============================================================================== ISOC Document 94-333 Title: Facilitating Internetworking in Africa Author(s): Randy Bush Date: 1994.12.05 Body: Board of Trustees Document: 94-333 Revision: basic Supersedes: - Status: Draft Maintainer: Randy Bush Access: unrestricted ---------------------------------------------------------------------------- Facilitating Internetworking in Africa a Proposal by The Internet Society Bush, Sadowsky, Landweber Goal: Empower Africans in the Electronic Information Age o Empower academics, NGOs, public and private sectors o First-class information citizens, not Nth class technology o Open networking, not closed private network enclaves or closed technology o Standard technologies, user and engineer knowledge are transferrable in and out o Cooperation with and enhancement of existing open infrastructures Means: o Technology: Transfer, training, . o Empowerment: To be eventually owned and operated by Africans o Sustainability: Self-funding models adapted and localized o Extensibility: Continuing interchange with the technology developers o Bottom-up, network academic etc. because that is what has worked in Africa Context Issues: o Assisting in-country partners with respect to negotiations with, education of, and consensus building with relevant Government and PTT agencies o Working with our partners to help them educate these agencies regarding the economic, social, and technical benefits of the Internet Examples o Network the main university to get most critical users and a source of engineers o Statistical and research ministries o Be available to critical development NGOs o Build an infrastructure which can be extended and attached to by others o Leave sockets for commercial, k12, so other projects can exploit, expand, and enhance o Build the critical roads, not all the towns The Basic Physics o A community of USERS to be served o ENGINEERS provide services, training, o PERMANENT NETWORK CONNECTION to the world over which to provide the services to the users and continue aid and assistance to the engineers o If the users are served and the cost-recovery model is sound and well-suited to local conditions, growth is well-assured User Training o DOS, Windows, and Mac clinets, not UNIX o Train user trainers and initial evangelists o Domain-specific training, librarians train librarians o Provide initial facilities and help acquire permanent facilities o Help Adapt material to local environment o Help establish collegial relationships with domain specific experts in developed countries; setting up the human links Engineer Training o Hands-on, in-country along with regional and international workshops o Engineers train engineers o Continued presence of engineer-trainers during deploymentand early follow-on phases o Continuing participation in the international community o UNIX, TCP/IP, uucp, DOS/Mac clients Ongoing International Workshops o Internet Society in-depth training workshops held annually in 1993, 1994 - About 300 paticipants from 90 countries trained intensivelyin dial-up, TCP/IP, network navigation and services, and national network management - Value in continuing such training, including regional events - Multi-country training strengthens self-sufficiency in region - Global workshops facilitate melding specialists into worldwide context and community of colleagues Add Regional Workshops o Work with groups in regions to promote and support in-depth training workshops o Identify regional and subregional training opportunities in Africa, leveraging the experience of early adopter countries to train others o Encourage migration of workshop training to national level as national responsibility as infrastructure evolves to support it Deployment o Appropriate in-country technology, but first-class, gateways to oddities (FidoNet) o Major metropolitan area(s) at first o Dial-Up for end-user clients on DOS/Mac o Dedicated IP where at all possible, by radio if necessary, and packet-radio for distant secondary sites o Store and forward IP and uucp where dedicated does not work o International links by 64kb satellite Three Stage Deployment Model o Setup: - Making friends - initial training, design, o Deployment: - IP links - Network Operations - User Sites, - User training, initial and training trainers o Continuity: - heavy and continuing end-user training - business issues of running a sustainable network - expansion of infrastructure and of the goods delivered Initial Setup o Finding partners, users, engineers o Initial engineer training o Training and cost-recovery model o Deployment engineering plan o Store and Forward temp comms links o NIC coordination Deployment o User Sites & User Training o NOC installation & Engineer Training o Earth Station o PTTs, radios, modems, o LANs at User Sites o Interface with NICs o Maintenance plan, repairs, spares, Ongoing Support o Support for continuing training of a large number of users o Business and cost-recovery issues to make a sustainable network o Integration of in-country engineers and domain specialists into world networking community with international net culture o Building in-country information providers o Countinued assistance as needed to assure self-reliance and self-sufficiency Cost Recovery Model o Start-up costs subsidized o Comms costs flat subsidy o Charge for servcices. If users won't pay, then you are providing the wrong service Evaluation and Flexibility o Criteria for success: - users: number of users and institutions, pattern of use - self-sustainence - ability to expand under local control and financing o Learning and Adapting - Complex inflexible plans are a primary cause of project failures in the experience of major development agencies - Advisory Council evaluates quarterly and gives tactical and strategic guidance - Alliances with existing development groups help to localize and make appropriate strategies Why this Team? o ISoc has deep international engineering contacts: IETF, IAB, anda Track record of Developing Country Workshops o Landweber has worked on infrastructure development and technology transfer for the past fifteen years o Bush has track record deploying networks in developing countries o Sadowsky has experience geting technology transfer results from large bureaucracies