The Network Startup Resource Center (NSRC)

Randy Bush

July 1996

National Science Foundation Grant No. NCR-9616597


The Network Startup Resource Center (NSRC) has proven to be an efficient, cost-effective mechanism for facilitating the establishment of computer networks in developing countries and for enhancing network operations globally. The existence of these networks has positively affected the US scientific community and US agencies, as well as the local communities in which the networks have emerged. By providing technical assistance to numerous countries around the world, and tracking international connectivity developments, the NSRC has established and maintains an extensive base of contacts willing to contribute their time and expertise to further these efforts. The NSRC's emphasis on training, gathering tools, disseminating information and providing technical support has contributed significantly to the development of sustainable networks, managed by local hands with local expertise. Widespread recognition of the importance of access and of the accrued benefits of networking is a recent global development, which the NSF and the initial NSRC project have successfully helped to promote. This proposal seeks NSF support to continue our efforts for another three years, in assisting the establishment and expansion of developing country networks.


The Network Startup Resource Center (NSRC) currently functions as a virtual global coordinating clearinghouse and service center working with individuals, organizations, and governments worldwide. The NSRC began as a volunteer effort in 1988. As the original volunteers lent their time and expertise to those requiring assistance to start computer networks, it quickly became clear that there was considerable need for the information the NSRC was providing within a large community having neither access to, nor funds for, private consulting assistance. (Indeed, insufficient funds among those who wish to establish networks is still a serious obstacle, and the NSRC receives many requests for funding.)

To date, the NSRC has contributed significantly to the creation of national or regional networks in Africa, Central America and the Caribbean, Central and Eastern Europe and several former Soviet states, and in South, Southeast, East, and Central Asia as well as to some start-up K-12 network systems in the US. Central to the success of the NSRC has been its philosophy of empowering local users - thereby fostering independence and, most important, sustainability - together with an emphasis on the use of appropriate, affordable, and scalable technology.

This new NSRC proposal would receive cooperative support from network services staff at the University of Oregon. Joanne Hugi, Director of University Computing, and Dave Meyer, Director of the Advanced Network Technology Center (ANTC) have offered their pro bono assistance and access to University Computing staff, as necessary. The ANTC staff is engaged in leading-edge research, engineering, and development of next-generation Internet protocol technologies. Recently, the UO's Network Services staff assisted the NSRC by hosting an intern from Kenya.


The Network Startup Resource Center (NSRC) helps US scientists gain ready access to international resources by nurturing networking in areas of the world in which computer-based communications have traditionally been limited. The NSRC gathers, organizes, and disseminates information, training, and tools to local organizations so they can acquire affordable networking technology and use it to join the Internet.

This proposal seeks funding for continuation and expansion of the NSRC's current activities and for new projects. Some of these contribute directly to network establishment and expansion, some to network management and administration, others to improving network efficiency and connectivity by improving the technology and our understanding of available technologies. This also includes continued maintenance and improvement of the NSRC connectivity database (, which disseminates information via the World Wide Web, gopher, and automated mail servers.

Statement of Need

Widespread recognition of the importance of access and of the accrued benefits of networking is a recent development, one which the NSF and the initial Network Startup Resource Center project have helped to promote. US research and engineering scientists rely on the Internet for collecting data from remote locations and for collaborating with R&E scientists in the developing world. Scientists travelling abroad use the Internet for communicating with colleagues at their home facilities in the US. The ease and timeliness of communication permitted by the dynamic functioning between computer networks contributes significantly to the success of US scientific endeavors, and benefits all others involved in US efforts around the world.

The NSRC has made significant improvements in the consistency and availability of data, tools, discussions of alternatives, and other documentation, but there is still more work to be done in this area. Specific examples include a need for detailed technical and development model histories, which would provide useful information to those currently engaged in developing new networks; documentation and publication about last kilometer technology alternatives (i.e., packet and spread spectrum radio), including writing RFCs and WWW pages, if appropriate; and publishing information, insufficiently documented elsewhere, about the complex administrative and managerial steps obligatory to establishing and maintaining a network today.

