the Uganda Connectivity Project, is run by the NGO and UK charity, Mission Mobile Education www.uconnect.org

[2] The majority of better equipped schools in Uganda have only one copy of a textbook for each of their teachers, the contents of which is shared with students by writing it out on a blackboard. Many schools are without electricity.

[3] A significant gift considering the local price for a new modem at the time was a few hundred dollars.

[4] Local calls to the server were 75 Ugandan Shillings a minute, about $4.50 an hour. But besides the actual cost of phone calls school administrators had to also consider the possibility of telephone fraud leading to an unduly large bill. And with no itemised bills, there is recourse to victims but to pay or be cut off.

[5] Dr. Chonia is a Ghanaian lecturer from University of Zurich's computer sciences department.

[6] Our HF radio data network also provides Internet (e-mail only) connectivity to rural communities, including schools. To read more about this other form of wireless Internet connectivity see Upcountry HF E-Mail Network As an Early Component of a Developing Country's Information Infrastructure by Daniel Stern www.uconnect.org/hfnetful.html

[7] Cylink AirPro 64 http://www.p-com.com/products_frame.html

[8] Instead of each school having to pay the $60 monthly fee to the ISP they could pool together to buy their bandwidth, distributing the bandwidth wirelessly, thus bypassing the expensive and inefficient telco network.

[9] Wi-LAN Hopper Plus 45-24 http://www.wilan.com/products/specs/plus4524.html

[10] Jedidiah Stern was appointed system administrator for the project after attending Track One of the INET98 Network Training Workshop. Jed is my eldest son.

Let the Children be Fed First: How We are Overcoming the Hurdles in

Providing Knowledge Networks for Schools in Uganda