Results from Prior NSF Support (NSF Grant No. NCR 9216064)

The NSRC was awarded an initial three-year NSF grant through the International Nutrition Foundation in October 1992. The purpose of that grant, which expires in September 1996, was to formalize the NSRC, which until 1992 had been staffed entirely by volunteers. With the help of the initial grant, the NSRC has built an organizational base for disseminating networking information and expertise to many recipients around the world, during a period when interest in networking has grown at an unprecedented rate world-wide.

Examples of achievements of the original NSRC proposal include:

o Development of toolkits for low-cost IP, UUCP, Fidonet, and related networking technologies, made available via FTP, Gopher, UUCP, WWW, mail-based archival services, and postal mail;

o Tracking and archiving documentation on the startup and early phases of networking in the developing world, and making this information available by the above means;

o Coordinating the provision of networking textbooks and documents to information-deprived areas of the world;

o Assisting US scientists travelling abroad with nurturing nascent networking services;

o Lobbying for the needs of emerging networks in the fora of international network engineering and administrative decision-making institutions; and,

o Hosted interns from Peru, Sri Lanka, Kenya - all of whom went home to establish their national network.

In March 1995, the NSRC received a supplement to the original core grant from the NSF's Research Experiences for Undergraduates (REU) program. Two University of Oregon undergraduate students worked with a UO graduate student and the NSRC project principals to move the archived gopher data to an interactive WWW site. The enhanced NSRC data repositories incorporate a set of clickable image world maps with better tools for indexing and accessing information.

Some examples of NSRC planning and engineering assistance to country networking initiatives include:

o Designed, taught about, and helped deploy a multi-country (South Africa, Botswana, Namibia, Zimbabwe, and later many others) network using various technologies, in order: Fidonet on dialup lines, UUCP on dialup lines, low-cost IP technology based on 9600 baud and below links using old PCs and publicly available PC-based SLIP routing software, and finally multiple dedicated medium speed (128-512kb) links to the US;

o The UNDP and Union Latina chose Randy Bush and the NSRC to do the technical work and training for the first networking within Peru and to establish the links to the US. The NSRC was Peru's UUCP link to the outside world for over two years until Peru was able to upgrade to a 64kb and then 128kb satellite internet link;

o The NSRC was chosen by the NSF to design, install, train, etc. for the first IP links in Kenya;

o NSRC staff organized, designed, and educated the engineers for Sri Lanka's entry into the Internet;

o Bush and NSRC affiliates were chosen by the US National Academy of Sciences to catalyze and design the Indonesian academic internet, IPTEKnet; and,

o The NSRC provides cost-recovery only PPP and UUCP links for R&E sites and NGOs in Belize, Cambodia, Guinea, Guatemala, Kenya, and around the US. In doing so, the NSRC is careful not to compete with commercial providers, or with NSFNET connectivity arrangements; we try to act as a stepping-stone to such more "normal" arrangements.

The NSRC's enthusiastic reception validates the organization's goals set forth in the original grant proposal. The current proposal is to continue the activities of the NSRC in the same areas, with a few changes and extensions:

o As the worldwide network infrastructure develops, first time networkers are bypassing low-end, minimum-cost solutions in favor of dedicated connectivity. "Entry level" now begins at a higher point, technically, than it did in the past.

o The NSRC's users have facilitated many startups, and are able to identify patterns that can be described and documented. Transcribing this information will expedite startups and provide useful models and lessons for others.

o The NSRC's user base is maturing, so the organization finds itself devoting an increasing proportion of its resources to issues of network management and longer-term development.

The following sections describe specific activities in which we propose to engage.

Proposed Activities

Establishing Local Infrastructure

US researchers cannot and should not need to bring the Internet with them wherever they go or wherever they need data. Typically when they do so, they spend excessive time organizing networks rather than doing research in their substantive areas. Costs tend to be higher than necessary because such single-user links cannot benefit from the economies of scale of shared infrastructure. In addition, networks that are perceived as being set up by and for "outsiders" tend to collapse as soon as the outsiders depart.

For new installations, the NSRC's focus is to foster and assist in the establishment of networks in place and to train local engineers to maintain them so connections become permanent rather than transitory. As experience and demand accumulate, connections become permanent rather than transitory. For maximum benefit to the US scientific community and their colleagues around the world, development and deployment are in local hands, using local expertise, and primarily at local expense.

When a visiting US scientist does set up a network facility, that facility should be structured so that it can expand into something permanent, serving local users and linked with other facilities in the region, rather than providing only a "call home" capability. In this way, the next visiting scientist is less likely to have to repeat the earlier effort, which would be a considerable waste of time and resources.

Technical Assistance and Coordination

The critical path toward facilitating network installation generally does not lie in buying equipment but in providing information: arranging for databases of, and easy access to, information about regional hubs, potential collaborators, technical information and alternatives, putting potential donors together with potential refurbishers and recipients, etc. An important part of this is drawing information together about efforts going on in parallel that might be coordinated in ways that leverage both into higher-quality services.

At the implementation level, the NSRC proposes to continue providing information in several ways to those who need it:

o Helping US scientists and network support personnel obtain data regarding availability of overseas connections. As a means of propagating this information, we maintain a World Wide Web site at, and we continue to respond to numerous email inquiries each week about the status of networking in country X;

o Sending books to engineers setting up supportive networks in developing countries;

o Answering technical questions and giving technical advice to engineers in developing countries who are maintaining and/or setting up networks that US scientists use;

o Facilitating top level domain (TLD) registrations;

o Maintaining and participating in mailing lists that support US scientists seeking connectivity overseas, (for example, the venerable INFO-NETS list) and lists in support of improving connectivity overseas (TCPWS); and,

o Lobbying for scalable technology at IETF and other internet design fora.

Producing New Documentation

In order to expand our activities, we are requesting funding for a project coordinator and three student assistants, whose tasks will be to write additional documentation, maintain and supplement the information contained in the data repositories, and otherwise assist the project principal. As part of this proposed award, we would seek REU support again as a way to provide network training opportunities for students and complete projects in a cost efficient manner.

Some of the topics we plan to document and publish include:

o Detailed technical and development model histories of successful networking in support of R&E in developing countries. These provide short but detailed models and lessons, both positive and negative, to those starting new networks. We have learned that case studies of "how others have done it and how to learn from that" are as important as specific instructions on how to use particular tools. The case studies, which will be prepared by NSRC staff and users, will be disseminated electronically via the World Wide Web, gopher, mail servers, and in paper format when electronic delivery is not possible;

o Descriptions of technical aspects of networking technology that are insufficiently documented elsewhere. For example, "last kilometer" technology alternatives (packet and spread spectrum radio), and mailbag protocols and techniques to radically reduce bandwidth requirements on expensive links, including writing RFCs if appropriate, WWW pages, etc.;

o Administrative procedures needed for establishing working relationships with the infrastructure in the US. Such procedures are mandatory in performing critical-path operations such as domain registration, obtaining and registering network address space, and routing database registration. We propose to help startup or expanding networks with navigating this increasingly complex maze of administrative procedures and routing registries, and assist them with completing the various templates, forms, and registrations.

Hosting Engineer Interns

Books and workshops are critical, but hands on experience in larger operating networks gives an engineer perspective and confidence which is not otherwise available by any other means. We have been doing this informally over the last years, and the results have been very positive.

A typical internship varies in duration from two weeks to two months, and consists largely of sitting shoulder to shoulder with peers operating real networks in the US. Examples of past internship activities include UNIX training, building BSD UUCP systems and SMTP gateways for UUCP, building and configuring a T1 POP, constructing prototypes of systems to be installed in home country networks (in the cases of Sri Lanka and Kenya, the actual systems were built and shipped after the engineers were trained on them), touring US university networks, and attending IETF.

The University of Oregon has assisted with hosting and training network engineer interns and would continue to do so under the auspices of its Advanced Network Technology Center. Housing and other living expenses would have to be provided on a cost recovery basis outside of the costs enumerated here.

Organizing Workshops

In-country training has proved to be the most effective tool for bringing new network staff up to speed. The NSRC staff and, more recently, many of their former students, have been involved in numerous training workshops around the world. The goal of some of these sessions has been actual installation of a network connection, while others have taken place in locations where network connections are already in place. Although local technical conditions have not always been optimal, face-to-face interaction between trainer and trainee has consistently proven useful.

NSRC affiliated engineers currently organize and participate in workshops for engineers from new and expanding networks in developing countries such as the INET workshops, NATO in the former Soviet Union, and in Latin America. NSRC would be willing to organize and help staff task-specific technical workshops at the request of NSF and other organizations that promote R&E networking in the developing world.

Workshops are not specifically budgeted as costs are extremely variable. As an extension of this task, the NSRC will, at the direction of and in collaboration with NSF, utilize supplemental resources to specifically seed network training and development activities in selected critical areas.

Direct Engineering and Planning Assistance

NSRC affiliated consulting staff have designed, installed, trained engineers and network managers, etc. for networks in the developing world, e.g. Kenya, Peru. We have done this in response to specific requests from the network managers in these networks.

Consulting is not specifically budgeted as costs are extremely variable. Each case has been different, and the funding agencies have varied from NSF to the World Bank, AAAS, and others. At the direction of and in collaboration with NSF, the NSRC would help locate training resources and appropriate personnel to provide engineering training for networks requesting assistance.

Low-Cost Startup Solutions

We are seeing interest in FidoNet, UUCP, and other message-forwarding technologies rapidly give way to a focus on TCP/IP. Efforts are in progress in many areas that had moved to FidoNet or UUCP connectivity only a few years ago to arrange more robust and dedicated TCP/IP connectivity; many of the latter type of links have already been installed and are running. Fueled by the explosive interest in the World Wide Web, non-Internet technologies have simply become uninteresting in most of the world, even in situations where they might still be appropriate from a technological perspective. If there is any possible way to install Internet connections, neither US scientists nor their international collaborators are willing to settle for other networking approaches - approaches they now consider second-rate.

We will, however, continue to make available non-IP technologies for those situations in which better solutions are not yet possible.

The NSRC makes available a variety of installation kits that permit users (including those who are not computer experts) to quickly and efficiently establish network connections. The focus of our efforts in this area will be improvements to documentation and automated procedures, which are as important as improvements to the technical networking parts of the kits. In keeping with trends toward ever-higher entry levels, we will also be more concerned with kits for initial IP installations.

If current trends in connectivity patterns continue, some installation kits that focus on configuration of national-scale routers and security mechanisms are likely to be desirable within the performance period of this proposal; we would seek and collect such toolkits, and try to make them available.


The Network Startup Resource Center (NSRC) has proven to be an efficient, cost-effective mechanism for facilitating the establishment of computer networks in developing country areas and for enhancing networking activity globally. The development of computer networks has had a demonstrable positive impact on the US scientific community, and other US agencies, as well as on local communities in which the networks are established.

The NSRC brings considerable experience and perspective in designing and deploying appropriate technologies in networking. We have also established and maintain an extensive base of contacts who are willing to contribute their time and expertise to enhancing network operations globally, as evidenced by their participation in past NSRC activities.

We propose minimal funding for another three years to continue assisting the local engineering and networking infrastructures abroad, by making available and transferring by any reasonable means - appropriate tools, technology, information, documentation, and training to improve global networking operations.